Does women's employment enhance women's citizenship? Open Access

In social investment (e.g. Kvist 2015) and EU (e.g. European Parliament 2000) discourses it has been argued that higher female employment rates benefit individual women and society alike, as the integration of women into paid work represents a major means to advancing financial independence of women, gender equality, economic growth, social cohesion, and – not least – the exercise of women's citizenship (Council of Europe 2005). The presumed employment-citizenship nexus rests on the argument that women outside the labour force are denied social rights as they have not made a ‘contribution’ and that employment is the basis of social respect and equal standing (Sainsbury 1994; Gutmann and Thompson 1996; Lewis 1997).

It remains an open question, however, whether women's integration into paid employment will certainly enhance women's citizenship. Researchers contributing to the literature on the employment of women argue that the feminisation of the labour force is gendered (Benería 2000). There are massive differences between the remuneration to men and women, often associated with segmented labour markets and vertical and horizontal job segregation (e.g. Grimshaw and Rubery 1997; Crompton 2002; Lewis 2003; Mandel and Semyonov 2005; Young 2010; Vosko 2011; Cuesta and Carcedo 2014). Women are employed in precarious and ‘atypical’ jobs more often than men, depriving them of full social security rights; many women work (short) part time hours, performing basic work in lower paid jobs (Gornick et al. 2008; Manning and Petrongolo 2008; Presser et al. 2008; Wilson and Ebert 2013), which results in new types of inequalities in the form of ‘working poor’ (even among full time working women) and income polarisation. Women work part time as a way of reconciling work and care and to meet work–life balance problems due to an uneven gendered distribution of domestic labour (Crompton 2006). Hence, the relationship between paid work and the exercise of women's ‘full’ citizenship may not be as clear-cut as what is sometimes presumed.

The research addressing the integration of women into the labour market is relatively extensive. Few studies, however, have empirically analysed the integration of women into the labour market from a citizenship perspective. As a normative ideal, liberal democratic citizenship is associated with equal status and the right of all human beings to equal treatment (Sweetman et al. 2011). Using citizenship as the key dependent variable, the overall aim of this paper is to consider how and under which conditions the employment of women enhances women's citizenship.

The paper is guided by two issues: first, in as much as employment matters for women's citizenship, it is expected to differ along variations in women's occupational status, which leads to Research Question 1: To what extent does citizenship differ among three categories of women? (a) the employed, (b) the unemployed, and (c) women outside the labour force? Second, work, wages, and living conditions are not the same for all working women. We therefore expect different groups of employed women to exercise citizenship differently. Research Question 2, then, is: To what extent does citizenship among working women depend on individual characteristics and their respective positions in the occupational structure?

Most studies have used a national or welfare regime approach when studying policies concerning female employment. In contrast, this paper employs a local approach. Thus, welfare policies of relevance for women's employment opportunities largely depend on local political decisions and priorities, and there are pronounced differences in welfare programmes between municipalities (e.g. Jensen and Lolle 2013). Such inter-municipal variations sometimes exceed the variations between countries, rendering it timely to view nation-states as a multitude of different ‘welfare municipalities’ (Kröger 1997), although universal national policies like maternity leave arrangements are of course relevant in local contexts.

Three European municipalities/cities have been selected for in-depth analysis: Aalborg (Denmark), Leeds (England), and Bologna (Italy). These cities are all social democratic or socialist strongholds, respectively embedded in social democratic, liberal, and conservative/Mediterranean welfare regimes. All three cities are also marked by a crucial similarity of potential relevance for women's citizenship: they all comply with the EU 2010 employment target, meaning that the employment rate in all three cities exceeds the EU 2010 60% threshold. Hence, we should be able to find similar traits in the three cities if there is a link between employment and citizenship.

Women's citizenship has been analysed previously, but usually in abstract and/or theoretical debates. As argued by Lister (2007), for instance, there is an imbalance between theoretical and empirical studies of citizenship. This paper aims to fill the empirical lacuna. The empirical analysis is based on fully comparable survey data designed to study citizenship and labour market participation among women aged 25–64.

The next section reviews the literature and main conceptions of citizenship, which will subsequently guide the construction of our dependent (citizenship) and independent variables. The empirical analysis is subdivided into two sections: a preliminary analysis addressing how citizenship is related to the labour market status of women (i.e. employed, unemployed, and outside the labour market), and regression analyses identifying relationships between different aspects of citizenship and the lives and work conditions of working women. The paper ends with some concluding remarks.

Since Marshall (1950), citizenship has been considered a central and overriding democratic ideal for contemporary societies. As a normative concept, it has been associated with notions of the ‘good life’, as it refers to the advance of the social ethic of equal social worth among all of the citizens in a given society. Citizenship is a multi-facetted concept embracing an array of dimensions. As a multidimensional, normative, and analytical concept, citizenship can be understood, first, as a status, referring to citizen's rights and obligations in different arenas of life, including the economic, social, and political arenas. Social and political rights allow citizens to live a life of a civilised being, free from unfavourable material living conditions. Second, as a practice, meaning to ‘act’ as a citizen (Johansson and Hvinden 2007: 36); for instance, people are not acting as citizens if they do not exercise their political rights by voting in elections etc., meaning that citizenship concerns participation in the economic, political, and social spheres of life. Third, citizenship is associated with a feeling of belonging and of having ‘equal social worth’ vis-à-vis fellow citizens (Janoski 1998; Lister 2002, 2004; Sweetman et al. 2011: 349). As such, citizenship is associated with fairness, including fair treatment and rewards in the work sphere, and a feeling of being fully part of a collectivity or community, which – contrary to gender discriminatory values and practices – is inclusive and promote institutional and social trust (Alesina and La Ferrara 2002; Chan and Chan 2006; Larsen 2013; Cho 2016).

The concept of citizenship is evidently complex and multidimensional. However, some dimensions of citizenship may be more important (or have more often been emphasised as important) than others for women's agency.

It is hardly surprising that the labour market is often considered to be the main route to women's citizenship (although not the only one). This has led feminists to single out the right to employment as a major citizenship issue (e.g. Yuval-Davis 1997), as exemplified by the idea of a woman's right to become ‘commodified’ (Orloff 1993), emphasising that labour force participation is considered to be essential to emancipation.

Of course, women's integration into the labour market is pre-conditioned by civil and social rights. Social rights of importance for women are public social programmes providing good quality and affordable (child and elder) care institutions (Scheiwe 1994; Knijn and Kremer 1997; Drew and Mahon 1998; Borchorst and Siim 2002; Pavolini and Renci 2008), epitomised as the ‘women-friendly’ welfare state (Hernes 1987; Voet 1998; Leira 2006). Thus, the ‘women-friendly’ welfare state supposedly relieves women suffering from problems of reconciling work and family life.

Political rights and participation are equally important for women's citizenship. Actually, some feminists argue that a political voice is essential (Phillips 2003) and that the most important practice of citizenship in the exercise of political rights in the political sphere consists in voting. ‘The work of being a citizen starts and ends at the ballot-box’ (Sweetman et al. 2011: 349).

For decades, it has been argued that women in their role as housewives are not full citizens. De Beauvoir (2009), for instance, argues that housewives live behind the scene; they are isolated from the world and suffer from not doing anything, or not being able to express themselves. Among others, however, Baldock (1994) has argued that women's participation in civil society activities such as voluntary work allow them to express social concern, make a social contribution, and get the recognition they lack in their roles as stay-at-home wife/mother. As such, social participation in the form of voluntary work may enhance citizenship among housewives.

It goes without saying that the ability to act as a citizen is pre-conditioned by access to basic material resources and means of life, since poverty or economic hardship prevents people from acting as full members of a community. The ultimate request from a feminist citizenship perspective is therefore that generous welfare benefits (if women are out of work) or adequate female wages, allow women to form and maintain an autonomous household, even outside marriage (Orloff 1993).

As suggested, citizenship includes rights, obligations, and practices in different arenas in society, and there is most likely a spill-over effect from one arena to another. For example, economic hardship may lead to social isolation, which furthers political alienation. Fraser (2003) has thus argued that material resources precondition participation, which has been confirmed by Iversen and Rosenbluth (2008) and Stockemer and Byrne (2012), who have shown empirically that female labour force participation has a positive impact on female political representation.

In this paper we make use of two types of data. First, we use (primarily) administrative data to describe how citizens’ rights are orchestrated in the local welfare systems of Aalborg, Leeds, and Bologna. Second, we make use of survey data to analyse women's self-reported material life-conditions, social and political participation, and self-perception from a citizenship perspective. Survey data constitute the core of the analysis. We do, however, draw on local administrative data to explain city variations in survey data, like, for instance, whether differences in economic hardship can be explained by differences in female wages or generosity of local welfare benefits.

The three cities have been chosen because they are social democratic/socialist strongholds. The local governments in all three cities are expected to conform to gender equality and universalistic principles (cf. Esping-Andersen 1990). The three cities are embedded in different welfare regimes with different vertical governance systems. Municipalities are subjected to a centralised governance system in England, a multilevel system in Italy, and a decentralised system of governance in Denmark (Kuronen and Caillaud 2015). Local autonomy is low in England, medium in Bologna, and high in Aalborg. Although the room for manoeuvring is low in England, there are differences between municipalities, including childcare coverage and costs. The differences, however, are much greater in decentralised governance systems, and differences can also be enormous in multilevel governance systems. Characteristics of the three cities are:

In 2009–2010, the overall female employment rate in Aalborg was 71, in Leeds 65, and 63 in Bologna. In all three cities, the occupational structure is marked by features of a service-based economy (Pfau-Effinger and Schwindt 2015), and mothers with young children (0–2 years of age) tend to have higher labour force participation than women in general. In Bologna in 2009, for instance, the activity rate of mothers with young children (0–2 years) was 74, while 67 for women in general (Maestripieri 2014). And in Aalborg, while the labour force participation rate for women whose youngest child was 0–2 years of age was 78, the employment rate for this group in general was 72. In 2009–2010, the unemployment rate for women was 3.3 in Aalborg, 6.0 in Leeds, and 4.1 in Bologna, which ties in with differences in the percentage of women working part time (39% in Leeds, 30% in Aalborg, and 30% in Bologna (Maestripieri 2015)).

Social rights in relation to child and elder care differ to some extent between the cities. Institutional childcare take-up rates for children 0–2 years of age is 69% in Aalborg, 41% in Bologna, and 35% in Leeds, while institutional childcare take-up rates for children aged 3− school age is 97% in Aalborg, 96% in Bologna, and 86% in Leeds (Kuronen et al. 2015).¹ Take-up rates for the provision of public home and residential care for people 65+ was (in 2011–2012) 35% in Aalborg, 5% in Leeds, and 4% in Bologna (Maestripieri 2015).

The national level of social participation in the form of voluntary work is very high in the UK (over 40%), high in Denmark (30–39%), and very low in Italy (less than 10%) (GHK 2010). While no local data exist, an in-depth study of Red Cross activities in Aalborg, Leeds, and Bologna seems to indicate that local patterns do not deviate much from national patterns (Jensen 2015).

Welfare benefits are available in Denmark, Italy, and Britain to individuals and/or families who are not covered by the (national) social security system and require socio-economic support. In Italy, such social benefits are totally left to the discretion of municipalities, as there are no overall or national regulations outlining benefit levels, entitlement conditions, or requirements (European Commission 2013). In Denmark, municipally financed social benefits have been defined by the central state, and different rules for different age groups exist. A provider over 30 years of age with children is entitled to DKK 14,416 per month (equivalent to Euro 1900 + housing benefits + subsidised childcare etc.); spouse's income and any personal net worth will be deducted from the benefit. Danish social benefits correspond to the UK income-based job seeker allowance, amounting to about £114.85 per week for couples and £73.10 for singles (equivalent to €612 and €390/month). Quality differences in local welfare benefits can be ranked on a scale of high (Denmark), medium (UK), and low (Italy), while the national defined maximum amount of paid postnatal (maternity + parental) leave is 9 months in the UK, 11.5 in Denmark, and 14 in Italy (Council of Europe 2009: 39).

In all three cities female wages are well below male wages. But women's wages relative to men do not differ much in the three cities. In 2007, the gender pay gap was 77 in Aalborg, 77 in Bologna, and 75 in Leeds (Flaquer et al. 2014).

As can be seen, city characteristics such as employment rates, gender pay gaps, maternity arrangements, and childcare for 3+ children support the view that the overall social and economic framework conditions for women's citizenship are not that different in the three cities. Still, the cities differ in areas such as unemployment, part time employment, childcare take-up rates for children aged 0–2, eldercare take-up rates, voluntary work, and generosity of local welfare benefits, which will be taken into consideration in the empirical analysis. It may thus be expected that differences in the quality of local welfare benefits to some extent mirror differences in the degree of economic hardship for unemployed women.

Survey data in this paper originate from telephone surveys conducted in Aalborg, Leeds, and Bologna in late 2012 and early 2013 (for a more detailed description of the data, see the introduction to this edited volume). The data offer a unique source of self-reported comparative information on the quality of living and work conditions across geographical locations from a citizenship perspective. In Aalborg, N = 809 with a response rate of 50; in Leeds, N = 800 with a response rate of about 25; and in Bologna, N = 800 with a response rate of about 28. The data from the 3 cities have been merged into a single dataset with 2409 respondents.

In principle, however, the merging of data undermines the opportunities to analyse how city characteristics affect citizenship. We have therefore included city dummies to allow for such comparisons. The merging of data is based on the assumption that the effect of a cause has the same direction in all contexts. This might be a false assumption. We have therefore analysed merged data as well as city data (not shown). In case of discrepancies between cities on specific variables (e.g. number of children), interaction variables have been included in the overall model (not shown) but without any significant results.

Citizenship is the dependent variable in this paper, and we use four survey questions as indicators for citizenship.

As a measure of material living conditions we use the following question, which is a measure for economic hardship: ‘During the last year, have you experienced any difficulties in paying for the running costs of your household (e.g. food, rent, household bills)?’ Information on the composition of the household is included in the analysis, which allows for the analysis of single women as well as women in different family constellations. Single women in particular provide insights into the economic challenges women face as individuals in different cities.

Participation in the social sphere of life is often measured in terms of the number of contacts (weekly or monthly) with friends and family. We have chosen to measure social participation as involvement in voluntary work, as it demonstrates attachment to the broader community. Thus, as a measure for social participation, we use the following question: ‘At present, do you do any unpaid voluntary work?’

As a measure for political participation, we analyse voting behaviour in elections, as this has theoretically been pointed out to be the most crucial dimension in relation to citizenship in the political sphere of life. We use the following question: ‘Do you always, usually, sometimes, or never vote in elections?’ Of course, a comparison of male and female voting behaviour in each city would have been beneficial. The data, however, do not allow such a comparison. It is also worth emphasising that voting is an overall normative citizen ideal. If women do not vote, they cannot be considered to be acting as citizens – whether men vote or not. It is therefore fully relevant to compare the voting behaviour among women in different cities, as the differences can be measured on a scale of more-or-less in terms of political citizenship.

A precondition for citizenship is that individuals have a feeling of ‘equal social worth’ and a feeling of being fully part of or belonging to a collectivity or community that promotes social trust. To analyse the trust-dimension of citizenship we use the following question: ‘Generally speaking, would you say that most people can be trusted or that you can't be too careful when dealing with people?’

The four citizenship indicators will be analysed independently of one another. To test the spill-over effect from economic to social and political marginalisation, however, the answers to the ‘economic hardship’ question will be used as independent variables in the analysis of social and political participation as well as how economic hardship affects trust.

Citizenship is a complex, multidimensional concept, while the indicators used in this paper are rather narrow. We do not claim that the four indicators for citizenship fully grasp the comprehensiveness of the citizenship concept (i.e. the validity is not perfect). But the indicators are without doubt valid in certain dimensions of citizenship. In addition, administrative data help to frame the analysis.

This paper is primarily interested in the relationship between employment status and citizenship. We are also interested in how bivariate effects are mediated by other respondent characteristics. As control variables, we have included a broad range of self-reported variables, including individual, demographic, structural, institutional, motivational, and geographical factors.

The variables possibly affecting the work–citizenship relationship and the status of women in society include individual and demographic variables, such as age and education (OECD 1989), marital status, occupational status of partner, and family composition, including the number of children (cf. Kneip and Bauer 2007; Vere 2007; Daly 2011), health and ethnicity.

Citizenship largely depends on the opportunities to find gainful employment (full or part time work) and how women are positioned in the horizontal (sector of employment) and vertical division of labour: are they employer or employee? Do they have managerial responsibilities or not? (Jaumotte 2003; Caruana 2006; Jarman et al. 2012).

The orientation of women towards the labour market, reflected as rational-actor-based preferences (Hakim 2002) or socially constructed dispositions (Bourdieu 1977), may generate social practices in relation to the labour market; but participation is possibly also involuntary. The question becomes whether it can be characterised as ‘citizenship’ if a woman with a disposition towards the domestic sphere is forced to participate in the labour market in order to alleviate economic hardship or the like. We have therefore asked working women the following question: ‘If you had a completely free choice, would you prefer a paid job or not to work at all?’

Finally, we have controlled for geographical location, as overall living conditions and economic hardship differ from one locality to another, to some extent determined by the generosity of the (local) welfare state. Appendix presents a thorough overview of the data.

This section analyses the relationship between labour market status and the four dimensions of citizenship, using Table 1 as our overall frame of reference.

The first columns reveal a significant association between women's labour market status and economic hardship measured in terms of difficulty paying running household costs. Unemployed women (35.8%) have far more difficulty paying household costs than working women (16.9%) and women outside the labour force (17.6%). It is also worth noting that the differences between working women and women outside the labour force are rather limited. Women who have chosen to be a homemaker – or for any other reasons have opted out of labour force participation – are not financially worse off than the gainfully employed.

Within this overall picture, we find marked differences between the three cities (not shown in Table 1). In Aalborg, only 3.4% of working women struggle to pay running costs as compared to 24.2% in Leeds and 25.3% in Bologna, while the level of economic uncertainty among the unemployed is 24.4% in Aalborg, 38.2% in Leeds, and 45.5% in Bologna. These patterns are in accordance with the overall differences in the generosity of municipal-based social benefits. In Bologna, however, women outside the labour force (16.5%) are better off than working women, while the opposite is true in Leeds and Aalborg. The special feature in Bologna might be explained to some extent by the fact that women 60–64 years of age (i.e. they are retired) make up a large share (66%) of the Bologna sample of women outside the labour force.² Still, the overall picture clearly indicates that unemployed women are largely subject to economic hardship and loss of citizenship as compared to working women and those outside the labour force.

Using engagement in volunteer work as an indicator of social participation, i.e. the exercise of citizenship in the social sphere of life, Table 1 shows that the expected relationship between labour market status and citizenship has been confirmed. Women outside of the labour market (and unemployed women) are more engaged in voluntary work than those who are gainfully employed. This ties in with the so-called scarcity theory (Marks 1977), which holds that time spent on market work or family responsibilities, including informal caregiving, usually reduces the time available for volunteer work, contrary to the argument by Haski-Leventhal et al. (2009), who contend that labour market participation positively affects the decision to participate in voluntary activities.

Table 1 reveals significant relationships between labour market status and voting behaviour (political participation). Unemployed women are less active than working women and women outside the labour force, who actually vote the most frequently. At the city level (not shown), it is interesting to note that voting behaviour is rather similar in Bologna (91.1%) and Aalborg (88%), while much lower in Leeds (47.4%), indicating that political integration and citizenship are less developed in Leeds than in Aalborg and Bologna; maybe also because the benefits to be achieved from voting in municipal elections are rather limited in centralised governance systems.

As to trust, there is a significant difference between those working and those outside the labour force. Working women have higher trust levels than those outside of the workforce. The magnitude of differences is minor, signifying that trust is not strongly associated with labour market participation. Women in Aalborg have higher levels of trust than do women in Leeds and Bologna (not shown), while it is not possible to detect any differences in the levels of trust among women in Leeds and Bologna.

We will now proceed to analyse how four dimensions of working women's citizenship (economic hardship, social and political participation, and general trust) intersect with individual, demographic, structural, institutional, motivational, and geographical factors. We only present the full model of all four regression analyses in this article (see Table 2), and each of these models will be commented on separately.

Differences in the experience of economic hardship (model 1) are linked to women's individual and family characteristics. A clear age effect is observed, where the experience of economic hardship decreases with age. This relationship is statistically significant for those below age 50. Moreover, the composition of the family has a marked effect on economic hardship, i.e. the greater the number of children in the household, the greater the odds of experiencing such hardship. Women living alone have a greater risk of experiencing economic hardship than those living with a spouse/partner, regardless the labour market participation. We also find that health and ethnicity are related to economic hardship; the poorer the health, the higher the odds ratio. Women born outside the three countries (Denmark, Italy, and England) have a greater risk of experiencing economic hardship and potential marginalisation compared to native-born.

Rather surprisingly, neither education nor a woman's position in the horizontal and vertical division of labour or weekly work hours has a significant effect on the experience of economic hardship, which leads to the conclusion that the risk of economic marginalisation among working women is not determined by their position in the occupational structure. As already discussed, there seems to be a clear connection between geography and economic hardship; working women in Aalborg are less exposed to economic hardship than those in Leeds and Bologna.

Participation in volunteer work – our indicator of social participation – does not interact with a woman's labour market position (cf. model 2), where Pseudo-R² is rather low (0.052), signifying that only a limited amount of the variance in the data is explained by the included variables. Some of the included variables are, however, significant. There is a clear educational effect. The least educated are less involved in the voluntary sector, whereas women with tertiary education have a higher participation rate. A recent study on volunteering in Denmark (Fridberg and Henriksen 2014) produced similar findings. But there are no spill-over effects from economic marginalisation to social participation; that is, problems with paying for running costs do not exclude women from social participation.

Furthermore, it is worth noting that family size and position in the vertical division of labour are related to the propensity to do volunteer work. The inclination of women to do volunteer work increases with the number of children, just as women with managerial responsibilities are more involved in volunteer work than those without such responsibilities, although only significant at a 10% level.

As to political participation (model 3), individual characteristics have an impact on whether or not women exercise their political right to vote. The youngest age group (25–29) in particular votes less than 30+ women. The higher the education level, the more women are inclined to exercise their democratic right to vote. Ethnicity also matters – the women born outside Denmark, Italy, and England vote less frequently than do native-born women. It is worth considering, however, that this group may actually include women not entitled to vote in elections.

Aside from marital status, family characteristics have no impact on voting behaviour. Single women are less likely to vote than married or co-habiting women, while neither the labour market status of the partner nor the number of children in the household has any effect on political participation.

Overall, the labour market positions and working time patterns of women have a rather limited impact on the political dimension of citizenship. The only significant result is that women working in industry tend to vote more frequently than those working in all other sectors. Geographical location also matters: women in Leeds vote less frequently than women in Aalborg and Bologna, but there are no signs of spill-over effects from economic hardship to political marginalisation, meaning that the economic situation of working women in no way prevents women from exercising their political rights.

Political orientations affect voting behaviour. Women reporting that they belong to the left or right of the political spectrum are more likely to vote than those who place themselves in the middle of the spectrum or women who are indecisive about their political orientation. This would appear to indicate that political engagement stimulates political participation.

When it comes to the level of trust (model 4), a clear relationship between trust levels and personal and individual characteristics exists. We find that the least-trusting women are the youngest (25–29).

Education level is highly influential on levels of trust; especially women with lower tertiary education are the most trusting. This category covers a number of the typical welfare state professions, such as nurses and teachers, which are typically associated with higher levels of trust. Women with a lower secondary education have lower levels of trust.

Turning to the household characteristics, having a partner (or not) only has marginal effects on trust, with the exception of women with an unemployed partner. This clearly indicates that having a spouse with insecure connections to the labour market is linked to women's trust levels. This result highlights an important methodological point: trust must be seen in the context of a woman's family situation; focusing on individual characteristics alone does not provide a full picture of the factors conditioning trust. Still, the number of children has no effect on women's trust levels.

Poor personal health is associated with low trust. Women reporting fair/poor or very poor health score lower trust levels than those reporting very good health. Similarly, the political orientation of women interacts with variations in trust levels. Left-wing women are high trusters, whereas women who are indecisive about their political orientation are low trusters.

The position in the division of labour and working time patterns has little effect on trust levels and no spill-over effects can be found. Experiencing economic hardship does not affect women's trust levels. Much of the variation in trust among women is due to city of residence; women in Aalborg are far more trusting than women in either Bologna or Leeds. This is hardly surprising, as similar differences in the levels of trust between the Nordics and other European countries can be found in most of the comparative literature on trust (see www.worldvaluessurvey.org).

The overall aim of this paper was to analyse the extent to which women's enrolment in the paid labour force improves their living conditions from a citizenship perspective. The paper is guided by the idea that citizenship depends on labour market status, position in the occupational structure and geographical location. Three European cities were studied – Aalborg, Bologna, and Leeds – and a most similar design has been employed. The three cities exhibit similar employment rates and working time patterns (part time/full time) for working women.

The female employment rate – also for women with young children – being relatively high in the three municipalities led by the Social Democrats in separate corners of Europe is rather remarkable, as there are considerable differences in the coverage (or take-up rates) of childcare institutions in the three cities. This reflects how childcare provisions are not a necessary prerequisites for the entry of women into paid employment. That is not to say that women in Bologna and Leeds, where coverage is low, do not face reconciliation problems; but such problems are solved by other measures, such as family care or part time work. For instance, women in Leeds, where we find the lowest coverage rates for childcare institutions, take up more part time work than do women in Aalborg and Bologna.

We have analysed three groups of women – employed, unemployed, and women outside of the workforce – and we have found that gainful employment does not automatically lead to citizenship. The relationship between labour market status and different dimensions of citizenship points in different directions. Political participation is more frequent among women outside the workforce as compared to labour force participants, while more working women have a feeling of ‘equal social worth’ (measured as trust) than those outside the labour market. A major demarcation, however, runs between the unemployed, on the one hand, and working women and women outside the labour force, on the other. That is, working women and those outside the labour force surprisingly share experiences and practices that stand in contrast to unemployed women, which is especially the case with regard to the experience of economic hardship.

Loss of citizenship due to economic hardship, however, can be counteracted by the local welfare state. Economic uncertainty is low among unemployed women in Aalborg, medium in Leeds and high in Bologna. These patterns mirror variations in municipally based social benefits, which are high in Aalborg, medium in Leeds, and low in Bologna.

It is nonetheless interesting to note that we find no spill-over effect from one citizenship arena to another, meaning that economic hardship does not tie in with lower levels of trust and political and social marginalisation. These findings are in line with Gallie et al. (2003), who found that unemployment led to poverty, but this did not lead to further social exclusion.

More specifically, if we examine the group of women who are employed, we find that there is no relationship between the position of women in the occupational structure and citizenship, though with a few exceptions: social participation and self-esteem depend to some degree on education and whether or not women have managerial responsibilities, while women employed in the industrial sector score higher on political participation than other women.

In sum, the major findings of this paper are that neither paid employment nor position in the occupational structure has major effects on women's citizenship. Of course, welfare state policies may alleviate social suffering, illustrated as to how differences in the quality of welfare benefits in Aalborg, Leeds, and Bologna counteract economic hardship to different degrees among unemployed women. From a gender perspective, however, viewing citizenship as an individual–state relationship is insufficient. Women's individual and family characteristics largely influence all of the dimensions of women's citizenship. A broader approach is therefore necessary for the analysis of women's citizenship. Further research must include the impact of economic relationships within families, and of the relations of families to the state, on the opportunities for women to exercise citizenship.

No potential conflict of interest was reported by the authors.

Per H. Jensen is professor of Social Policy at Center for Comparative Welfare Studies (www.ccws.dk), Aalborg University, Denmark. He has published widely in the fields of comparative welfare state analysis, formal and informal work, elder care, comparative labor market analysis, early exit/retirement, and the sociology of family and gender relations.

Rasmus Juul Møberg is associate professor in Department of Sociology and Social Work at Aalborg University, Denmark. His research interests include comparative welfare studies, work-life balance, and female labour market participation.