Women at risk: the impact of labour-market participation, education and household structure on the economic vulnerability of women through Europe

ABSTRACT While increasing female employment has contributed to reducing gender inequality, it has also exposed women to higher economic insecurity. The contribution of this paper is to understand the social conditions that might expose women to economic insecurity in different European cities. Specific aspects have been considered: (a) reduced (part time) work, (b) hampered labour-market participation (unemployment, involuntary inactivity due to care tasks), (c) different household structures (single/couple; with/without young children) or (d) educational level of both partners in the household. Data are based on a survey carried out in 2012 in seven European cities, representative of the different welfare/gender regimes in Europe. The results show that the most important divide is between women cohabiting/not cohabiting with a partner. Splitting the analysis on these two groups of women, differentiated configurations of conditions exposing women to economic insecurity have emerged in different welfare/care regimes. While Nordic, Central-eastern and Anglo-Saxon cities substantiate an individualised model of exposure to economic insecurity mostly driven by women’s participation in the labour market, in Continental and Mediterranean cities insecurity mainly depends on the educational levels (mainly of the partner in the case of coupled women) and the organisation of the household (presence of children).


Introduction
The last two decades have seen an increasing level of female education and a corresponding growth in female labour-market participation (Esping-Andersen 1999), with markedly positive social consequences. Higher female labour-market participation has increased the number of people with access to independent income and has helped to reduce the general level of poverty and social exclusion. Moreover, it has contributed to reducing gender inequality in the labour market, in income distribution, and in access to social citizenship in countries where welfare benefits depend on the position of individuals in the labour market.
On the other hand, increased gender equity has produced further differentiations in resources, power and independence, thereby increasing inequality and social vulnerability among women (Mandel 2012). To date little research has investigated the social differentiation occurring among women, and the factors explaining the higher vulnerability of some of them. Given that nowadays women play a key role in mediating between the demands of the labour market and societal needs (Esping-Andersen 2009), investigating the factors jeopardising that role seems crucial.
The aim of this paper is to contribute to filling this gap. Based on original data gathered in seven European cities, its main goal is to determine the principal factors causing female economic insecurity. By economic insecurity we mean a situation in which households have strong difficulties to autonomously attain an adequate living standard even though they are not necessarily in a situation of income poverty or material deprivation (Curatolo and Wolleb 2010). This concept has been proposed in order to shift the focus from poverty or current deprivation to economic stress and exposure to shocks (Whelan and Maitre 2010). While poverty and material deprivation show the absolute or relative lack of financial resources or basic goods in a household, measures of economic insecurity provide a complementary perspective by showing to what extent available financial resources allow households to autonomously pay their running costs or face unexpected economic contingencies. It is precisely economic insecurity the most severe difficulty that women are undergoing in crisis times: their increased participation in the labour market and greater individual autonomy from family relationships have exposed most of them to economic uncertainty and income instability. It can be said that autonomy and individualisation have their economic costs, of course, but the issues at stake here are inequalities among women and the exposure of their households to economic insecurity.
Previous research has already identified the main factors contributing to poverty risk for women. The most relevant ones are: partial or temporary inclusion in the labour market, unemployment, childcare duties due to the presence of young children, family breakdowns, low qualifications with consequent low wages, difficulty in conciliating work and care because of lack of welfare support and/or family ties. In this paper, various factors contributing to economic insecurity are jointly examined. On the one side, we considered variables related to individual characteristics, such as women's participation in the labour market and their educational level; on the other side, we considered variables related to the composition of their households and, for women living in couple, the educational levels of their cohabiting partners, controlled by homogamy.
Moreover, configurations emerging in different cities were compared in order to understand the role played by welfare systems in protecting women from economic insecurity. While there is extensive research on women's socio-economic conditions, few studies have considered all these factors together, and how their impact is shaped in different contexts.
Therefore, the main contributions of the analysis carried out are as follows: (a) to shift the focus from poverty to economic insecurity in order to capture the individual situations characterised by instability/ uncertainty rather than by hardship or material deprivation; (b) to develop a multidimensional explanation of economic insecurity and (c) to consider how these factors are distinctively combined in different urban contexts.

Literature review
Women's exposure to poverty risk or material deprivation has been explained by considering either micro or macro factors: on the one hand, individual and household characteristics (such as age, labour-market integration, social class or care responsibilities), and on the other hand aspects that are mainly related to the welfare system (Sainsbury 1996).

Individual and household factors
Age is one of the most important differentiating factors regarding economic insecurity among women. Although it depends also on welfare states, studies have shown that even in Western countries older people are more likely to be poor than working-age adults (Brady 2009). Because of women's longer life expectancy, the share of women among the elderly is higher, and older women are more likely to be poor (Ginn 1998;Bianchi 1999;McLanahan and Kelly 1999). More relevant for our analysis, results pertaining to age groups among working populations vary from country to country in Europe: for instance, in most cases the young are more likely to be working poor on an individual basis (Crettaz 2013).
This effect is closely connected to the influence of women's household/ marital status and employment affiliation. Notwithstanding significant cross-national variations in the extent of poverty for single mothers, research shows that single mothers tend to be poorer than all other social groups (McLanahan et al. 1995;Christopher et al. 2002;Hansen et al. 2006;Misra et al. 2007;Ranci 2010;Barcena-Martin and Moro-Egido 2013;Kim and Choi 2013) and that they experience more constraints in combining work, care and leisure in daily life (Bakker and Karsten 2013). Regarding female-headed single parent households, in most of the European countries the general trend shows that few women can provide for their household without the intervention or assistance from their extended family or the state (Misra et al. 2007;Gradin et al. 2010;Chzhen and Bradshaw 2012;Barcena-Martin and Moro-Egido 2013).
Compared to single women, married women or women cohabiting with a partner are less threatened by poverty owing to their partner's earnings, income pooling, fixed cost sharing and other income sources (Crettaz 2013). The poverty risk is especially lower in dual-earner households, although in these cases economic insecurity may increase with the birth of children. It has been shown that association between poverty and the number of children may vary across welfare states, for example, in Britain the households with three or more children are more deprived compared to households with one or two children, but this is not the case in Sweden and Finland (Halleröd et al. 2006). The birth of children, especially in the case of a third child, heightens a household's immediate poverty risk in two main ways: the addition of another household member increases the financial burden on the household, and the need to provide parental care impairs women's earning capacity (Ruspini 1998). Parentsusually the mothermay reduce their time devoted to paid work or withdraw from the labour market, with loss of income as a consequence. Evertsson (2013) points out that even in Sweden, the entry into motherhood leads to temporary lower work commitment. Besides increasing women's personal vulnerability, this also increases the risk of deprivation for the household (Halleröd et al. 2006).
As regards female labour-market participation, several studies have confirmed that full-time work significantly lowers the risk of poverty (Halleröd et al. 2006;De Graaf-Zijl and Nolan 2011;Barcena-Martin and Moro-Egido 2013). Halleröd et al. (2006) point out that the unemployed, the early retired and even part-time workers are more deprived than full-time workers. Hansen et al. (2006) maintain, on the basis of German and Norwegian data, that the welfare state is unable to protect households against income losses when the head of the household only works part time or is unemployed. Thus, women are exposed to poverty in two ways: on the one hand, they are dependent on their husbands' educational and occupational resources (Gesthuizen and Scheepers 2010); on the other hand, their own opportunities to enter and stay in the labour market are more limited compared to those of men (Rubery et al. 1999).
While women are more likely to be unemployed, inactive, precarious or confined to short work hours compared to their male counterparts, this effect is greater when their professional profile is lower (Rubery et al. 1999). Moreover, the differentiation in the levels of poverty risk among women on the basis of their education has recently increased. Misra et al. (2007) point out that highly educated women are less likely to be poor and, as argued by Barcena-Martin and Moro-Egido (2013), women's higher education reduces the probability of entering into poverty and increases the probability of exiting from poverty by about 60%. The capacity of higher education to reduce poverty is partly due to the employment status of the most educated women: that is, higher educated women are more likely to participate in the labour market (Keck and Saraceno 2013). Moreover, there is still high educational homogamy among couples in European countries (Kalmijn 1998;Smits 2003): higher educated women have partners with a high education and thus better earning capacities. Homogamy increases also the magnitude of the differentiation among households: families with highly skilled dual earners have a limited risk of poverty compared to households with only one low-skilled breadwinner or two low-skilled workers.
Finally, families with an immigrant background run a higher risk of poverty, and women in immigrant households are even more vulnerable. Depending on national social assistance regulations, they might depend more on social assistance than the rest of the population, as it happens in Sweden (Mood 2011). On the other hand, the gender gap in poverty among immigrants may be lower compared to the main population (Elmelech and Lu 2004).
To sum up, there is evidence that women are more exposed to poverty risks than men. The main factors explaining this situation are the following: mothering and consequent care-work reconciliation problems; more difficult access to full-time employment and a higher concentration in precarious or part-time jobs; greater exposure to unemployment or inactivity; fewer opportunities to use high qualifications in order to develop top careers; migration status. Moreover, for women more than for men, the risk of poverty is concealed, and may be revealed due to union dissolution or loss of the male breadwinner's salary. This means that on a household basis, women may very well cope with the situation due to the male breadwinner's salary and costs sharing. However, this kind of economic coping is very fragile: union dissolution or loss of the male breadwinner's salary may lead rapidly into high poverty risk. We can assume that these same factors can be differentiated among women and that accordingly for some of them the exposure to economic insecurity as well as to poverty can be higher.

The welfare system as a frame for women's vulnerability
The extension and generosity of welfare services facilitating female labourmarket participation are of crucial importance for the prevention and/or mitigation of women's economic insecurity (Lewis 1992;O'Connor 1993;Sainsbury 1994Sainsbury , 1996O'Connor et al. 1999;Daly 2000). In general, women's labour-force participation is higher in Western-European countries with 'progressive' welfare systems, characterised by extended family-oriented services and large accessibility to public childcare facilities (Mandel and Semyonov 2006). Although in post-communist countries female employment has been very high, the services and the quality of formal care provision has been lower (Deacon 1992). As Ferge (2001) concluded, the collapse of the Soviet Union meant that the paternalistic care of the state was replaced by family self-reliance strategies, where women had to take over the informal care-giving. A similar tendency is mentioned by Zdravomyslova (1995). Despite these differences among welfare states, in many European countries municipalities maintain crucial responsibilities in providing publicly organised and subsidised childcare services (Kuronen et al. 2012). Therefore, this 'progressive' context is not determined only by the welfare state but is also supported or limited by policies and social measures provided for at municipal level (Kröger 1997).
Not only women's access to the labour market but also their work conditions depend on the public regulation of the labour market. While a high level of decommodification of services tends to stifle the growth of lowwage jobs in the private sector (Scharpf 2001;Iversen 2005), it increases intra-class inequality since the absence of uniform social rights makes part time and intermittent workers more vulnerable to wage discriminationand these workers are women in their majority (Mandel and Shalev 2009). Stier et al. (2001) found an association between the level of women's labour-market integration and the level of disadvantages related to intermittent or part-time work. Female labour forces in countries with highly supportive policies for female labour-force participation are more heterogeneous and not penalised by intermittent or part-time work. Whereas, in countries where the traditional gender division of labour and the preservation of women's familial roles are maintained, full-time work after childbirth is less established and labourmarket penalties for 'nonstandard' employment are the highest. However, women's labour-market integration also depends on the availability of services provided at the local level (Flaquer and Ranci 2011).
To conclude, the role of the welfare system has been considered crucial in supporting not only women's labour-market inclusion but also their employment stability and level of wages: for example, the presence of childcare services facilitates full-time work, while part time and temporary jobs become more frequent when such services are less generous. Moreover, inequality among women depends on institutional factors: the higher the level of commodification, the more likely inequality and dualisation become. However, most of the public services supporting female labour-market participation are provided at the local level under the direct responsibility of municipalities. While this local variability has often been acknowledged, so far no research on female poverty and vulnerability has included the local level in cross-country analyses.

Data and method
Based on this discussion, this paper explains women's exposure to economic insecurity in the context of their households on the basis of their and (possibly) their partners' individual characteristics as well as of their household composition (having a cohabiting partner or dependent children). We considered seven European cities separately and contrasted them in order to obtain configurations that can be interpreted as associated to different welfare systems. The database used was produced by the EU project FLOWS. Please see the introduction of this issue for the design of the research. The seven cities considered in this analysis cover all the European welfare regimes and are: Leeds (UK); Aalborg (Denmark); Nantes (France) and Hamburg (Germany); Brno (Czech Republic); Bologna (Italy) and Terrassa (Spain). The study was conducted on a total amount of 4836 women, with a range between 630 and 740 women for each city.
We use the following variables. Economic insecurity is defined as 'the exposure of a household to the likelihood of suffering economic hardship as a result of negative events affecting its ability to ensure that each of its members enjoys the living standards prevalent in the country of residence' (Curatolo and Wolleb 2010: 58). In our study, economic insecurity is measured by means of an indicator combining two questions: (a) During the last year, have you experienced any difficulties in paying for the running costs of your household (e.g. food, rent and household bills)? (b) If needed, could you pay an unexpected bill of ***? (The amount of the bill was different for each country and it was calculated as around one quarter of the average monthly net earnings 1 ). The outcome variable is a dummy: a woman's household is considered economically insecure if the interviewed answered affirmatively to one of these two questions.
Women's labour-market integration is calculated as a categorical variable assuming full-time employment as the reference value. Women are distinguished among: inactive (pensioners and students are excluded from the sample), part-time workers (less than 30 hours per week) and unemployed/discouraged women (by 'discouraged' we mean women who have become inactive because they see no jobs available). Other independent variables include: age, migrant background (at least one parent originally from a poor country), level of education of the women and of the (possible) cohabiting partner, presence of a cohabiting partner and possible number of children aged 18 or under. In order to control for educational homogamy, a dummy indicating if a woman living in couple has the same educational level as her partner was included in the model. Table 1 shows the main characteristics of the participants in each of the cities included in the study. We used a logistic regression model to assess the risk of economic insecurity experienced by women, running two different models for each of the cities included in the study: one for women with a cohabiting partner and one for women not living with a partner. The results for each city were compared in order to assess the different configurations of factors by city (Tables 2 and 3) and will be discussed in the following paragraph.

Results
Preliminary data investigation showed that the presence of a cohabiting partner was the only explanatory factor present in all cities. This result is consistent with previous studies demonstrating that the household composition is the key factor influencing individual exposure to poverty risk or material deprivation (Curatolo and Wolleb 2010;Fusco et al. 2011;Kim and Choi 2013;Maestripieri 2014). Not living with a partner not only is a major factor of vulnerability, but it also exposes women to specific factors of insecurity related to their work conditions or educational level, and to the actual chances of reconciliation should they have dependent children. On the other hand, women with cohabiting partners are more likely to be exposed to factors that pertain to their partner, to the gendered division of labour dominant in the couple and to specific opportunities offered to couples by the welfare system. In order to understand these different situations, the analysis was split into two parts: one focused on women cohabiting with a partner, the other related to women not cohabiting with a partner. The presence of cohabiting children under 18 years of age was analysed in both cases.

Women cohabiting with a partner
The most common insecurity factor for women cohabiting with a partner, in almost all the localities (with the exception of Leeds), is the educational level of their partners (Table 2). However, female education is significant only in two cities, Nantes and Terrassa, where also the partners' education level plays an important explanatory role. Control by educational homogamy (Table 2) shows that this aspect does not explain economic insecurity, in all the cities with the exception of Aalborg where, in the case of monogamous couples, the likelihood of economic insecurity is reduced. The most common factor explaining the economic insecurity of women with a cohabiting partner is, therefore, related to the educational level of the male partner, while women's educational level counts only in specific cities.
Besides this general result, the model shows four different configurations of factors in different groups of cities. The first configuration is distinctive of Nantes (France), Terrassa (Spain) and Hamburg (Germany). Here economic insecurity depends on the educational level of both partners (in the case of Nantes and Terrassa) or of the sole male partner (in the case of Hamburg), and is not influenced by the labour-market position of individual women (only in Terrassa, parttime positions are significant). Furthermore, in Nantes and Terrassa the educational level of women has significant weight independently of that of their partners, showing that lower qualified women are the most exposed to economic insecurity due to greater difficulties in entering the labour market and/or lower wage levels. To sum up, in this configuration exposure to economic insecurity basically depends on mechanisms of social reproduction by which access to education is unequally distributed among women and, even more crucially, among their partners.
The second configuration is typical of Bologna (Italy) and can be defined as a 'weakened male breadwinner' configuration. In this city, economic insecurity is significantly affected by women's inactivity, combined with their partner's educational level. In this case, a household structure based on a single male earner determines the condition of risk and the educational level of women's partners increases household economic insecurity, although the two conditions are reciprocally independent. Both factors indicate a weakening of the traditional male breadwinner household, which has become less financially sustainable due to the need of households to have a dual-earner income in order to bear current costs and unexpected expenditures (Esping-Andersen 2009).
Even in this configuration a clear effect of the educational level is visible, which is based on the male's position.
The third configuration is characterised by a 'predominance of the dual worker family', and it is characteristic of Aalborg (Denmark) and Brno (Czech Republic). Economic insecurity in this case is not due to role asymmetries in the household occupational structure (configuration 2) or to the educational level of the household (configuration 1). Here the main explanatory factor is female unemployment, as long as it reduces the household's income and is not adequately replaced either by unemployment benefits or by the partner's earnings. Education of both partners, controlled by homogamy, does not play an important role in this respect.
Finally, Leeds (UK) is characterised by a model of 'labour-market exclusion', where economic insecurity depends on three main factors: the exclusion from the labour market (inactivity), being a foreigner, and age. In this city, segregation from the labour market and ethnic or age discrimination represent major stress factors and increase the chance of losing economic independence.
Before concluding, it is worth mentioning that the existence and number of children aged under 18 has no significant effect on women's economic insecurity, with the sole exception of Terrassa. Childcare responsibilities, in other words, may have an impact on the level of female labour-market participation and produce disadvantages for women with respect to men; but they do not seem to play a significant direct role in producing financial inequality at household level when women are cohabiting with a partner.

Women not cohabiting with a partner
In the case of women not cohabiting with a partner, typical configurations are less consistently found (Table 3). There is no common factor explaining the economic insecurity of these women across the studied cities. However, it is possible to highlight several patterns that persistently emerged in cities both for women with partners and women living alone.
A first configuration characterises again Aalborg and Brno and can be defined 'labour market driven'. For single women in those cities the most important factor protecting from economic insecurity is full-time employment: being unemployed (as it is for women cohabiting with partners) or inactive (in the case of Aalborg) is actually the real problem. Furthermore, in the case of Aalborg, in addition to employment-related factors, women with a migrant background are particularly exposed to economic insecurity when they are not cohabiting with a partner. In these cities, the presence of young children significantly affects the exposure to economic insecurity.
The other cities show an opposite configuration, here defined as a 'family-driven' configuration. In this case, the care load due to the presence of children aged 18 or under and women's educational level are the two main explanatory variables. Conversely to what was shown in the previous configuration, in these cities women's labour-market position is not consistently of significant importance, with a few exceptions (unemployment is significant in Terrassa and part-time work in Nantes).
Finally, in Hamburg, single women show an intermediate pattern of exposure to economic insecurity because almost all the factors investigated non-participation in the labour market, presence of young children, low educational levelsplay a major role in determining the exposure to economic insecurity. However, age, migrant background and part-time employment do not significantly affect the dependent variable. Therefore, the accumulation of a series of factors (activation, care load and educational level) explains, in a very affluent city such as Hamburg, the conditions of economic uncertainty experienced by a minority of women (economic insecurity affects 29% of women in this city, which is significantly lower than the other cities in our sample with the sole exception of Aalborgsee Table 1).
In conclusion, it is important to mention that for women not cohabiting with a partner, the presence of at least one dependent child is a significant factor that affects the household economic insecurity in almost all studied cities (with the exception of Brno and Aalborg) as opposed to women living in couple, where the care load was relevant only in Terrassa.

Conclusions
In many ways women's greater labour-market participation has often been considered as contributing to stronger economic independence and higher social equality. Welfare policies have been considered to play a crucial role in achieving this through the delivery of care services, leave programmes and support for labour-market insertion. However, our analysis only partially supports this contention. The presence of different configurations of factors explaining women's exposure to economic insecurity means that, according to our analysis, women's position in European societies does not depend on the same circumstances everywhere. Furthermore, the constellation of factors causing female economic insecurity is very wide and only partially based on their labour-market participation.
Two main conclusions can be drawn from this analysis. First of all, there is a close overall association (with only one exception in this study) between economic insecurity and the presence of a cohabiting partner. Indeed, women not cohabiting with a partner tend to suffer more from economic uncertainty, regardless of age, children under 18 or whether they work.
Secondly, labour-market participation, household composition and educational level jointly influence women's exposure to economic insecurity. However, the relevance of such factors and their association are differently shaped in our cities, giving rise to specific configurations of factors that are summarised in Table 4. From this scheme, a general consistency emerges between configurations related to women with cohabiting partners and those related to single women, confirming that specific local conditions peculiarly shape female economic insecurity.
Four basic patterns emerge from our analysis. In Aalborg and Brno, unemployment, or inactivity, or even part-time work are the main drivers of economic insecurity for all women. Age or migrant status matter as they increase the probability of being in a difficult position in the labour market. The fact that both women cohabiting with a partner and women living alone face the same configuration of problems confirms that labour-market exclusion is the crucial determinant of insecurity in these cities.
A second configuration is proper of three Continental cities (Nantes, Terrassa and Hamburg), where the organisation of households (presence of a partner and/or children) and the educational level of women and/or of their male partners are the dominant factors. In these cities, the level and quality of female participation in the labour market significantly influence insecurity only for women living alone, while these aspects are not relevant for women cohabiting with a partner. In such cases women's exposure to insecurity basically depends on women's and their partners' education levels (controlled by homogamy). In these cities a 'social reproduction' configuration is dominant, characterised by insecurity mainly explained by the low educational level of one or both partners. A third configuration is found in Bologna, where insecurity is mainly explained by the presence of households organised around the traditional male breadwinner model, where women are inactive and their partners mostly have a low educational level: a situation that highlights the demise of the male breadwinner family model.
Finally, in Leeds we found a configuration characterised by 'labourmarket exclusion', where economic insecurity basically dependsboth for women living alone and women cohabiting with a partneron the exclusion from the labour market, reinforced by ascribed characteristics (such as age and migrant background), and with no role played by different forms of labour-market participation or a specific household composition.
Our final typology (Table 4) shows that, in Europe, female economic insecurity depends on various aspects. Women's weak labour-market position is a major source of economic insecurity in cities where a dual worker family model is dominant. This significantly happens in cities located in a Northern or in a Central-Eastern country. In other contextsincluding continental and Mediterranean citieseconomic insecurity strongly depends, despite the increase in female employment until 2008, on the educational level of women's partners, controlled by homogamy in the case of women cohabiting with a partner. For women cohabiting with a partner, a weak position in the labour market does not directly affect the economic insecurity of their households; a fact that can be explained by a number of different reasons, including access to unemployment benefits, the combination of female inactivity or part-time employment with a high level of education of their male partners, and/or the limited relevance of the income depending on female employment. Therefore, in these contexts insecurity is mainly the result of female dependence on low-educated partners. In contexts where gender imbalance in the labour market is still widespread, as in the case of Bologna, distinctions based on partners' education are combined with traditional household characteristics to mainly explain women's economic insecurity. In Leeds, insecurity mainly depends on labour-market exclusion, reinforced by a migrant status for women cohabiting with a partner and by the presence of children in the case of women living alone.
Finally, female insecurity is also expected to depend on the characteristics of local welfare systems. From our analysis we can hypothesise that the more the local welfare systems are individualised and gender neutral, the more women's insecurity depends on their labour-market inclusion; the more local welfare systems are based on a conception of family/household to be preserved and the more they are gender biased, the more women's insecurity is affected by educational levels and household compositions. The consistency of our results both for women with a cohabiting partner and single women shows that contextual differentiation is significant throughout Europe. Two main clusters emerged: a Northern-Central-Eastern cluster and a Continental cluster. The latter is further differentiated into two sub-groups depending on the level of diffusion of traditional male breadwinner households. It is likely that the level of care services and support for women's labour-market inclusion play a role in explaining this distinction. Future research will be necessary to clarify more precisely the distinctive role played by different welfare systems in explaining female insecurity.
To conclude, if we consider economic insecurity to be a characteristic that pertains to households rather than individuals, our general results show that the individual contribution of female full-time employment to the protection of households from economic insecurity is significant only in Northern, Central-Eastern cities and British cities, while in Continental and Mediterranean cities this is not substantial. In the former, households are likely to be economically insecure if women have a marginal role in the labour market or they are excluded from the labour market or with great difficulties to reconcile care and work in the case of mothers living alone; their work greatly contributes to their household material conditions. Whereas in the latter, the household economic robustness is still grounded on the educational level of the male. In one city (Bologna), the inactivity of women is a relevant factor of vulnerability for households structured according to the very traditional male breadwinner model. More generally, in these cities women who work part time, or are unemployed or inactive, do not significantly affect the exposure of households to economic insecurity. While this aspect may depend on various factors that should be further investigated on the basis of closer empirical inspections, our research clearly shows that in continental European cities exposure to economic insecurity still crucially depends on women's strong economic dependence on their partner's earnings.