Cause and effects of female labour force participation in local welfare systems

Introduction

For quite some time, it has been argued that the division of labour between the genders in family and society begets relations of dominance and subordination between men and women. Already towards the end of the nineteenth century, Engels (1972) argued that it was not possible for women to achieve equality with men as long as they did not take part in production, referred instead as they were to domestic labour. In the last 30–40 years, this argument has been advanced by feminist scholars claiming that women’s emancipation is preconditioned by a ‘right to be commodified’ (Orloff 1993: 318) and the formation of the dual-earner model (e.g. Gornick and Meyers 2008). More recently, however, the integration of women into the labour market has not merely been linked to the question of social equality and women’s independence; female labour market participation has increasingly been justified in economic terms: more women in the labour market will increase the number of taxpayers, thereby ensuring the sustainability of the welfare state (Esping-Andersen et al. 2002). This economic justification for increased female labour market participation has clear connections to the European Union (EU) discourse concerning the social investment state, which resulted in Lisbon in 2000 in an EU objective to increase the employment rate for women to more than 60% by 2010.

It is thus hardly surprising that female labour market activity has been the object of comprehensive theoretical and empirical studies in the course of the last 20–30 years. Thus far, however, most comparative studies of the timing of female labour market integration, the forms that it takes, the conditions under which it supports gender equality, and the extent to which paid employment reduces social inequality, have been based on national data and national comparisons. This is rather remarkable, since intra-national variations in female employment rates in some instances are even greater than the variations between countries. In 2012 in Italy, for instance, female employment rates varied between 27.5 in Campania and 61.4 (i.e. beyond the EU 2010 target) in Emilia Romagna.

Similarly, social, institutional, and economic structures conducive to female employment vary within countries. Urban environments are particularly supportive to the employment of women, just as female employment opportunities are especially favourable in localities with a strong service economy. Moreover, it is important to keep in mind that the municipalities are generally responsible for the welfare policies that supposedly impact female labour market opportunities (e.g. Kröger 1997). As local bodies of government have some autonomy with regard to policy formation as well as the design and implementation of policies, considerable regional differences in the provision of social services, such as child and eldercare, can be found (e.g. Trydegård and Thorslund 2010; Jensen and Lolle 2013). In Italy in 2010–2011, for instance, the percentage of children (0–36 months) in municipal day care centres varied from 2.6 in Calabria to 29.4 in Emilia Romagna.

Given the massive local variations in institutions, policies, processes, and female employment practices, the aim of this special issue is to contribute to our knowledge of the causes and effects of female employment in local contexts. The special issue consists of four articles trying to fill the empirical lacunae about (1) the factors conditioning female labour market participation and (2) how female employment has an impact on women’s living conditions in local contexts. For comparative purposes, the articles draw selectively on a pool of nine cities embedded in different welfare regimes which were selected for in-depth analysis: Aalborg (Denmark), Jyväskylä (Finland), Brno (Czech Republic), Leeds (England), Hamburg (Germany), Nantes (France), Bologna (Italy), Terrassa (Spain), and Hsinchu (Taiwan). Different combinations among these cities are analysed by the four articles included in this special issue.

The first two articles deal with women’s integration into the labour market. The aim of the first article, by Per H. Jensen, Rasmus J. Møberg, Ralf Och, and Birgit Pfau-Effinger (Explaining differences in women’s working time in European cities), is to explain differences in women’s formal work hours that have so far been under-researched. The paper employs a complex explanatory framework including family-cultural orientations, welfare institutions, individual and family characteristics, and the character of local labour markets and work conditions. The second article, by Yueh-Ching Chou, Birgit Pfau-Effinger, Teppo Kröger, and Costanzo Ranci (Impact of care responsibilities on women’s employment), revisits the discussion about how women’s work pattern is influenced by child or eldercare responsibilities – even work time pattern among women with grandchildren is considered. Apart from analysing grandparents’ care responsibilities the major innovation of this paper is that European cities are compared with a city embedded in an East Asian (Taiwan) welfare state regime.

The next two articles focus on how women’s formal and informal employment influences their life conditions. The third article, by Per H. Jensen and Rasmus J. Møberg (Does women’s employment enhance women’s citizenship?), analyses how women’s employment status has an impact on women’s citizenship. Understanding citizenship as the ‘de facto full membership of society’ (cf. Andersen 2005), was to a large degree a reinvention of the concept by feminists so as to describe how women are affiliated to dominant and subordinated social positions (Siim 2000; Lister 2002). The concept of citizenship, however, has mainly been used in abstract and/or theoretical debates, and to a lesser extent in empirical analyses. The paper seeks to bridge theoretical and empirical analyses of women’s citizenship with the aim of exploring the extent to which citizenship differs between three groups of women: (1) gainfully employed, (2) unemployed, and (3) women outside the labour force.

Female employment has contributed to a reduction in gender inequality. It is often argued, however, that the feminization of the labour force also has been accompanied with a social polarization among women and that some groups of women have become subject to higher degrees of uncertainty and vulnerability. In the fourth – and final – paper, by Kairi Kasearu, Lara Maestripieri, and Costanzo Ranci (Women at risk: the impact of labour-market participation, education and household structure on the economic vulnerability of women through Europe), the aim is to analyse the factors causing economic insecurity among women in seven European cities. Major explanatory factors are individual and household characteristics, but differences in welfare systems protecting women from economic insecurity are also considered.

All four articles build on survey data collected as part of the EU project FLOWS¹ among women aged 25–64 years. When the survey questionnaire was constructed, we drew on social science traditions that have already discussed the phenomenon at the aggregate national level and formulated concepts that are central for understanding female labour market participation and living conditions. It soon became clear, however, that findings were contradictory and that no single causal factor could explain either the inclination of women to enter the labour market or the effects of female labour force participation.

Thus, the novelty of the four papers included in this volume can best be understood against the backdrop of earlier scholarship. We have used existing literature – and research gaps in this literature – to formulate hypotheses that have been transposed into survey questions (for a full overview of the questionnaire, see Jensen and Møberg 2016). The hypotheses tested stem from a series of previous works. It is important to present at this point a theoretical synthesis of that literature, which was the basis for the survey that all four papers draw on.

Hypotheses

Our aim was to identify factors that in the existing literature have been assumed to influence female labour market activity, and the discussion of these factors is organized as follows: Firstly, the factors explaining women’s entry into paid employment are scrutinized, where a distinction is drawn between individual characteristics, household composition, labour market (demand-side) features, welfare policies, and cultural factors. Secondly, cultural factors are especially discussed, as the link between cultural factors and the disposition of women towards the labour market has been under-researched in studies of female labour activity. Thirdly, research on the effects of female labour activity is surveyed.

Factors conditioning female labour force participation

A long range of studies have cast light on how individual characteristics such as age, health, and education are strong predictors of female labour force participation (OECD 1989; Gustafsson et al. 1996; Gornick et al. 2008; Grunow et al. 2012). Labour force participation wanes with age and poor health, while higher levels of female employment are associated with higher levels of educational attainment and the opportunity for vocational training and life-long learning (Yeandle et al. 2009).

Moreover, household composition is relevant for the employment and work time patterns of women, as households are generally the basic unit for decision-making (Becker 1965). Whether women offer their labour thus depends on their marital status, the economic situation of the household, and their husband’s income (Hakim 2002; Matysiak and Steinmetz 2008; Stähli et al. 2008). Married women with a very low income are more inclined to maintain their connection to the labour market than are women with a high-earning spouse. Factors such as divorce, childbearing, number of children, and the age of the youngest child have also been found to have an effect on female labour supply (Kneip and Bauer 2007; Vere 2007).

Numerous studies have analysed the extent to which labour market features are conducive to women’s labour force participation. It has been argued, for instance, that wage rates, minimum wages, and the male–female wage gap have a great impact on the female labour supply (Blau and Kahn 2007; Attanasio et al. 2008): the higher women’s wages (relative to men), the higher the utility of paid employment. In particular, it is assumed that high minimum wages stimulate the labour market participation of lower-educated women, including lower-educated mothers (Cloïn et al. 2011). Financial incentives alone, however, do not suffice for that outcome. Occupational and workplace characteristics are equally important for female labour activity. The service economy has proven to be ‘open’ towards female workers (e.g. Pissarides et al. 2005);² that is, the service economy is a major arena for the demand of women workers. At the same time, women have been oriented towards the service sector because it offers flexibility and the opportunity to work part time (Jaumotte 2003; Caruana 2006; Plantenga and Remery 2009), which allow for the reconciliation between work and care obligations. It is nonetheless questionable whether part time work holds emancipatory potential, since part time employees are usually positioned in subordinate positions in the labour market. Based on the Nordic experience, however, Nätti (1993) has shown that part time work may function as a bridge for full-time integration into the labour market.

Another factor contributing to the decision about whether or not to work is if work is associated with expenses. The price of childcare will, ceteris paribus, have significance for the decisions made by mothers as to whether or not to work (OECD 1985; Berger and Black 1992; Blau and Kahn 2007), which is of particular relevance for women in low-paying jobs. Similarly, reliable and affordable public transport makes a difference (Buckner 2009: 61) as women are more likely than men to work locally (Yeandle 2009: 5). Transport expenses do not necessarily have an impact on the decision made by women as to whether or not to work if such expenses can be written off as tax deductions, as is the case in Denmark. As such, the actual costs associated with working might largely depend on the support of the welfare state.

It is almost undisputed that welfare state policies are crucial for female labour force participation, as the welfare state is the major institution in modern societies redistributing income, opportunities, and life chances by means of cash benefits, welfare services, and taxation policies.

As to cash benefits, it is often argued that generous maternity, paternity, and parental leave schemes allow women to remain in contact with the labour market during maternity (Ondrich et al. 1998; OECD 2002; Gornick et al. 2010), although the actual impact of these schemes on female labour force participation is somewhat disputed (Morgan and Zippel 2003; Aisenbrey 2009). Cash benefits also signal whether (part-time working) women are considered to be an integrated part of the workforce. In Denmark, part-time workers have been enrolled in the unemployment benefit system since the late 1960s, primarily to counteract the so-called discouraged worker effect; that is, eligibility for unemployment benefits is meant to prevent part-time working women who become unemployed from returning to the family and become full-time homemakers.

Comparative welfare state research has strongly argued that welfare services such as public childcare and eldercare institutions are major factors in women’s employment prospects (Daly and Lewis 1998; Lewis 2002; Kangas and Rostgaard 2007; Hegewisch and Gornick 2011; Pfau-Effinger and Rostgaard 2011). There has been particular focus on public childcare facilities for mothers with young children (Orloff 1996; Leira 2006; Lister 2009), while care obligations seem to have a limited impact on the decisions made by older women (Frericks and Meier 2012). However, recent research has further shown that care policies tie in with immigration policies, as female immigrants from poorer countries are important as potential carers in some European countries (Pfau-Effinger et al. 2009).

The taxation system supposedly has a huge impact on female labour force participation. It is often assumed, for instance, that the tax level – or the post-tax wage – helps structure female labour market activity (Smith et al. 2003; Saczuk 2004). Thus, rational actors will refrain from working if taxes eat away the utility of working or if the social welfare benefits are higher than the obtainable wages (Keeley 1981; Summers et al. 1993). Joint taxation systems are also expected to discourage female employment (e.g. Jepsen et al. 1997) (an assumption that has been questioned by Dingeldey 2001), while individual and progressive taxation systems have rendered it advantageous for women to work. In the case of Sweden in the late 1980s, for instance, it has been estimated that the husband can reduce his work hours by about 10 hours per week and the family maintain its available income if the wife takes on a job for about 5½ hours per week (OECD 1990: 163).

The twist of culture

Factors linked to or affecting female labour force participation have been studied extensively, and most social scientists agree that institutions and policies condition the behaviour of women. The role of culture has been less scrutinized, although culture is looming in the background in studies adhering to the regime approach developed by Esping-Andersen (1990). Thus, the regime approach argues that regimes are based on some basic principles (different perceptions of ideology, solidarity, equality, etc.), which help structure the role of major institutions in society (i.e. the family, market, and welfare state). Though Esping-Andersen originally argued that his three welfare regimes were based on three distinct and qualitatively different social logics, from the very beginning the regime approach was inscribed in linear and one-dimensional understandings (cf. continual high–medium–low degrees of de-commodification, etc.), with the state as the central structuring mechanism.

Some studies by feminist scholarship that adheres to the welfare regime approach, tend to be marked by linear state-centrism (very women-friendly or women-unfriendly welfare states), as well as essential understandings of the nature of women (e.g. women have an irresistible urge to be in the labour market, as this leads to emancipation) (e.g. Orloff 1993). Thus, the entry of women into the labour market is perceived as a more or less automatic or mechanical reaction to the opportunity structure of the welfare state (more or less women friendliness); that is, the idea is that women mechanically react to welfare policies in a specific, homogeneous, and predictable manner.

Seldom, however, is it possible to identify a straight line between policy programmes and the behaviour of women. For instance, it is widely assumed that the coverage of publicly funded childcare institutions structure the integration of women (mothers) into paid employment (Meyers and Heintze 1999; Siim 2000; Plantenga and Remery 2009). This argument disregards the oft-existing discrepancy between policy outputs and outcomes. In Finland (since 1996) and Germany (since the mid-2000s), all minor children have had a universal right to full-time care, but actual take-up rates (and the employment rates of mothers with small children) have been rather low (Kröger et al. 2003; Mätzke and Ostner 2010). The fact that mothers do not make use of public childcare and opt out of the labour market is therefore less because of any lack of childcare and more due to culturally embedded norms regarding social behaviour: In Finland and Germany, it is a widely held belief that good, loving mothers take care of their own children. Childcare institutions are thus established with different purposes in mind and have different effects in different countries, meaning that identical institutions are accorded different meanings and produce different outcomes in different cultural systems.

A corresponding collapse occurs when economists regard human behaviour as an informed and premeditated response to the economic utility of a given behaviour; that is, when actions in relation to, for example, female remuneration in the labour market are seen as a function of rational and utility-maximizing calculations. Economic incentives such as female wages, however, do not necessarily have the same effect on women’s labour force participation across social contexts. High female wages may actually limit the occupational opportunities for women in certain contexts while supporting female employment in others.

In Denmark, the struggle for equal wages largely unfolded in the 1960s. This battle was initiated by women under the slogan: ‘Equal pay for equal work!’, and was primarily based on egalitarian ideas regarding gender in all domains of social life. In contrast, very high remuneration for women was already introduced in Italy in the 1940s (Lorini 1975). In Italy, these demands were advanced by male-dominated trade unions, the purpose of which was to render it financially unfeasible (or less attractive) for employers to employ women. Thus, high wages for women had an exclusionary purpose in a male-breadwinner-dominated cultural model (Jensen 1996).

These examples clearly demonstrate that culture can increase or modify the impact of institutions and policies, as the manner in which women grasp and make use of institutions and policies is culturally embedded. Thus, policies do not automatically trigger action; rather, policies condition decision-making, including with respect to how women combine various formal and informal work activities in different social and cultural contexts (e.g. Bourdieu 1977; Giddens 1984). Nonetheless, with few exceptions (e.g. Pfau-Effinger 2000, 2004; Genre et al. 2005; Oorschot et al. 2008), culture is usually offered marginal treatment in explanatory frameworks of women’s labour force participation, which is quite remarkable since variations in cultures are associated with variations in employment practices (Pfau-Effinger 2004).

Cultural ideals, however, are not by definition homogeneous across a national society. Within a given national culture, divergent or even contradictory values and ideals may exist. For instance, women of different classes and localities (e.g. differences between Campania and Emilia Romagna) are differently disposed towards the employment system. As to class, C.C. Harris already pointed out in the 1980s that ‘we must be careful not to suppose, because middle-class women obtain satisfaction from their jobs apart from the financial rewards, that this is true of everyone else’ (Harris 1983: 74); that is, low-skilled and low-paid mothers in anti-social and unfulfilling jobs may find it more important – and more fulfilling – to be with their children as compared to highly educated, well-paid women.

Differences in the orientations and preferences of women towards the labour market have been conceptualized by Catherine Hakim (2002), who draws a distinction between (1) home-centred, (2) work-centred, and (3) adaptive preferences. This distinction may serve as a guideline for empirical studies. Theoretically, however, the challenge is to explain how and why differences in preferences emerge in a social, economic, and historical context. Thus, individual preferences or dispositions cannot be understood independently of the historical, social, and cultural conditions that constitute them in their specificity for a given society and a given moment in time.

Effects of labour force participation

Some feminist scholars, researchers supportive of the social investment strategy (e.g. Kvist 2015) as well as EU institutions share a common understanding that female labour market participation more or less automatically will lead to women’s financial independence, emancipation, and social cohesion. It should be emphasized, however, that the impact of female labour market participation largely depends on which forms, at which skill level, and with what work conditions and pay women are integrated into the labour market.

Labour markets have widely been described as being segregated vertically as well as horizontally and that women hold subordinate positions in the gendered division of labour (e.g. Cyba 1998; Blackburn et al. 2002; Rubery 2002; Charles 2003). Horizontal occupational gender segregation refers to the unequal distribution across branches and industries, while vertical segregation refers to status differentials within a given sector, and to how men dominate the top of the hierarchical ladder.

These disadvantages are often coupled with women working part time or in ‘atypical’ jobs. Part-time employment is often associated with low-quality jobs and low levels of social security (Hegewisch and Gornick 2011), but part-time work is not by definition precarious. Nonetheless, studies have shown that full-time work among women significantly lowers the risk of poverty (Halleröd et al. 2006). Even among women with full-time work, however, the proportion of the ‘working poor’ has substantially increased in recent years (Lohmann 2009), indicating that an increase in female labour supply does not necessarily enhance social cohesion nor lead to the true socio-economic empowerment of women (Lim undated).

New measures supportive to qualitative changes in the functioning of labour markets have been introduced on a Europe-wide scale over the last decade. To advance female labour force participation, the EU has adopted legislation against gender discrimination in pay and advancement chances; the Roadmap for equality between women and men (2006–2010), the European pact for gender equality (Council Conclusions March 2006), and the Community programme ‘PROGRESS’ all aim to promote comprehensive gender mainstreaming. Although initiatives have been multi-faceted, they have not always been implemented effectively (Boeri et al. 2005).

Choice of cities and data

The core data in the four articles in this special issue are survey data collected in 11 European cities belonging to 5 different welfare regimes; that is, social democratic, post-communist, liberal, conservative, and Mediterranean. The survey was also replicated in a city in an East Asian welfare regime. Due to differences in data quality only 9 of these 12 cities have been selected for in-depth analysis in this volume (in different combinations): Aalborg (Denmark), Jyväskylä (Finland), Brno (Czech Republic), Leeds (England), Hamburg (Germany), Nantes (France), Bologna (Italy), Terrassa (Spain), and Hsinchu (Taiwan). All nine cities exhibit rather high female employment rates (i.e. close to or above the EU 2010 60% employment target for women, with the exception of Brno) and are thus suited to illuminate the problematics of this special issue: What are the causes and effects of female employment in local contexts? In 2007, the employment rate among women was 70% in Aalborg, 69% in Jyväskylä, 45% in Brno, 69% in Leeds, 65% in Hamburg, 56% in Nantes, 66% in Bologna, 62% in Terrassa, and 55% in Hsinchu (women of working age).

The overall aim of the survey was threefold: (i) to collect empirical evidence about women’s culturally constructed dispositions and practices towards the labour market; (ii) to identify factors women perceive as facilitators or barriers for labour force participation; and (iii) to gain knowledge on the working and living conditions of working and non-working women. The questionnaire was subdivided into four parts: (1) background variables, including education, family and household composition, trust, social participation, and care obligations. The other three parts were adjusted to women in different positions vis-à-vis the labour market. (2) For the employed/self-employed, questions focused on job characteristics, the work–life balance, work history, experience with maternity leave, and enrolment in lifelong learning. (3) For the unemployed, questions focused on the duration of unemployment, reason(s) for unemployment, type of welfare benefits received, prospects for re-acquiring employment, and enrolment in activation measures. (4) For women outside the labour market, questions focused on the reason(s) for being outside the labour force, previous work experience, and prospects for re-entering the labour market.

Survey data were collected between October 2012 and March 2013. As oversampling was not possible in most of the cities, the collection of data in each city was based on a simple, random sampling technique among women aged 25–64 years residing in the city. The standardized survey questionnaire was conducted differently in the different cities. In Brno, data were collected by means of face-to-face interviews. In the rest of the cities, telephone interviews have been conducted. In Hamburg, Bologna, Terrassa, and Taiwan, only landline numbers have been called, while a combination of landline and mobile phones have been used in the other cities. In Aalborg, the telephone interviews were followed up by a postal survey of the non-respondents.

The aim of ending up with at least 800 valid questionnaires for analysis was achieved in all cities. The number of valid interviews (N) was 809 in Aalborg, 806 in Jyväskylä, 805 in Brno, 800 in Leeds, 814 in Hamburg, 802 in Nantes, 802 in Bologna, 802 in Terrassa, and 815 in Hsinchu. In total, 7255 valid interviews were conducted in 9 cities.

There were significant variations in our response rates. This is due to different sampling procedures and ways of calculating response rates (50% in Aalborg, 21% in Jyväskylä, 53% in Brno, 24% in Leeds, 21% in Hamburg, 5% in Nantes, 28% in Bologna, 16% in Terrassa, and 29% in Hsinchu). Although response rates in Nantes and Terrassa were rather low, it is noteworthy that an analysis has been made comparing the survey sample with administrative data in all of the cities regarding age, education, and employment status (we have used national register data in the few cases where city register data were not available). No major differences have been found. Interestingly, however, the data are skewed in a similar direction in all of the cities. Thus, data are skewed towards older age groups, women with tertiary education, and medium or higher income. Still, this does not pose a major problem, since data only to a minor degree will be used for descriptive purposes. Data will primarily be used in regression analyses where the effects of each category are estimated irrespective of the actual number in the category.

Core findings

The survey data are used to analyse the motives underlying the decisions made by women about entering the labour market in the policy, cultural, and economic context at the local level, and how labour force participation relates to women’s living conditions. What makes the research innovative is that the basic unit of analysis is the city. The strength of this approach is that it renders it relatively easy to keep track of women’s social, economic, and institutional environment. This is not possible if national data are used, since many factors, for example, the economic structure and welfare policies, within countries differ from one locality to another.

As to women’s integration into the labour market, the core finding is that family-cultural orientations associated with a specific economic structure (service economy) within an urban environment are the main determinants behind women’s work practices. Of course, individual (e.g. health, education, etc.) and family characteristics (number of children) are also important for women’s work practices. It has come as a surprise, however, that welfare policies are less important. Female employment rates may be high while coverage of day care institutions is low (e.g. in Bologna). Furthermore, the Hamburg experience shows that coverage in childcare may be high while take-up rates are low. An overall finding is that women with care responsibilities tend to reduce their attachment to the labour market – even in cities such as Hamburg and Jyväskylä where child care facilities are available for everybody.

Concerning the effects of women entering the labour market, the core finding is that housewifery is not associated with loss of citizenship. In terms of citizenship, hardly any differences between women outside the labour force and women gainfully employed can be found. Unemployed women, however, experience a loss of citizenship. Unemployed women suffer particularly from economic hardship. As to economic insecurity, however, the major dividing line is between women who cohabit or do not cohabit with a partner.

Notes on contributor

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.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 ‘Impact of Local Welfare Systems on Female Labour Force Participation and Social Cohesion’ (FLOWS) is a research project funded by the 7th Framework Programme of the EU, 2011–2014; www.flows-eu.eu

2 Open labour markets ‘[do] not deny participation to anyone who wishes to join’ (Weber 1978: 43).