Introduction

Participation rates in adult education tend to be unequal. Educational attainment, having a job or not, and age have been repeatedly found to be major determinants of participation (Boeren, 2016; Desjardins, 2017). Those with the highest qualifications, with knowledge-intensive jobs, and those who are younger are more likely to take part. While girls are known broadly to have caught up with boys in initial education systems in recent decades, as adults, women in many countries tend to receive less support for participation in work-related training (EIGE, 2019).

Recent research by Boeren (2019) has also demonstrated that adult migrants living in Europe tend to participate less in education. While separate variables such as gender, ethnicity and social class are known to correlate with participation in adult education, this chapter aims to extend knowledge on participation issues using the lens of intersectionality. Current knowledge on adult education and intersectionality is based mainly on small-scale studies. Instead, in this chapter, we draw on large-scale representative data from the European Social Survey (ESS) and Adult Education Survey (AES) to further understand the role of gender and intersectional dimensions in adult education. This also allows us to move away from studying gender and intersectionality in one specific setting and to take into account the structural differences in the economy, labour markets and education systems that exist between the diverse countries of Europe. As previous studies show, adult learning systems are indeed ‘embedded in specific economic and social arrangements’ (Ioannidou & Jenner, 2021, p. 321) and ‘lie at the intersection of a variety of other systems including a nation’s education and training system, labour market and employment system and other welfare state and social policy measures’ (Desjardins, 2017, p. 21).

In this chapter, we first focus on theoretical explanations of gender inequalities, intersectionality and participation in adult education. We highlight variations between European countries and link these to the theory of Bounded Agency (Rubenson & Desjardins, 2009), as used throughout this book. We discuss European policies on gender equality and lifelong learning before turning to the methodological underpinnings of our analyses. The analysis of the impact of intersectionality on participation in adult education, including the simultaneous negative effects of gender, social origin (parents with basic or lower education) and ethnicity, enables us to contribute further to the research on vulnerable groups undertaken within Enliven (Maiztegui-Oñate et al., 2019). Discussion of our results leads to a conclusion with recommendations for policy, practice and research.

Gender Segregation: Theoretical Explanations

Lindsey (2005) argued that segregated gender roles peak once children enter a family. Men feel strongly responsible for being the breadwinner while women move into being the primary carer, looking after the children and undertaking household duties. Because of their stronger involvement in the labour market, men are more likely to participate in vocational non-formal education and training which nowadays is very dependent on job-related characteristics. This mechanism can also be explained through human capital theory (Becker, 1985; Livingston & Guile, 2012; Aleandri & Refrigeri, 2013; Knipprath & De Rick, 2015). As fathers feel very responsible for maintaining their family financially, they are more likely to invest in their skills through education and training (Dieckhoff & Steiber, 2011). This will then further strengthen their position in the labour market.

Yet focusing on the biological differences between men and women as part of the reproductive process, and how this carries over in their diverging roles once a new-born child has arrived, is not enough. The role of discrimination against women has been well-documented in the literature. Neilson and Ying (2016) wrote about ‘taste discrimination’. This concept refers to people’s preferences to work with others who are most similar to themselves. Because of the higher proportion of men among senior managers, there is also a higher likelihood of appointing men to these roles. This leads to what Schuller (2017) has labelled the Paula Principle, a situation in which many women work below their levels of competence because of sustained gender stereotypical thinking among managers. In relation to education and training, those in managerial positions receive more chances to participate, and this is strongly age-dependent (Tennant, 2007; Boeren, 2016). As women tend to reproduce early on in their careers, or at least in the first half of their working lives, the advantage they have when younger is also disrupted.

Leathwood (2006) describes participation patterns in lifelong learning as strongly classed, raced and gendered due to a combination of reinforcing factors known in the academic literature as ‘intersectionality’.Footnote 1 For example, the costs for lifelong learning—underpinned by an economic rationale—are higher and rewards are lower for low-educated women from ethnic minority backgrounds, especially when they keep on facing discrimination afterwards and remain in low-paid jobs. The social justice rationale is expressed in the concept of inclusion, which goes beyond being employed or mastering basic skills. Inclusion means widening access to social, cultural and material capitals that can lead to increased levels of self-confidence (Bartky, 1990).

In her article ‘The Complexity of Intersectionality‘, McCall (2005) insists on the multi-layered complexities of intersectional analysis. She introduces three basic levels of complexity: (1) ‘intra-categorical’—‘focusing on particular social groups at neglected points of intersection’; (2) ‘anti-categorical’—deconstructing the very categories of analysis as they ‘do not readily allow for the diversity and heterogeneity of experience to be represented’; and (3) ‘inter-categorical’ (or categorical)—seeking to ‘document relationships of inequality among social groups and changing configurations of inequality among multiple and conflicting dimensions’ (McCall, 2005, p. 1773). In our analysis, we implement the first of these levels of the analysis of intersectionality, applying the intra-categorial approach to intersectionality by identifying internal inequalities among women based on social background (measured by parental level of education), and their own education and ethnicity.

The role of women in society is perceived differently across Europe. Esping-Andersen (1990) extensively researched differences between European welfare states, for example, focusing on the social-democratic nature of the Nordic countries and their lower levels of discrimination (Veggeland, 2016). These contrast with policies in the Southern Mediterranean countries where women are expected to concentrate on family duties. This translates into lower demand among women to participate in education and training. As Schroder (2016) pointed out, these differences are strongly embedded in countries’ cultures and are therefore difficult to change, and mean that women’s agency to participate in education and training is bounded by the wider structural contexts in which they live. The European post-socialist countries represent a separate welfare regime, characterised by a longer tradition of combining full-time employment and family, of high public sector employment, and of a well-established infrastructure of childcare and extended paid parental leave. Given the importance of differences between countries, different welfare regimes will be included in our statistical models presented below: social-democratic, conservative, Mediterranean, liberal, and post-socialist regimes.

European Policies on Lifelong Learning and Gender Equality

Europe’s core policies on education and training in the last 20 years have centred on the use of benchmarks and indicators (Boeren, 2016). By 2020, member states were expected to have 15% of their population between the ages of 25 and 64 participating in at least one formal and/or non-formal learning activity. This was measured on a four weeks basis using data from the Labour Force Survey.

Gender equality is an important focus point of the European Pillar of Social Rights, introduced in 2017 (European Commission, 2017a). Twenty pillars are distributed among three strands: (1) equal opportunities and access to the labour market, (2) fair work conditions, and (3) social protection and inclusion. A New Start Initiative further underlined the need to restore imbalances in pay and careers, which often favour men (European Commission, 2017b). Interestingly, the Initiative focused on paternity leave, parental leave, care leave, and flexible working arrangements but does not mention lifelong learning, education, or training. The cultural norms that lead men to see themselves as the major breadwinners may be difficult to break and take considerable time to change.

Lombardo and Agustin (2011) evaluated European policies and their effects on vulnerable individuals and groups through an intersectional lens. Their work suggests that reducing inequalities through policies needs to be related to class, gender, and ethnicity. Inequalities are not only an accumulation of different individual dimensions such as gender, ethnicity, and class but also need to be understood in the light of broader societal levels of deprivation. Lombardo and Agustin (2011) focus on three policy topics: (1) gender-based violence; (2) intimate citizenship—policies regulating the intimate or private sphere—divorces, marriages, reproductive rights; and (3) non employment policies such as taxes, work-life balance, gender pay gap, and equal treatment. They suggest that these policies need to be dealt with in an explicit and visible way, through the articulation of intersectionality and inclusiveness. Policy suggestions need to take a transformative approach that takes structural power hierarchies into account and challenges the main privileges in society. In doing this, they argue, it is important to avoid stigmatising particular groups and to consult civil society organizations.

The good intersectionality approach should be evaluated against public investments in education, underpinned by the concept of equality of opportunities (Cefalo & Kazepov, 2018). Social investments through public expenditure aimed at achieving inclusion and equity are also worth evaluating from the intersectional perspective. For the European Union (EU), social investment ‘means policies designed to strengthen people’s skills and capacities and support them to participate fully in employment and social life. Key policy areas include education, quality childcare, healthcare, training, job search assistance and rehabilitation’.Footnote 2

Taking into account differences in policy approaches between countries, this chapter applies the theoretical concept of bounded agency (Rubenson & Desjardins, 2009) from a gender-sensitive perspective. We work with the distinction made by Cross (1981) that defines situational, institutional, and dispositional barriers to participation in adult learning. Roosmaa and Saar (2016) observed, in quantitative comparative research, that women tend to experience more situational and institutional barriers to participation in adult learning than men, but fewer dispositional barriers. According to Saar et al. (2014), perceived barriers can be removed by institutional and structural solutions: this is a core underpinning of the bounded agency concept, which argues that adults are ‘bounded’ by structural elements, effecting the opportunities of their individual actions. Boeren (2011) investigated the gender gap related to barriers preventing participation. She found several disadvantages for women, such as financial difficulties in paying course fees and the experience of time pressure due to difficulties in combining family responsibilities with other life tasks. In this chapter, we analyse mainly quantitative data on the differences within the category of women, based on the view that intersectional research should reject a singular category of gender and avoid applying generalizations valid for some women to all women (Verloo et al., 2009).

Although the analyses in this chapter draw theoretically primarily on perspectives of intersectionality and bounded agency, discussions of vulnerability also throw light on societal inequalities. Maiztegui-Oñate et al. (2019) discuss the individual level characteristics that lead to vulnerability; these are included in our models as control factors: sex, low educational attainment, ethnic minority status, and unemployment. As structural factors, we also add family background, specifically low-educated parents as a proxy for low income and lower socio-economic background, leading to poverty, which tends to be intergenerational. In this chapter, we aim to explain in more depth the negative effects of intersectionality on participation in adult education.

Gökşen et al. (2016) mapped vulnerability by gender across class and ethnic differences. They also used regime type typology for studying vulnerability in school-to-work transitions—universalistic (Denmark and the Netherlands), liberal (United Kingdom), employment-centred (France and Belgium), and sub-protective (Spain, Greece and Turkey). Our research takes this further by focusing on intersectionality in adult education as a critical factor—among women from ethnic minority backgrounds and from families with a low educational background—in increasing the risk of poverty and becoming vulnerable. By including family background, we can account for the complex family situations of ethnic minority women, including culture, traditions, and material poverty. We also include the post-socialist, central, and eastern European countries, with the aim of focusing on the institutional level, the policy measures taken, and their effects on women’s participation in adult education. The authors in the STYLE projectFootnote 3 discovered strong intersectionality effects between gender and youth and linked them to migrant status. Central and East European countries relate in different ways to ethnic minorities, especially to the Roma population. There are estimated to be around 10–12 million Roma, half living in the EU (Martinidis et al., 2015); they form very vulnerable groups in several European post-socialist countries where they are part of the native population and cannot be treated as migrants. Kóczé (2009) argued that there was no significant intersectional understanding of Roma women’s social position in policy-making.

For the EU, there are on average no gender differences in participation in adult education (Fig. 6.1). Bigger gender differences can be observed in Cyprus, the Czech Republic, Estonia, and Sweden. However, these data do not give us a clear picture and the data do not answer the question of whether there are significant gender differences in adult education when we consider other important factors such as education, class (own and parents’ educational background), or ethnicity.

Fig. 6.1
Bar graph of males and females of different countries. The graph for females is showing an increase.

Gender differences in participation in adult education (Source: Eurostat, AES 2016, data code: [trng_aes_100])

Full size image

Our main research questions are: Which individual factors cause differences (1) between men and women, and (2) between women with different socio-economic characteristics and social background? and (3) How efficient is the role of institutions and policies in different European countries in overcoming the barriers caused by intersectionality for access to lifelong learning?

Data and Methodology

In this chapter, we use data from the European Social Survey 2016 and the Adult Education Survey 2016. The European Social Survey is a biannual cross-national survey, representative for the population aged 15 and over. We have selected data from the eighth round, carried out in 2016 (European Social Survey Round 8 Data, 2016). This is because we wanted to analyse data from two different datasets with the same reference year (2016). We have limited the analysis to 21 countries.

The Adult Education SurveyFootnote 4 2016 provides information on adults’ participation in education and training and is one of the main data sources for EU lifelong learning statistics. The reference period for participation in education and training is the 12 months immediately preceding the Adult Education Survey interview. It also covers themes such as access to information on, and obstacles to participation in, education and training. The Adult Education Survey covers the resident population aged 25–64. The 2016 Adult Education Survey scientific use file which we received from Eurostat includes micro data from 32 countries. We have limited the analysis to 29 of them: 27 EU countries,Footnote 5 Norway, and Switzerland.

We use four dependent variables. The first is a yes/no variable indicating whether a person had improved their knowledge or skills by attending a course, lecture, or conference in the last 12 months or not. The second is a yes/no variable indicating whether an adult had mentioned family responsibilities as an obstacle to participation in education and training. The third is a yes/no variable indicating whether an adult had received ‘partial or full payment’ for the first non-formal education (NFE) activity they had attended from ‘public employment services’ or ‘other public institution’. The fourth is a yes/no variable indicating whether an adult had received partial or full payment for the first NFE activity they attended from an ‘employer or prospective employer’.

As independent variables, we have included variables at individual and country level. At the individual level, we have included in the analysis of the first dependent variable gender as a main independent variable (female (1) or male (0)). We have controlled for respondents’ level of education (low = ISCED 0–2, medium = ISCED 3–4, and high = ISCED 5–8); social background measured against parents’ educational background (0—of low educational background: people neither of whose parents have upper secondary, post-secondary or higher education) or (1—of a high educational background: those with at least one parent who has upper secondary or higher education); whether the respondents have children living at home (1) or not (0); whether they were living with a husband, wife, or partner (1) or not (0); whether respondents belong to a minority ethnic group (1) or not (0); their main activity at the time of the survey (0—paid work, 2—education, 3—unemployment, 4—inactivity); and their age (as a continuous variable).

Regarding the analysis of the other three dependent variables, we included the same independent variables at individual level except for belonging to a minority ethnic group (the Adult Education Survey does not collect data on this). However, some of the categories differ slightly. Thus, in the case of the current labour market status the Adult Education Survey allows for four categories: 0—full-time employed, 1—part-time employed, 2—unemployed, 3—inactive. In the case of cohabitation, there is a variable which distinguishes whether a person is living in a consensual union (1) or not (0).

At country level, we have included a variable which distinguishes the welfare regime to which a country belongs. For our analyses, the countries in the European Social Survey are classified in the following way. Denmark, Finland, Norway, and Sweden are coded as social-democratic regimes. The liberal regime type encompasses Ireland, Iceland, and the United Kingdom. Italy, Portugal, and Spain are classified as Mediterranean regimes. Austria, Belgium, France, Germany, and Switzerland constitute the conservative type. We classify the following countries as post-socialist: the Czech Republic, Estonia, Hungary, Lithuania, Poland, and Slovenia. For the Adult Education Survey, we have also classified Norway as a social-democratic regime. In the Adult Education Survey, there is only one country belonging to the liberal regime (the United Kingdom); Cyprus, Greece, and Malta are classified as Mediterranean regimes; Luxembourg as conservative; and Bulgaria, Croatia, Latvia, Romania, and Slovakia as post-socialist.

We implement a multi-group analysis, investigating the existence of a gender gap related to the barriers preventing men and women from equal participation in adult education and the role of different factors in gender deprivation. We estimate the effects of education, social origin measured by parents’ education, ethnicity and family situation, labour market participation, and age on access to adult education, both for the entire sample and for men and women separately. This approach makes it possible to test whether estimates are equivalent across gender groups.

We employ multilevel regression models (see Rabe-Hesketh & Skrondal, 2012). Multilevel models are useful, especially in handling nested data as individuals are ‘nested’ into countries. The assumption is that the attitudes or behaviour of people may be influenced not only at the individual level (level 1)—characteristics such as educational level, sex, age, social background—but also by group characteristics (level 2) such as the welfare regime of the country where people live, labour market policies, education, family, and social assistance. In order to analyse the factors which prevent men and women from participating equally in adult education and to consider family responsibility as an obstacle to participation, we estimate three models for each dependent variable: a model which includes all respondents and two separate multilevel models for women and men.

Factors for Gender Differences in Participation in Adult Education

We will now discuss the factors that prevent women from equal adult education participation with men. Table 6.1 demonstrates that there are significant gender differences in participation in lifelong learning. Being a woman is associated with a higher probability of participation than being a man, when control variables are kept constant. However, when looking into the control variables, we can observe additional negative effects strongly related to gender. Thus, the results show that inactivity in the labour market decreases the probability of participation in adult learning. Women taking care of newborn children are over-represented in this category. Chances of getting back into employment are unevenly distributed among women and depend on opportunities for participation in additional training. Living with a partner is also associated with a lower probability of participation in adult learning. This underlines the importance of studying adults’ family situation separately for men and women.

Table 6.1 Participation in adult education (Results of two-level random intercept logistic regression models on whether a person has improved his/her knowledge/skills: course/lecture/conference, in the last 12 months)
Full size table

Having a parent with a high or medium level of education increases the probability of participating in adult learning compared to a person with low-educated parent. Belonging to an ethnic minority group is associated with a lower probability of participating in adult learning compared to other adults. That raises the importance of investigating intersectionality further and evaluating the impact of European policies on vulnerable individuals and groups affected by multiple dimensions of inequality. The estimates also provide evidence that a country’s welfare regime matters for participation in adult learning. More specifically, they show that adults who live in the countries that belong to the conservative, Mediterranean, and, especially, post-socialist regimes are less likely to participate in adult learning than those in social-democratic countries.

The differentiated models estimated for women and men show that the effect of higher and secondary education on adult learning participation is greater for women than men. Our analyses also show some gender differences relating to family situations. Having a child is associated with a higher probability of adult learning for men but has no association for women. Living with a partner has a negative effect on the probability of participating in adult learning for women, but no effect for men. We see similarities in the effects, which are important for the phenomenon of intersectionality. Belonging to an ethnic minority group has a stronger negative effect for women than for men. The statistical effect of parents’ educational background is stronger for men than for women. Last but not least, we observe that men who live in countries with liberal and conservative regimes have a lower probability of participating in adult learning when compared to men living in social-democratic countries. For women, there are no such differences. Finally, welfare regimes typology can explain the slightly higher proportion of the country-level variance in the case of men (but not women).

Obstacles Preventing Participation in Adult Education from a Gender Perspective

Figure 6.2 demonstrates that the strongest barrier preventing participation in adult education in the EU 28 relates to not seeing any need for (further) education and training (75.4% of women and 79% of men). ‘Family responsibilities’, a situational barrier, are mentioned by 39.8% of women but only 24.2% of men, indicating a large gender difference. Not being able to meet the ‘costs’ was mentioned by 30.5% of women and 26% of men.

Fig. 6.2
The bar graph is plotted for males and females in different services. The highest value for males and females is no need for further education or training at 79 and 75.4 respectively.

Obstacles to participation in education and training for adults in the EU 28 countries, aged 25–64, by sex. (Source: Eurostat, AES 2016, data code: [trng_aes_176])

Full size image

Information in Table 6.2 demonstrates that gender is an important factor in adults’ chances of improving their education and qualifications. More specifically, the probability of women perceiving family responsibilities as an obstacle to undertaking learning is higher than for men, if control variables are kept constant. Two variables at individual level influence men and women differently: part-time vs full-time employment and higher education vs lower education. Part-time employment has an insignificant effect for men but is significant for women, increasing the negative effect of the ‘family situation’ as an obstacle for their participation in adult learning. Higher education decreases the importance of the family situation as an obstacle for women and increases its importance as an obstacle for men. Higher age decreases the importance of family responsibilities as an obstacle. We also observe some gender differences at the country level. While women in countries other than the social-democratic welfare states see family responsibilities as an obstacle, for men this is true only in the Mediterranean regime. Welfare regime explains a higher proportion of country-level variance in the case of women than it does for men (respectively 42.77% vs 33.48%).

Table 6.2 Family responsibility as an obstacle for participation in adult education (Results for two-level random intercept logistic regression models on whether an adult has mentioned family responsibilities as an obstacle to participation in education and training)
Full size table

Gender Gaps Related to Cost

Gender differences also occur in relation to the institutional barrier represented by the ‘costs’ (who pays for additional qualification or training). We found significant effects of gender, education, and social origin measured by the educational level of parents (see Table 6.3). Adults with parents educated at upper secondary or higher education level have a lower probability of receiving additional training paid by a public institution. The probability of receiving financial support from a public institution is higher for women than for men. Compared to countries from a social-democratic regime, only respondents living in Mediterranean countries are more likely to receive payment from public institutions for their training.

Table 6.3 Results for two-level random intercept logistic regression models on whether an adult had received ‘Partial or full payment for the 1st NFE activity: public employment services’ or ‘other public institution’
Full size table

In general, women with medium levels of education are less likely to receive payment from public institutions for their training than women with low education. This effect is not present for men. For women, increased age has a positive effect on the probability of receiving financial support from a public institution; this effect is not present in the case of men.

Women are less likely to receive payment from their employer for training compared to men (see Table 6.4). People with secondary and higher education also tend to receive more payments from their employers than those with low education. Differences between countries (grouped into different welfare regimes) in the probability of receiving partial- or full-payment from an employer or prospective employer are not significant. The only exception is for those living in Mediterranean countries, who are less likely to receive training paid by employers compared to those living in social-democratic countries.

Table 6.4 Results for two-level random intercept logistic regression models on whether an adult had received ‘Partial or full payment for the 1st NFE activity: employer or prospective employer’
Full size table

Discussion and Conclusions

This chapter has investigated the individual factors and barriers (situational and institutional) preventing men and women from participating equally in adult education. The results allow us to respond to our three research questions.

Based on our analyses, ‘family responsibilities’ presented the largest gender difference. It represents a situational barrier, which has been discussed in previous research. The issue of family responsibilities is also strengthened for women through part-time employment or inactivity; this is not the case for men. The assignment of primary care roles to mothers (O’Reilly, 2010; Siddle, 2011) explains the first element of the gender differences observed in participation. It can, however, also be explained through human capital theory (Becker, 1985). The idea of family responsibilities preventing women from participating in adult education decreases for highly educated women. In fact, when both partners are highly educated, we can see a positive effect on transforming the traditional gender division of the care role between men and women in the family (Stoilova et al., 2020).

We also see gender differences in the probability of men and women receiving financial support for participation from their employer, which tends to be lower for women. Women receive less employer-funded training than men. There is, however, a higher probability of women receiving subsidies from public institutions to support their participation in adult education. In that respect, we see a positive effect of social investments as part of active policies, aiming at minimizing the negative effects of gender and social origin (Cefalo & Kazepov, 2018).

In discussing our results, we return to the concept of vulnerability, investigated by Maiztegui-Oñate et al. (2019) and applied in the Enliven project. Women form a deprived social group because family responsibilities are assigned to them. They are discriminated against by employers in offers of additional work-based training, especially in Mediterranean welfare regime countries. This supports the initial statistical data, showing that men participate more than women in Cyprus, Italy, and Malta. In addition, ethnic minority women tend to have limited support within the family, and their parents are also often low educated. Taken together with their low educational levels, family obligations may lower self-confidence (a dispositional barrier). For such multiple deprivations, a different set of policy measures is needed.

Implementing policies and practices such as affordable childcare and pay while participating in lifelong learning seems to be a pre-condition for low-educated women to be involved. This is especially important for women from minority ethnic backgrounds and those with low-educated parents. They need more opportunities to escape part-time jobs and inactivity. But that is not enough. Establishing a positive attitude to inclusion in adult education is an important policy task. Combating discrimination is a pre-condition for ethnic minorities and migrants to be motivated for additional and long lasting investments—particularly of time and scarce resources—in improving their qualifications.

In this chapter, we have added a fourth barrier—namely, structural inequality in societies—to the three identified by Cross (1981): institutional, dispositional, and situational. A critical message from our research is the need to look at individual level characteristics and at different categories of intersection such as gender, ethnicity, and social background as generating multiple risks of vulnerability.

In our analysis, social origin has been defined through the educational level of parents as an indicator for class. Intersectionality has been observed in the multiple disadvantages in participation in adult education that include the simultaneous negative effects of gender, class, ethnicity, and age for women compared to men, applying an inter-categorial approach and—among women—an intra-categorial approach to intersectionality (McCall, 2005). We found the negative effect of ethnicity on women to be stronger than on men. The negative effects of intersectionality for participation in adult education could not be effectively reduced under the Mediterranean and post-socialist welfare regimes when compared to the social-democratic type.

The present research also contributes to the concept of bounded agency as it relates to the account of factors and barriers at individual, group, and national level (Rubenson & Desjardins, 2009). An important finding of our analyses is that country-level welfare regimes help explain a higher proportion of the variance in the case of women than men. This holds true for all our analyses of our dependent variables. We suggest this indicates that family policies are mainly targeted at women and explains why women’s participation in lifelong learning differs considerably according to the welfare regime of the country in which they live.

Our contribution to the idea of intersectional policy evaluation, as discussed by Lombardo and Agustin (2011), lies in its extension to the policy field of adult education. We have focused on the impact of gender, ethnicity, and social origin (as an indicator of class) as barriers to participation in adult education, treating men and women as complex categories. The explicit analysis of disparities among men and women is significant not only for policy makers but also for society more widely. To increase women’s motivation and ability to participate in adult education, policy makers should adapt measures and programmes to people’s individual situations.