The gender gap in political efficacy: The accelerating effect of classroom discussions

The gender gap in political self-efficacy originates early in life and is among the most persistent inequalities in political attitudes across Western democracies. The difficulties in accessing data for early adolescents have resulted in limited research available to provide an understanding on how gender differences in political self-efficacy are developed. In this article we provide unique evidence on the development of the efficacy gap during secondary school. The Citizenship Educational Longitudinal study data from England on young people between the ages 11–16 shows that boys increase their levels of political self-efficacy significantly during this period while girls do not. Secondly, we show that teaching methods that are expected to enable the development of internal political efficacy, such as classroom discussions, are accelerating the gender gap in perceived abilities to get engaged in politics. Our findings show that, despite having the same access to classroom discussions in schools, this experience has a positive effect in the development of political efficacy for boys but, if anything, a negative effect for girls. Furthermore, even a small number of boys in the classroom discussions appears detrimental to girls' political efficacy.

Introduction

Decades of research have consistently shown that women have less confidence in their skills to influence the political realm than men (Bennett 1997; Gidengil et al. 2008). In Europe, data from the European Social Survey show a substantial gender gap in political self-efficacy across all 23 countries examined (Fraile and de Miguel 2022). The scarce empirical evidence on the development of the gender gap on political self-efficacy points to an origin in early adolescence (Husfeldt et al. 2005). Equally, it is these early years that are particularly formative for young people's socialization into gender roles (Blakemore et al. 2013; Bos et al 2022). The theory of gendered political socialization has recently been identified as a promising approach for research addressing when and how gender gaps in political attributes appear (Fraile and Sánchez-Vítores 2020; Bos et al. 2021). Yet rigorous statistical research on gender and political socialization is rare (Lawless and Fox 2010). Among the potential causes of this lacuna of knowledge are the high costs and complexity of undertaking research with children and young adolescents. In this article we use data that enables us to use repeated measures for the same students over time during early adolescence.

Despite the difficulties in studying young adolescents, there is a great deal of potential in this approach. Identifying the socialization experiences that are having differential effects on the learning of political self-efficacy for girls and boys¹ would provide the first step in attempting to address inequalities in political self-efficacy. One important socialization experience for learning the attitudes and dispositions for political engagement is the school (Keating and Janmaat 2016). Identifying the role of school is important as with universal education provision, at least in Europe, this institution provides a possible policy lever to create change and to reduce the gender gap. Furthermore, scholarship on political socialization in schools has shown that a specific teaching method, in which teachers lead discussions in class through the creation of an environment where students feel able to talk about political and social issues, referred to in citizenship education literature as an open classroom climate, promotes the development of internal political self-efficacy (Hoskins et al. 2021; Kuang, Kennedy and Mok 2018; Claes et al. 2017 and Knowles and McCafferty-Wright 2015; Torney-Purta et al. 2001).

The contribution of this article is three-fold. First, to the best of our knowledge, this is the first time that empirical evidence shows that not only is there already a gap on political efficacy between boys and girls at an early age, but also that this gap increases significantly during early adolescence. The second innovation is the combination of literature on political efficacy from multiple disciplines to create a holistic understanding of the development of this phenomenon. It brings together knowledge and theories from political science and sociology on gendered political socialization (Bos et al. 2021) that explain how gender and political socialization work in tandem to recreate gendered roles in politics, together with education and social psychology research on how self-efficacy is developed from a gender perspective (Bussey and Bandura 1999). In addition, the article draws on citizenship education research highlighting a teaching method on classroom dynamics, the open classroom climate, which is associated with increasing political self-efficacy (Torney-Purta et al. 2001). The third contribution of this article is showing that the popular teaching method, the 'open classroom climate,' does not reduce inequalities but rather contributes to the development of the gender gap in political self-efficacy. Specifically, boys significantly increase their levels of internal political efficacy from this experience, whereas girls do not.

The findings are derived by analyzing the Citizenship Education Longitudinal Study (CELS) data from England on young people between the ages of 11 and 16. The panel structure of the CELS data allows the possibility for analyzing the effect of an open classroom climate on levels of internal political efficacy by controlling the previous levels of this same measure from the year before. The results indicate that perceived access to an open classroom climate is similar for boys and girls, but the experience of the open classroom climate has a large positive effect for boys only. Girls experience a negative effect from an open classroom climate on their levels of political-self efficacy when in the same class with boys; and when girls are taught on their own, the experience has no effect on their levels of political self-efficacy. The results are interpreted using theories on gendered political socialization (Bos et al., 2021) and gender and social cognitive learning theory (Bussey and Bandura 1999). The implications for classroom and teaching practice, as well as for our understanding of the persistent gender gap in internal political efficacy, are then discussed.

The gender gap in internal political efficacy

As introduced before, the gender gap in perceptions of individuals' own capabilities to participate, understand and influence the political system is persistent, stable over time, and generalizable throughout Western democracies. For instance, Gidengil et al. (2008) shows a large and statistically significant gap in self-perceived political understanding from 1952 to 2000 in the US and from 1965 to 2004 in Canada. Surprisingly, despite changes in overall levels of internal political efficacy over time, the difference between men and women remains pretty much the same. When it comes to the EU, Fraile and de Miguel (2022) show a statistically significant gap in internal political efficacy across 23 countries included in the European Social Survey. Furthermore, when political efficacy and political interest are taken into account in later life, then there are rarely any gender differences in any form of political engagement (Verba et al. 1997).

Political efficacy refers to the feeling that one can have an impact on the political process (Lane 1959; Campbell et al. 1960[1980]; Milbrath 1965). In a further refinement of the concept, a distinction was made between internal political efficacy—capturing an individual's level of self-confidence in her own ability to politically engage—and external political efficacy—the expectation that the political system will be responsive to an individuals' demands (Lane 1959: 149; Kahne and Westheimer 2006). In this article, we are focusing on internal political efficacy or political self-efficacy, the language used for the same concept within education and psychology research. Political self-efficacy is crucial as it has been repeatedly identified as one of the major determinants of future political engagement (e.g. Almond and Verba 1963; Hope and Jagers 2014; Levy and Akiva 2019).

Recently in political science the focus for understanding the gender gaps in political involvement has increasingly moved towards gendered political socialization (Bos et al. 2021). To start with, young people grow up in an environment in which politics is still a male dominated world (Jennings 1983; Lawless and Fox 2010). Yet the research also shows that girls need role models as the presence of female politicians are found to encourage girls to engage in political discussions and political engagement (Campbell and Wolbrecht 2006; Wolbrecht and Campbell 2007). Furthermore, children and adolescents are socialized differently and predominantly according to commonly held perceptions of their biological sex and corresponding gender. Boys are typically steered to be confident in their abilities to lead and take up public roles whilst girls typically are steered towards identifying their strengths in terms of compassion and caring activities (Eagly and Wood 1991; Fraile and Sánchez-Vítores 2020). Most recent evidence shows that children not only perceive politics to be gendered, but that differences in political interest and attitudes increase with age during childhood (Bos et al. 2021). These dynamics also influence how they relate to the political world. Socialization in gender stereotypes such as leadership, conflict avoidance or risk aversion could explain women's lower levels of internal political efficacy. For instance, Wolak (2020) shows that self-confidence is an important psychological asset that promotes, among others, feelings of personal competence in politics. Furthermore, women declare higher levels of conflict avoidance than men and, not only that, but men also report to have comparatively higher levels of enjoyment of arguments and disagreements (Wolak 2020), which are common characteristics in politics.

When it comes to internal political efficacy, the scarce empirical evidence available for children and adolescents also suggest looking at early socialization to identify the roots of the gender gap. To start with, the limited literature points to an early origin of the gap. Husfeldt, Barber and Torney-Purta (2005) find significant differences in levels of political self-efficacy in favor of boys in 23 out of 28 countries among 14-year-olds (see also Barber and Torney-Purta 2015). Secondly, among adults and contrary to other attitudes such as political interest, political self-efficacy does not seem to increase with age once in adulthood (Fraile and de Miguel 2022). Finally, extant empirical evidence shows an early development of somewhat related basic political orientations such as political interest (Prior 2010; Neundorf et al. 2013; Fraile and Sánchez-Vítores 2020; Janmaat and Hoskins 2022). In summary, political socialization research has repeatedly shown that basic political orientations develop early in life, during the socalled impressionable years and have a strong persistence across an individual's life span (Kinder and Sears 1985) and only change in the event of dramatic changes or radically new circumstances (Dinas 2013). For these reasons, what happens at an early age appears to be key to understanding women's relatively low confidence in their ability to influence the political system.

Nevertheless, the detailed study of early age, the developmental process and gender dynamics in political efficacy is surprisingly absent. Socialization and political socialization research were at its peak in the late 20th century and only recently, research has moved back to children and adolescents to understand the origins of the gender gap in political interest, political ambition, and political knowledge (Van Deth et al. 2010; Abendschön and Tausendpfund 2017; Bos et al. 2021). A major limitation in this field is the scarcity of longitudinal datasets that follow children and adolescents over time, and thus our study is the first, to our knowledge, to provide empirical evidence on how levels of political efficacy for boys and girls evolve over time. In the section that follows we explore how the gender gap in political efficacy develops at an early age and then discuss the potential of school classroom dynamics to compensate—or accelerate—the differences between girls and boys.

Where does the gender gap in political self-efficacy come from?

To explore the origins and evolution of the gender gap in internal political efficacy we draw on theories from social psychology and education and in particular social learning theory, which has extensively theorized the dynamics in which children and adolescents learn gender roles and general self-efficacy.

In a broader sense than internal political efficacy, a generalized sense of self-efficacy has been defined by Bandura (2010 p.1) as "people's beliefs in their ability to influence events that affect their lives" and it was this broader concept that has then been narrowed down and used in the field of politics to generate the concept of political-self-efficacy/internal political efficacy. Self-efficacy forms a crucial part of the personal agency that individuals are said to have within any given environment and provides the motivation to act (Bussey and Bandura 1999). Bussey and Bandura (1999), using the social cognitive theory, identify four ways of learning self-efficacy. The first way is through graded mastery—"success builds a robust belief in one's personal efficacy; failure undermines it" (Bussey and Bandura 1999; Bandura 2010). In terms of learning political self-efficacy in the classroom—this can be achieved by providing opportunities for young people to voice their opinions in the classroom beginning with less complex topics and gradually increasing the difficulty to build the levels of confidence of the students in a student paced manner. The second method of learning self-efficacy is through observing people similar to one's self succeeding at the activity in question (Bussey and Bandura 1999). In this article, this would be watching other students of one's own gender having their voice heard and speaking regularly in the classroom and seeing the positive reaction (or not) of others to their performance. The third method of increasing self-efficacy is through significant others, parents, teachers or peers positively or negatively appraising the individual's own performance. In the classroom the reaction of other students and teachers to an individual's attempt to voice their thoughts in the classroom will influence the development of their political self-efficacy. The final method of developing self-efficacy is through psychological therapeutic interventions but we propose that this is beyond the scope of this article.

Can these processes explain how gendered socialization results in inequalities in political self-efficacy? The learning of gender is argued to occur through parents treating children differently according to their particular understanding of the child's biological sex (Blakemore et al. 2013). For example, a rigorous study showed that while there are no scientific differences between the crawling abilities of babies according to sex, mothers believed that their sons would be able to crawl down steeper slopes than their daughters (Mondschein et al. 2000). Fagot (1977) in a major study in the field researched toddlers at a nursery school. At the start of the study the girls and boys behaved in a similar way but the teachers responded differently. Teachers during that year responded to girls positively when they were gentle and quiet and ignored them when they were aggressive and loud. In contrast, teachers in the study gave attention to boys when they were loud and aggressive and ignored them when they were quiet and gentle. At the end of the yearlong study the boys were unsurprisingly found to be louder and more aggressive than the girls and the girls were quieter and gentler compared to the boys (Fagot 1977). Blakemore et al. (2013) suggested that there are likely to be long term consequences of this reinforcement of perceptions of abilities on levels of self-efficacy according to a child's gender from parents and teachers. There has been a great deal less research completed on gendered political socialization process compared to gender socialization more broadly but it is expected that parents and teachers are more likely to positively reinforce male students when they put forward their ideas at home or in the class than for girls, reinforcing boys' political self-efficacy (Fraile and Sánchez-Vítores 2020; Bos et al. 2021).

Bussey and Bandura (1984) experimental research has shown that children learn through observing how people are treated according to their gender and then imitating typically the gender that they are identifying with. Their experiments showed that this identification went as far as imitating the level of power the person performing the gender role had (Bussey and Bandura 1984; Frey and Ruble 1992). Children are said to imitate and perform gender in the home or school and then also learn from the reaction of others who either reward and reinforce or punish the child for their performance (Bussey and Bandura 1999). These reactions can be added to with additional specific gendered instructions and tuition from a teacher or parents, who can generalize the performance to rules about what all girls or boys should do (Bussey and Bandura 1999). In addition, peers can also serve to maintain gender norms and can pressure their peers to conform to the expected gender role (Bussey and Bandura 1999). Knowledge from observing the action/performance, reaction and tuition is then stored (the cognitive part of this theory) to guide future behavior and becomes a tool for self-regulation as much as the regulation of others.

Gendered behavior in the classroom

From 1980s to 2000 considerable empirical research was undertaken observing gendered performances in the classroom and identifying inequalities (Nielsen and Davies 2016) which provides some evidence of the extent to which political efficacy may have been learnt or not according to gender in the classroom. Kelly (1988), in a meta-analysis of 81 quantitative studies of primary and secondary schools, showed that in all countries studied, across all ages, school levels, subjects, and socioeconomic and ethnic groupings, girls received fewer instructional contacts, fewer high-level questions and academic criticism, less behavioral criticism, and slightly less praise than boys. Closer to our interest, male students in American schools were said to be controlling the classroom conversation and asking many more questions (Sadker and Sadker 1994). In another study in the US, girls in the classroom were found to silence or self-censor themselves to maintain good relationships with other girls and keep in line with constructs of femininity (Brown and Gilligan 2013). Although girls were performing well at school it was noted that self-esteem of girls decreased during secondary school education (Lees 1986).

Considering previous evidence showing a gap in internal political efficacy at an early age and following social learning theory expecting a reinforcement of gender roles during childhood and adolescence, we expect the gender gap in internal political efficacy to already be visible, and to grow, during adolescence. The use of longitudinal data allows us to explore whether the gap widens among the same students over time. In the form of our first hypothesis, we expect the gap in political efficacy between boys and girls to be larger at age 16 than the gap at age 11 or 13 (H1).

The more recent literature on gender and classroom dynamics does identify a new more confident and active voice as being used by some of the more privileged girls in classroom interactions (Nielsen 2004). Nevertheless, there appears to be a significant gap in recent years in this field (Nielsen and Davies 2016). In contrast, there is, however, research on promising methods for teaching political self-efficacy that focus on classroom dynamics which has the potential to relate to both theories on learning political self-efficacy and the learning of gender roles. We will now review this citizenship education literature on the most cited teaching method for increasing political efficacy, "the open classroom climate" (Torney-Purta et al. 2001; Kuang, Kennedy and Mok 2018; Claes et al. 2017 and Knowles and McCafferty-Wright, 2015) and propose three potential ways in which this classroom dynamic can influence the internal political efficacy for girls and boys: differential access, differential effect and classroom composition.

Gender and differential access to an open classroom climate

The widely held consensus in the field of citizenship education research is that the method in schools for developing student's political self-efficacy is through a particular type of classroom discussion which is called "an open classroom climate" (Torney-Purta et al. 2001; Kuang, Kennedy and Mok 2018; Claes et al 2017 and Knowles and McCafferty-Wright 2015). This teaching method is where students are provided the opportunity to freely express their opinions and have their voice respected and listened to in classroom discussions. Students' self-reported access to the experience of this teaching method has over the last 20 years been repeatedly found to be a very effective method for developing students' political self-efficacy. For example, analysis of large-scale comparative studies of citizenship education such as the IEA International Citizenship Civic education study undertaken in 2009 and 2016 have demonstrated the association between this experience and political efficacy outcomes across diverse countries and at the two different time points (Hoskins et al. 2021; Kuang, Kennedy and Mok 2018; Claes et al 2017 and Knowles and McCafferty-Wright 2015 Manganelli et al. 2015;). A lack of student self-reported access to this learning experience has been found to be one of the major reasons for students from disadvantaged socioeconomic backgrounds having lower levels of political self-efficacy (Hoskins and Janmaat 2019, Hoskins et al 2017). It is important to note here that self-reported access to the experience of an open environment for discussion is not the same as capturing the amount of talking that an individual undertakes in a classroom discussion which would be the preferred measure to capture gender differences. It is, however, the closest measure available within the existing datasets to capturing this construct and, to the best of our knowledge, there has been no examination of how gender is associated with self-reported access to an open classroom climate, how gender interacts with an open classroom climate and how this influences the development of political self-efficacy.

To develop our next hypothesis on how the gap on political self-efficacy develops we combine the theory on gender classroom dynamics together with the empirical findings from the citizenship education field. Drawing on the above literature on gendered classroom dynamics that suggests that boys dominate classroom discussions (Sadker and Sadker 1994), we would therefore expect that boys report a higher level of access to an open classroom climate. The Citizenship Education literature we have just reviewed suggests that then the consequence of boys accessing and experiencing more open classroom is that boy's levels of political self-efficacy increase more than girls simply as they have greater access to this experience. Our second hypothesis, therefore, is that girls report less access to an open classroom climate than boys (H2).

Differential effect of open classroom climate

A second reason that can explain the different patterns of development of political efficacy between boys and girls is through a particular education method having differential learning outcomes for different social groups (Hoskins and Janmaat 2019; Campbell 2008). The usual explanations for differential learning outcomes from a particular education experience for a social group is that this group typically have less of this skill prior to the intervention and then they gain more from having experienced this opportunity. Alternatively, a learning activity is pitched at a high level and a certain level of skills is necessary to be able to learn from the experience and only those students with an already high level of the skill are then able to learn from the experience. Following the same rationale but building from the theories of gendered learning (Bussey and Bandura 1999), we propose another possible explanation here. Due to the theory that how students engage in classroom discussions is likely to be gendered (Kelly 1988; Sadker and Sadker 1994) and that boys are, first, expected to do more talking in class than girls do and, second, are more likely to receive positive feedback from their peers and teachers we then posit that boys are likely to benefit more from an open classroom climate than girls. Girls, inversely, are less likely than boys to speak in the class—regardless of how open they perceive the environment to be—and when they do speak, they might receive less positive feedback. Thus, hypothesis three is that boys are able to develop their political self-efficacy more than girls in the context of an open classroom climate (H3).

The effects of classroom composition

Finally, we explore in more detail the effects of student's peers. In studies that have explored gender behavior in classroom interactions the gender composition of the classroom has been identified as being important. A study from Ireland found that when the classroom composition was approximately equal or had more boys, the boys participated more in class discussion but when girls were the majority, they participated more than the boys in the class (Drudy and Chatháin 2002). In addition, experiments on gender and adult deliberation (Karpowitz and Mendelberg 2014; Mendelberg and Karpowitz 2016) identify the importance of the composition and proportion of women in the group in influencing the amount they will contribute to the debates. The analysis of the data benefits here from the fact that about 12% of English state funded secondary schools were single sex between the years 2000–2010 and that this is reflected in the schools within the CELS dataset (see details below). Our fourth hypothesis then is that girls will benefit more from an open classroom climate and develop higher levels of political self-efficacy when they are in the majority in the class or when the class only contains girls (H4).

Data and methods

We used CELS to test the hypotheses. The CELS dataset is unique in that it combines a panel study with repeated measures of the same adolescents over time with a questionnaire tapping a wide range of civic attitudes. The study includes data from a cohort of young people aged 11 and 12 (Year 7; first year of secondary school) when they were surveyed for the first time in 2003. The same students were then surveyed every 2 years until 2014 (Round 6). The data was collected from a nationally representative sample of 112 state-maintained schools in England—representative in terms of region, General Secondary Education Certificate attainment and percentage of students on free school meals (Keating and Benton, 2013). Within each school all the students of target grade were selected. We used the data from Rounds 1, 2, and 3, when pupils were aged 11/12, 13/14, and 15/16, respectively. At the oldest age group, the young people are students in Year 11, which is the final year of lower secondary education and the last year of citizenship education as a statutory component of the curriculum.

The CELS dataset like other longitudinal datasets has considerable attrition. While the study started out with 18,583 respondents in Round 1, only 6155 of this original sample (33%) took part in Round 3. These 6155 respondents constitute our analytical sample. Following the method adopted by Eckstein et al. (2012), we compared respondents who only participated in the first round with those who took part in both rounds in order to test whether this attrition was selective. Significant attrition effects were found for the Round 1 variables "Books at home" and "Parental educational attainment", with respondents participating in both rounds reporting significantly higher numbers of books and higher levels of parental education.

To prevent this selective attrition from skewing our results main models presented below include a key measure of parental socioeconomic status, the number of books at home, in the model. In addition, sensitivity tests run with parental education (see Table A2 in the Supplementary Information File) show no significant differences with the main results. These two variables, parental education and the number of books at home, are used as control variables in the prediction of the dependent variable, students' internal political efficacy (see next section for a thorough description of the measures used in the present study). It is unlikely there will be any remaining attrition bias if the variables associated with the attrition are included in the analysis as the key independent or control variables (Paterson 2013).

As students are nested in schools, the independence of observations assumption is violated, which indicates that multilevel analysis is necessary. Two levels were distinguished in the data: the individual level (n = 5833) and the school level (n = 67). To take advantage of the panel structure of our data and to identify changes over time, our main models include levels of internal political efficacy in wave 3 as dependent variable and the prior levels of political efficacy in the previous wave 2 as a control. With this strategy we can focus on the intra individual development of internal political efficacy.^{2 3}

Measures

Political self-efficacy is measured with levels of agreement to the standard statement "Politics is so complicated that I can't understand". This scale is often used in political behavior research as a single indicator of internal political efficacy when other indicators are not available (Craig and Maggiotto 1982; Niemi et al 1999; Morrel, 2003; Mendelberg and Karpowitz 2016) and has been used regularly in electoral surveys since the 50s. In addition, it has been proposed as a particularly useful item to tap into the "ability" or "skills" component of internal efficacy (see Gidengil et al 2008). Furthermore, we conducted sensitivity tests with an alternative indicator, agreeing with the statement "I know less about politics than most people my age" and found no significant differences in the results (see figures A1 and A2 in the Supplementary Information File). Therefore, the standard measure was preferred. Respondents were provided a five-point Likert scale to answer (strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree). The scale was reversed before running the analyses so that higher levels imply higher levels of internal political efficacy. The school-based source of learning that we examine in this article is open classroom climate. This is a scale consisting of six items measured from 1 "not at all" to 5 "a lot" and taps whether students: (1) felt they could bring up issues for discussion, (2) felt encouraged to make up their own minds, (3) felt free to express opinions, (4) felt free to disagree with the teacher, (5) were under the impression that their teacher provided with several sides of an issue, and (6) felt teachers respected students' opinions (See the online supplementary material for table A1 for descriptive statistics). The scale showed a sound level of internal coherence (Cronbach's alpha = 0.811 in 2007) and represents the saved output of a factor analysis (i.e. the factor scores). These variables were constructed in the same way for each of the three rounds of data.

To measure gender, we use the item that asked the young people to identify if they were a girl or boy⁴. As highlighted above, we consider social background to be a key influence on political engagement both in a direct and indirect sense. Indirectly, social background shapes political engagement by determining access to learning sources. We measure social background with an indicator of the number of books. This measure focuses on Bourdieu's concept of cultural capital that may be understood as an individual's gain from parental level of education and cultural artefacts such as books (Bourdieu and Passeron 1990). The number of books at home has also been found to be strongly correlated to parental income (Schütz, Ursprung and Wößmann 2008; Baird 2012), making it a reasonable replacement for parental income and occupation considering that more obvious measures of social background are not available within the CELS dataset (Keating and Benton 2013). Sensitivity tests also use parental educational attainment (see table A2 in the Supplementary Information File); this variable is not included in the main models due to its high number of missing cases. In addition, our models include controls for ethnicity with a variable that differentiates between students that self-describe themselves as White British from other categories. Finally, to examine if the number of girls and boys in class affects the development of political efficacy, a variable was created that distinguishes whether the classroom has "only boys or more boys than girls (between 0 and 45% girls)", approximately "same number of boys and girls" (between 45 and 55% of girls), and "more girls than boys (between 55 and 99 % girls) or only girls". Descriptive statistics for all variables are available in the Supplementary Information File (See the online supplementary material for table A1).

Figure 1.

Levels of internal political efficacy in three waves for boys and girls. Note: Estimates are means with 95% confidence intervals.

Results

To address our first hypothesis, we examine levels of internal political self-efficacy for boys and girls in the three waves. As descriptive statistics in figure 1 show, boys' average levels of political self-efficacy increase significantly in each subsequent wave. However, levels of political self-efficacy of girls start from a slightly lower level and do not increase over time but instead remain rather stable. This preliminary descriptive result suggests that, during the key formative years (from 11 to 16 years old), boys develop confidence in their abilities to understand and participate in the political world while girls do not. When using the long format of the data, the average increase of internal political efficacy for boys and girls is significantly different (P ≤ .0000), 0,18 points per year for boys, and 0.009 for girls. Overall, the data provides support for our first hypotheses that the gender gap in internal political efficacy widens significantly during the key formative years of adolescence.

Gender and access to an open classroom climate

One possible explanation of why boys increase their levels of internal political efficacy significantly more than girls is our second hypothesis which expects that girls report less access to an open classroom climate. After all, previous research has shown that an open classroom climate promotes internal political efficacy and also that boys tend to talk more in class than girls. However, a comparison of mean levels of perceived open classroom discussion in both wave 2 and wave 3 in the survey show that this is not the case. Girls report significantly higher levels of open classroom discussion in class in both waves (P ≤ .000, see table 1). Furthermore, they also report higher levels in each of the items that compose the open classroom climate index (see fig. 2 including data for wave 3 as an example). Thus, the evidence does not support our second hypothesis, since girls perceive the same, or even higher, access to an open classroom climate than boys.⁵ These unexpected results could be due to the measure which is capturing self-reported access to an open classroom climate and not actual levels of participation in classroom discussions.

Differential effect of open classroom climate

Our third hypothesis suggests that boys gain more from the open classroom climate experience than girls do due to gendered dynamics in class. To test this proposition, we use a linear hierarchical model with political efficacy in wave 3 as the outcome. To examine if the open classroom climate has the same effect on girls and boys' levels of internal political efficacy, we

Table 1.

Perception/access to open classroom climate for boys and girls.

Figure 2.

Perceived open classroom discussion by boys and girls in wave 3. Note: Estimates are means with 95% confidence intervals.

include a multiplicative term between the individual perceptions of an open classroom climate and gender. As introduced before, political self-efficacy in the previous wave, the number of books at home and respondents' ethnicity are included in the equation as control variables (Model 3). In addition to the multilevel equation model with the interaction term, table 2 presents two previous models as reference points. Model 1 is a baseline model that shows that most of the variation in levels of political efficacy takes place at the individual level and not at the school level (Variance Partition Coefficient = 0.009). Model 2 incorporates the open classroom climate and the control variables. It shows that quite surprisingly the open classroom climate, once controlled for by other factors, has no direct effect on the development of political efficacy from wave 2 to wave 3. Political efficacy in the previous wave and the number of books at home has the expected positive relationship to levels of internal political efficacy in wave 3. Furthermore, the results presented in table 2 for model 3 show that, as expected (H3), there is a negative correlation between being a girl and an open classroom climate (see Model 3).

To further illustrate the results, figure 3 presents the marginal effects and shows how the higher the perceived open character of the class the higher the gap in levels of political efficacy between girls and boys. This is the result of the positive development of internal political efficacy as the classroom discussion becomes more open for boys. While girls, if anything, develop lower levels of internal political efficacy in the same setting. That is to say, as the class incorporates more topics and discussions in which the students feel free to participate and give their opinions, girls' political self-efficacy is reducing. As a result, while the difference in overall levels of internal political efficacy between girls and boys in a classroom with the minimum level of open discussion is of 0.23 points in favor of boys in a scale from 0 to 4, the gap is .62 in the scenario in which the discussion is perceived as most open. That is, the gap becomes 2.7 times the original difference.

Table 2.

Political efficacy and classroom climate. Multilevel linear equations.

Classroom composition

The results so far suggest that girls do not seem to develop internal political efficacy when exposed to an open classroom climate. Our proposal, based on gendered learning outcomes, is that boys might benefit most from this type of learning since they are more likely to talk in class in the first place and to receive positive feedback when they decide to talk than girls. If this is

Figure 3.

Predictive margins of political efficacy by classroom climate for boys and girls. Note: Estimates come from model M3 in table 2.

the case, we should observe that levels of political self-efficacy get worse for girls when in the presence of such an open climate and there are boys in the classroom, and that internal political efficacy stays the same when there are no boys present. To test our fourth hypothesis that girls learn political self-efficacy more when there are less boys in the class we have incorporated a measure of the number of boys in class to our previous models (see model 4 in table 2). Specifically, we have incorporated a triple interaction term to examine what is the effect on an open classroom climate for boys and for girls according to the number of boys in the classroom.

Figure 4 provides the margins of political efficacy by classroom climate for boys and girls in three scenarios. In schools with only boys, or with more boys than girls, an open classroom climate does not seem to foster levels of political efficacy for boys or girls as illustrated by the flat trajectories in the first graph in figure 4. The most common scenario is that of a similar number of boys and girls in class, which shows the differentiated effect of an open classroom discussion illustrated above for boys—a positive development—in comparison to girls—a negative effect. Finally, the scenario in which the gender gap in internal political efficacy becomes particularly acute is that in which there are more girls than boys in class. In this third case, an open classroom climate has a positive and significant relationship with the development of internal political efficacy for boys and still no effect for girls. These results suggest that the presence of some male students might be changing already the dynamics in class, and clearly show an acceleration in the development of self-efficacy in male students when it comes to understanding the political world. These striking results are further discussed below considering the social learning literature.

The results also suggest that the teaching method of an "open classroom climate", which has been found to be able to compensate for gaps in other political attitudes such as political interest (García-Albacete 2013; Neundorf et al. 2016), appears to have an unequal effect that increases the persistent gap in women and men's efficacy in the political world. These results open the door to a new research puzzle which is urgent to address.

Discussion

The interpretation of these striking results suggests that boys develop political efficacy in an open classroom climate while girls do not, even if boys are a minority in class, can only be speculative at this point. In order to provide possible explanations, we return to our review of the literature and in particular social learning theory. Building from the experimental research

Figure 4.

Predictive margins of political efficacy by classroom climate and number of boys and girls in the classroom for boys and girls. Note: Estimates come from model 4 in Table 2.

and gender socialization theories of Bussey and Bandura (1999) we suggested that political self-efficacy is likely to be developed through the dynamics in classroom discussion in two ways. First, when a student participates in the classroom discussion and receives positive feedback from their peers and teachers or second, through the process of observing a student of the same gender participating in the class discussions and receiving positive feedback. Thus, our results suggest that boys are likely to be participating more in the classroom and likely to be receiving more positive feedback from teachers and peers. In contrast, girls are likely to be participating less in the classroom discussion and receiving less positive feedback from their peers and teachers. This appears to be the case even though the perception of the openness of the classroom itself does not substantially differ. It is also the case that boys' political self-efficacy appears to benefit from the positive feedback from girls in the class as when boys are alone their political self-efficacy does not appear to rise. Theories of the gender socialization process in previous research suggested that girls appeared not to support their own gender's development of political self-efficacy but form part of the self-regulation process of gender roles (Brown and Gilligan 2013). We can see this as when girls are by themselves their levels of political self-efficacy also do not rise. However, the effect of even a small number of boys in the class is becoming detrimental to girls' development of political self-efficacy. Thus, the effect of boys' regulation of femininity within classroom interaction appears here to be stronger. These results suggest that the gendered norms of expected classroom behavior in discussion have not appeared to have substantially changed from the 1990s when most of the gender research in classroom practice to 2003–2007 when this research was conducted and that classroom dynamics appears to be negatively affecting girls learning of political self-efficacy. This interpretation is speculative, unfortunately, as longitudinal data is not available to test these questions on young people in the last decade.

Conclusion

This article has addressed the development of the gender gap in political self-efficacy which has been identified as one of the key contributors to gender inequalities in political leadership—one of the pervasive barriers to democratic equality. The first contribution of this article is identifying that the gender gap in political self-efficacy develops between the ages of 11–16 and that, although is already significant at age 11, widens substantially through these early adolescent years. This result adds to recent research that combines theories of gender and political socialization together to understand the development of the gender gap in political leadership (Bos et al. 2021) and provides the first test of these theories, to our knowledge, using repeated measures for the same students. As a potential solution, we have focused on the school and on one teaching method that has previously shown to be successful for learning political self-efficacy along with many other skills for political engagement, the open classroom climate. The currently available measures used to capture this construct only measure self-reported access to this experience and not levels of actual participation in discussions and we acknowledge that the existing measures are limited for research into understanding gender classroom dynamics.

Nevertheless, and contrary to our early expectations, the findings suggest that self-reported access to an open classroom climate is currently only benefiting boys and appears at least to be partly responsible for increasing the gender gap in level of political self-efficacy. Furthermore, this is the case even when girls are the majority in the classroom. We have no reason to expect England to have specific patterns in terms of gendered political socialization compared to other Western democracies; nevertheless, it would be interesting to test similar evidence in different contexts.

These findings are a novel contribution to research on gendered political socialization and political behavior and thus open a number of questions. The first one relates to potential explanations. The literature suggests that gender and political socialization processes in the school classroom environment may well be at work in forming the gender gap on political self-efficacy. Building from Bussey and Bandura (1999) social cognitive theory of the learning of gendered behavior, we suggest that boys are likely to be benefiting from positive feedback from girls, boys and teachers when they participate in the classroom discussion and in contrast girls are unlikely to be having this same experience.

The literature on deliberative pedagogies, that formalize the concept of an open classroom climate in their education programs, suggest that one of the basic principles behind the success of discussions is that all sides are on an equal footing, are equally respected and not dismissed or ignored because of who is speaking (Gibson 2020). We can hypothesize from our research findings that this equal respect is not being fulfilled in classroom discussions and that girls are losing out. Observational research on gender in the classrooms is needed to verify if this indeed the case and to understand if these dynamics are still relevant today. These conclusions are similar to Morehouse Mendez and Osborn (2010) who examined adult discussions and found that both men and women perceived women as being less competent on politics and the authors identified that the gender dynamics within these deliberations as being the likely source of these gendered perceptions. Following our results, it is necessary for educators to re-examine classroom discussions to reflect on the extent that they are actually involving all social groups and to ensure all voices are heard and equally respected.

In addition, there are several plausible education interventions that could be tested to see if they reduce inequalities in the development of political self-efficacy. First, developing teacher training, so that teachers are able to ensure that the whole class and teachers themselves give genuine equality of respect to girls who speak in the classroom and provide more encouragement for them to get involved. Second, to support girls directly and work with them to encourage them to raise their voice and challenge concepts of femininity in classroom interaction through the use of female role models and mentors.

Supplementary Material

Click here for supplementary material.