<<

Democracy by Assocation: A Comparative Exploration of the Effects of Inequality and the State on Civic Engagement

by

Nanda S. Purandaré

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Political Science University of Toronto

© Copyright by Nanda S. Purandaré 2011

Democracy by Assocation: A Comparative Exploration of the Effects of Inequality and the State on Civic Engagement

Nanda S. Purandaré

Doctor of Philosophy

Department of Political Science University of Toronto

2011

Abstract

The dominant civic engagement literature has focused on the many positive outcomes that stem from leading an active associational life, linking it to lower crime rates, economic growth and a healthy democracy. However, it has been less effective at recognizing how much of a dependent variable civic engagement actually is, exploring what shapes it and how.

Yet, in light of its centrality to the democratic process and the benefits that accrue from strong, active communities, it is important to understand what shapes civic engagement to establish who is in a better position to participate and why. Drawing on personal interviews with single , policy analyses, and World Values and ISSP survey data, this dissertation explores how inequality and the state shape civic engagement. The findings underscore the impact of class- and status-based inequalities on civic engagement, focusing on women as a case study. Women’s dual roles as caregivers in the home and paid workers in the labour market contribute to the gender gap in participation. However, the presence of

ii

children is linked to higher levels of participation for women, and parent-, - and care- related groups are found to build trust, foster a sense of community, and act as a catalyst for civic involvement. The thesis also highlights the extent to which the state structures citizenship and participation, focusing on welfare regimes as case studies. It develops theories that test the effect of interventionism, egalitarianism and statism on the civic engagement levels of welfare regime-types. The findings suggest that while egalitarian policies may help reduce the impact of inequality on civic life, comprehensive social policies alone do not necessarily lead to more active . The way political authority is structured can have a deep impact on civic habits, and creating openings and opportunities for citizens to participate can inspire collective .

iii

Acknowledgments

I am most grateful to my dissertation committee for their unceasing guidance and support throughout the dissertation process – from the proposal stage to the defense and beyond. Thank you to my supervisors Jennifer Nedelsky and Jeffrey Kopstein for their commitment to the project in all of its incarnations, constructive feedback, as well as their unwaivering encouragement over the many years it takes to tackle an interesting research question. Thank you to Lorne Sossin, who inspired me to pursue graduate school in the first place, and for his feedback and counsel over the course of research and writing the thesis.

I am also extremely grateful to Peter Loewen and Livianna Tossutti for their invaluable insights and advice at the later stages of my research, and to Sylvia Bachevkin, for her helpful comments earlier on.

I gratefully acknowledge the generous funding for my research from the University of Toronto School of Graduate Studies’ Doctoral Thesis Completion and Travel Grants, the Special Matching Fund Dissertation Fellowship from the Munk Centre’s Joint Initiative in German and European Studies, and the Sir Val Duncan Travel Grant.

Last but not least, I wish to thank my friends and , especially mom, dad and Jason, for their patience and support through thick and thin.

iv

Table of Contents

Acknowledgments ...... iv

List of Tables……………………………………………………………………………...,,,,,,...vi

List of Figures………………………………………………………………..…………………viii

Glossary of Terms ...... xi

Introduction: Ingredients for Participatory Democracy ...... 1

PART I: INDIVIDUAL-LEVEL FACTORS ...... 26

Chapter 1: All Things Being Equal: Gender, Citizenship and the Authors of an Indifferent Democracy ...... 27

Chapter 2: Exploring the Roots of Gendered Civic Life ...... 68

Chapter 3: Count Me In: Taking Stock of the Networking Habits of Single Mothers .... 109

Appendix: Interview Schedule ...... 151

PART II: COUNTRY-LEVEL FACTORS ...... 155

Chapter 4 The Impact of Welfare Worlds on Citizenship ...... 156

Appendix: Comparative Social Policies and Outcomes ...... 202

Chapter 5: A Hinderance or a Catalyst? Theorizing the Role of the State in Civic Engagement ...... 224

Chapter 6 Interventionism, Egalitarianism, and Statism: The Impact of Three State Dimensions on Associational Involvement ...... 278

Conclusion A Recipe for Participatory Democracy ...... 330

References ...... 351

v

List of Tables

Table 2.1 Predicting the Causes of Gender Differences in Civic Engagement: OLS Regression Results…………………………………………...………p.105

Table 4.1 A General Outline of Three Welfare Regime-Types……………...…...... p.203

Table 4a.1 Provincial Demographic Profiles…………...... p.206

Table 4a.2 Overview of Provincial Social Assistance Histories...... p.207

Table 4a.3 Provincial Family-Related Supplements, Subsidies and Allowances...... p.208

Table 4a.4 Provincial Leave Policies…………………...... p.209

Table 4a.5 Provincial Child Maintenance Enforcement Programs……………………………………………...... p.210

Table 4a.6 Information on Early Childhood Education and Care by Province……………………………………...... p.211 . Table 4a.7 Provincial Social Assistance, Rance and Adequacy of Benefits, and Employability Programs………………...... p.212

Table 4a.8 The Status of Women by Province: Poverty, Work and Political Representation…………………..…...... p.213

Table 4a.9 The Status of Women by Province: Violence Against Women, Shelters and Victims Services……...... p.214

Table 4b.1 Family and Childrearing Leave Policies in Canada, France and Sweden………………………...... p.215

Table 4b.2 Early Learning and Child Early Learning and Childcare Policies in Canada, France and Sweden...... p.216

Table 4b.3 Family or Child Benefits and Services in Canada, France and Sweden…………………...... p.218

Table 4b.4 Housing and Housing Policies in Canada, France and Sweden………………………………………...... p.219

vi

Table 4b.5 Income Security Policies in Canada, France and Sweden……………………………………...... p.221

Table 4b.6 Comparing Welfare Regimes: The Status of Women in Canada, France and Sweden...... p.222

Table 5.1 Predicting Civic Engagement by Religious Denomination: OLS Regression Results………………...... p.231

Table 5.2 The Cross-National Voluntary Sector, by Field…………………………………………………………...... p.254

Table 5.3 Sources of Nonprofit Sector Revenue………...... p.256

Table 5.4 Predicting the Effect of Social and Political Trust: OLS and Logistic Regressions………………………………...... p.271

Table 5.5 Dimensions of Variation Between Selective and Universal Welfare Policy...... p.274

Table 6.1 Hypothesizing the Effects of Interventionism on Civic Engagement Across Welfare Regime-Types...... p.282

Table 6.2 Gross, Net and Share of Private Social Expenditure Across Welfare Regimes…………………...... p.284 Table 6.3 Hypothesizing the Effects of Egalitarianism on Civic Engagement Across Welfare Regime-Types...... p.288

Table 6.4 Income Inequality and Poverty Rates by Regime-Type...... p.289

Table 6.5 Hypothesizing the Effects of Statism on Civic Engagement Across Welfare Regime-Types...... p.293

Table 6.6 Hypothesizing the Effects of Three State Dimensions on the Civic Engagement of Welfare Regime-Types...... p.300

Table 6.7 OLS Regression: Predicting the Effects of Welfare Regime Dimensions on the Civic Engagement……………………..……...... p.315

Table 6.8 Hierarchical Linear Models Evaluating Country-Level Factors as Predictors of Civic Engagement, with Cross-Level Interaction...... p.320

vii

List of Figures

Figure 1.1 Gender Differences in Civic Engagement……...... p.29

Figure 1.2 Gender Differences in Voluntary Membership Type…………...... p.33

Figure 1.3 Gender Differences in Voluntary Activity by Association Type……………………………...... p.35

Figure 1.4 Membership in Voluntary Organizations by Gender, Controlling for Religious Attendance...... p.37

Figure 1.5 Membership and Activity Volume by Gender………...... p.39

Figure 1.6 Political Action, by Gender…………...... p.40

Figure 2.1 Gender-Differences in Socio-Economic Status……………...... p.70

Figure 2.2 Voluntary Membership by Gender, Controlling for Education Level …………………………………………………...... p.73

Figure 2.3 Voluntary Membership by Gender, Controlling for Socio-Economic Status…………………………………...... p.75

Figure 2.4 Voluntary Membership by Gender, Controlling for Ethnicity……………………………………………...... p.77

Figure 2.5 Education Level for Men and Women, Controlling for Race………………………………...... p.80

Figure 2.6 Income Level for Men and Women, Controlling for Race………………………………...... p.81

Figure 2.7 Voluntary Membership by Gender, Controlling for Family Status……………………………...... p.85

Figure 2.8 Political Orientation by Gender……...... p.86

Figure 2.9 Voluntary Membership by Gender, Controlling for Interest in Politics………………………...... p.87

Figure 2.10 Voluntary Membership by Gender, Controlling Following Politics in the News……………...... p.88

viii

Figure 2.11 Frequent Networking with Kin and …………………………………………...... p.90

Figure 2.12 Membership in Voluntary Organizations by Gender, Controlling for Frequent Visits with Kin and Extended family...... p.91

Figure 2.13 The Effect of Involvement in Politically-Inclined Groups on Social Trust……………...... p.95

Figure 2.14 The Effect of Political Action on Social Trust……………...... p.96

Figure 3.1 The Voluntary Membership Habits of 60 Single Mothers in Canada, France and Sweden…...... p.117

Figure 3.2 The Informal Networks of 60 Single Mothers in Canada, France and Sweden ………………...... p.124

Figure 3.3 Social Trust Levels Among 60 Single Mothers in Canada, France and Sweden………………...... p.139

Figure 5.1 Cross-National Civic Engagement Levels…...... p.227

Figure 5.2 Cross-national Level of Catholics and Protestants…...... p.230

Figure 5.3 Cross-National Religious Attendance Levels………...... p.233

Figure 5.4 Cross-National Immigration as a Percent of the Population………………………………………………………...... p.236

Figure 5.5 Cross-National Levels of Trust in ‘Different’ Others……………………………………………...... p.239

Figure 5.6 Cross-National Level of Family Ties……………...... p.244

Figure 5.7 Annual Registration of New Associations and Laws Extending the French Welfare State…...... p.262

Figure 6.1 Overall Civic Engagement Levels by Welfare Regime-Type...... p.301

Figure 6.2 Five Civic Engagement Indicators by Welfare Regime-Type………………………………………………………………...... p.303

Figure 6.3 Membership and Activity Volume, by Welfare Regime-Type………………………………………………...... p.304

ix

Figure 6.4 Organizational Involvement by Group-Type, Across Welfare Regimes…………………………………...... p.305

Figure 6.5 Unpaid Work by Group-Type, Across Welfare Regimes ……………………………………………………...... p.307

Figure 6.6 Membership in Religious Organizations and Labour Unions, by Welfare Regime-Type………………………...... p.308

Figure 6.7 Measuring the Impact of Unions and Religious Organizations on Membership and Unpaid Work Levels Across Welfare Regimes...... p.309

Figure 6.8 Informal Social Ties by Network-Type, Across Welfare Regimes …………………………………………...... p.311

Figure 6.9 Political Action by Type of Activity, Across Welfare Regimes …………………………………………...... p.313

x

Glossary of Terms

Bonding Network: In civic engagement literature, Corporate-conservative regime-type: One of three denotes social ties between people that are mostly alike welfare regime-types, according to Esping-Andersen’s with similar values and worldviews. This type of typology, represented by Western European countries network alone is not considered desirable for building such as Germany and France. It has a Catholic and generalized trust and fostering further networking. authoritarian legacy, and is known to reinforce social Examples of bonding networks include those with status hierarchies, for instance via occupationally- people from the same culture, race or family. based insurance schemes. It is a de-commodifying, socially stratifying and familializing regime-type.

Bridging Network: In civic engagement literature, Crowding-out theory: Adopted by the dominant civic denotes social ties across divergent types of people or engagement literature, it posits that strong, large groups. This type of network is considered favourable interventionist states dampen civic involvement by for building generalized trust and making new contacts making citizens reliant on state provisions rather than to encourage further networking. Examples of this form on self-help via social networks. of network include those with a mix of people from different races or ethnicity, age group, income level and gender.

Citizenship: Membership status and belonging to the De-commodification : In relation to welfare regimes, community, awarded through rights and entitlements it refers to the extent to which citizens are free from that are not available to non-members. Scholars identify market dependency. It means that people are able to two dimensions: a) status stemming from rights and maintain a livelihood without reliance on the market entitlements of the state and b) practice involving duties since the state offers a range of generous benefits. and expectations of citizens in return for their status. The According to Esping-Andersen, corporate-conservative latter can range from paying taxing and mandatory and social democratic regimes de-commodify. military service, to civic participation.

Civic engagement theory: Suggests formal and De- : According to Esping-Andersen’s informal networks and norms of generalized trust lead to regime typology, it is when the households are relieved individual benefits such as job contacts and developing of some of their family burden via social provisions. new skills, and community benefits such as economic Nordic countries are considered de-familializing. growth, lower crime rates and responsive, accountable and effective institutions.

Childrearing leave: A lengthy (typically lasting more Egalitarianism: A theory about the role of the state in than one year per child) which is shaping civic engagement, developed in this thesis characterized by a lower, flat-rate or means-tested based on the literature that views the state as a catalyst benefit level. It is considered a familialist approach for civic engagement. It suggests that egalitarian which reinforces traditional gender roles because such welfare states such as in Nordic countries, as opposed lengthy leaves are said to reduce women’s incentives to to non-egalitarian ones, foster civic engagement by return to work by weakening their skills, limiting career reducing social stratification and build general trust. paths and opportunities, and reducing earning levels.

Commodification : In relation to welfare regimes, Familialism : According to Esping-Andersen’s regime refers to the extent to which citizens are free from typology, it is when households are assigned the brunt market dependency. It means that people find it difficult of family burden via social policies. Corporate- to maintain a livelihood without reliance on the market conservative Western European countries like France since the state offers few benefits. According to Esping- and German are considered familialist. However, in Andersen, liberal regimes commodify this study, countries which provide few social provisions also qualify as familialist – assigning the brunt of the family burden implicitly – as is the case with Anglo-Saxon countries.

xi

Formal network : In civic engagement literature, a Maternity leave: Leave available immediately prior to type of network that requires membership dues or cards and following the birth and, in some cases , of and which has a formal organizational governance an infant that is only open to mothers. This type of structure – such as unions. leave can be used to offer mothers early leave if experiencing complications. They tend to be short and in some cases unpaid. They typically are followed by a lengthier parental leave period.

Generalized trust: Level of trust in other individuals Parental leave: Leave available following the birth that we do not know, as opposed to particularized trust and in some cases, adoption, of an infant that is open to which is the level of trust between individuals in close either or , or a combination of both. In ties such as familial or communal bonds. It is a key most cases, they are paid leaves. ingredient in civic engagement, produced by social ties and, in turn, responsible for reproducing social ties.

Gross social public expenditure: Before-tax social Paternity leave: Leave available following the birth spending. It does not include tax schemes or private and in some cases, adoption, of an infant that is only spending. This is a measure commonly used to open to . They tend to be short and are often determine the degree of state interventionism. unpaid leaves.

Gini coefficient: A standard statistic frequently used to Poverty rate: The score for any post-tax and -transfer measure inequality in income distribution, which varies income that falls below 50% of the median household from 0 to 1, with 0 representing perfect equality, and 1 disposable income. It is a useful measure to determine representing perfect inequality. It is a useful measure to state egalitarianism. determine state egalitarianism.

Homophily: With regards to social networks, refers to Pronatalism: The implementation of policies designed associations with individuals that are similar to us (i.e. to influence the reproductive choices of – to birds of a feather flock ). increase the national birth rate and encourage large families. It is associated with welfare states which reinforce traditional gender roles.

Informal network: In civic engagement literature, a Net total social spending: After-tax social spending, type of network that does not involve membership dues inclusive of tax schemes and private spending. This is a or cards and has no formal organizational governance measure commonly used to determine the degree of structure – such as a regular card game with friends. state interventionism.

Interventionism: A theory developed in this thesis Non-statism: See statism. about the role of the state in shaping civic engagement, based on the literature which views the state as a hindrance to civic life.

Liberal regime: One of three welfare regime-types, Non-interventionism: See interventionism. according to Esping-Andersen’s typology, represented by Anglo-Saxon countries like Canada and the U.S. It has a liberal legacy, favouring private market solutions and small . It is commodifying, socially stratifying and is implicitly familializing.

xii

Non-egalitarianism : See egalitarianism. Social stratification: In relation to welfare regimes, it refers to the way states rank citizens via social policy - whether it emphasizes or maintains social cleavages, or strives to build between citizens. According to Esping-Andersen, liberal and corporate-conservative regimes are socially stratifying.

Regime: See welfare regime-type Statism: A theory developed in this thesis about the role of the state in shaping civic engagement, based on the work of Jepperson. It suggests that, as opposed to non-statist countries, states in which political authority has historically been entrenched in the central state and disconnected from the people, such as in Western European Germany and France, encourage civic engagement.

Share of private in total social expenditure spending: Welfare regimes: Families of welfare states that share The amount of private spending out of the total of public similar practices and logics over time. This underlying and private social expenditures. This is a measure logic is reinforced by the social, political and economic commonly used to determine the degree of state arrangements which can be traced back to a country’s interventionism. historical legacy, and shape the design of their systems of social provisions. Esping-Andersen’s dominant typology identifies three regime-types: liberal, corporate-conservative and social democratic.

Subsidiarity: A mode of social organization which Welfare state: consists of: a) the rights and dictates that higher social units (the state) should not entitlements which derive from social policies, but intervene in the affairs of lower social units (individual, often overlap into other policy areas such as family, tax family) unless the latter has demonstrated incapacity. and labour policy; b) the manner in which resources are This is a distinctively corporate-conservative trait, redistributed, for instance, eligibility criteria, and passed down from its Catholic legacy. duration and amounts of benefits; and c) the outcome in terms of how it works to order such as whether it targets some or treats all citizens equally. Its strategies and discourses are shaped by an underlying logic that works to direct policy implementation, design and outcomes.

xiii

1

Introduction Ingredients for Participatory Democracy

This study identifies and explores key factors which shape civic engagement. It demonstrates that class- and status-based inequalities at the individual-level, and the organization of political authority and the welfare state at the country-level, have a significant impact on associational life. Civic engagement theory posits that networks of sociability build generalized trust which together, over time, lead to a virtuous circle – a self-reinforcing system where networks and trust are reproduced. While a sizeable literature has developed, chronicling civic activity outcomes, little is known about the variables that influence civic activity in the first place. Over time, civic culture and social capital studies have accumulated to form what can be referred to as a civic engagement literature. One of the great contributions of this literature has been the production of a vast amount of theory and empirical evidence demonstrating the value that civic norms and networks can bring to both individuals and communities at large. In highlighting the centrality of civic engagement to democracy, the dominant civic engagement literature has shown that networks of sociability allow individuals to develop civic skills and political knowledge; provide an important outlet for citizens to express their views to decision-makers; promote favourable civic values like trust and solidarity; and facilitate the solution of collective action problems.

Even apart from its instrumental outcomes, civic engagement is also understood as an end in itself; it is the expression of individuality and human agency, and the exercise of self- governance and liberty. Civic engagement is a useful concept because it suggests that social networks are vital to democratic governance at a time when scholars are pointing to a decline of deference to, and a rampant disdain for, formal political institutions and

2

politicians. Moreover, leading an active associational life is favourable because just as it can provide goods and services, improve quality of life, trigger social innovation, and encourage effective governance, low levels of participation and trust can lead to a myriad of social and political ills, including an increase in crime and high school dropout rates, and a rise in corrupt, unresponsive institutions.

Hence, the dominant civic engagement literature has been effective at outlining what associational life can do, treating it as an independent variable. However, it has been less effective at recognizing how much of a dependent variable civic engagement actually is, exploring what shapes it, and how. In light of its centrality to the democratic process, and all the benefits that can accrue from strong, active communities, it is important to understand how civic engagement works to establish who is in a better position to participate and why, and what can be done to generate civic norms and behaviour generally.

This study is fundamentally about civic involvement in Western advanced democracies. At its heart, it explores the ingredients for participatory democracy. It aims to gain a better understanding of the types of policies and institutions required to build capacity and foster civically active communities. Specifically, the study focuses on two areas for which there is a paucity of research: it explores how inequality and the state affect associational life. As such, this study underlines the importance of considering participation from two levels of analyses: the individual- and country-levels. At the individual-level, social and economic context have a tremendous influence on civic networks and norms that are seldom accounted for. Despite the relationship between participation and democracy, the dominant civic engagement literature is blind to the power dynamics which characterize societies and structure associational life because it rests on a limited view of citizenship.

3

Citizenship provides membership status and belonging to the community, awarded through rights and entitlements that are not available to non-members. Prevailing civic engagement studies have adopted a view of citizenship that is grounded in formal equality, and therefore presumes that all individuals are the same and are being treated in the same way. In this idealized world, unfettered by the constraints of socio-economic status (SES), participation is understood to be the result of sheer free will. According to this perspective, civic engagement is merely the behaviour of individuals seeking to optimize benefits by combining with others to achieve something they might otherwise not have accomplished on their own, while reducing the risks involved in collective action by entering into a reciprocal relationship that will provide a pay-back in the future.

However, in practice, many of those who withdraw from associational life do so not because they lack the will, but because social and economic conditions limit their ability or opportunity to network. In order to address these factors, this study turns to a feminist citizenship perspective, grounded in substantive equality, premised on the understanding that individuals should be treated differently to make them equals. Drawing on a citizenship model that is grounded in substantive equality can account for the social and economic conditions of individuals which translate into class- and status-based inequalities – all of which are overlooked by dominant civic engagement studies. The importance of paying attention to difference and SES is illustrated in this study by the case of women. Women’s dual roles in the family as primary caregivers, and in the labour market as paid workers, is found to place them at a continued socio-economic disadvantage, and constrains their ability and opportunity to participate. Competing family and work duties structure men and women’s lives in different ways, leading to a gender gap in participation, with women being

4

generally less involved than their male counterparts, while also having certain forms of networks they belong to disregarded as ‘civic’ activity because it relates to or children. Common explanations for the enduring gender gap in participation, and inextricably linked to women’s citizenship, are considered here, including having: inadequate resources such as time, money or skills required for participation; fewer opportunities to be recruited into diverse groups due to a limited pool of contacts; and low levels of interest in politics and efficacy which may lead individuals to turn inwards and become isolated rather than connect with fellow citizens. In addition, the dominant civic engagement literature falsely presumes that political and civic life stop at the front door of the family home; that these exist in a space that is cut-off and disconnected from the family and the so-called private sphere. Focusing on women’s civic activity, and broadening the definition of ‘civic network’ to include parent-, child- and care-related groups, this study reveals the value of such networks to civic life, building a sense of community, norms of trust and generating further social ties.

While class- and status-based inequalities have a definite impact on associational life, it would be misleading to suggest that individual-level factors alone shape civic engagement. This study shows that citizenship is also greatly influenced by country-level factors, namely the way political authority is structured, as well as via the policies and provisions of the welfare state. Prevailing country-level theories seeking to explain cross- national variation in participation are reviewed. These theories posit that religious denomination (Catholicism versus Protestantism); the strength of family ties; and the degree of racial or ethnic diversity have an important effect on civic activity. However, this study contends that such approaches are not persuasive because they fail to account for the

5

important influence of the state over civic engagement. Such prevailing theories require that we overlook: the transfer of power and responsibility for social welfare from church to state; the role of the state in reinforcing racial tensions through ineffective social and immigration policies, and; the gender implications of having the family as the main source of welfare provision. Ultimately, it is argued that it would be impossible to fully comprehend democratic participation without also considering one of the key players in the project of democracy: the state.

This study uncovers two schools of thought regarding the role of the state in shaping civic engagement. The first, adopted by the dominant civic engagement literature, views the state as a hindrance to associational life, encouraging idleness and isolation by creating a citizenry that is reliant on generous state provisions rather than self-help via networking.

The second school of thought views the state as a catalyst for civic engagement, with certain political systems and welfare states fostering participation and trust. However, not only is this literature scarce and varied, but the findings contradict one another, and tend to be incomplete, focusing, for instance, on only one aspect of civic engagement (trust or participation) and one dimension of the state (egalitarian policies or state size or strength).

In short, there is no adequate theory of the role of the state that can explain variation in levels of civic engagement in a comparative context. Therefore, this study develops three theories of state based on existing these approaches, and tests them to determine which best explains the role of the state in shaping networking and trust. Each theory, including interventionism, egalitarianism and statism, posits how particular dimensions of the state might lead to regime-specific participation outcomes. The interventionism theory suggests that non-interventionist states, which avoid meddling in the market, family and the everyday

6

lives of citizens, will generate civic activity because the absence of generous state-run social programs and benefits will drive citizens to fend for themselves through mutual aid networks. The egalitarian theory suggests that states actively working to implement policies which reduce inequalities and social stratification will foster social cohesion through norms of trust, and encourage networking across all groups in society. Finally, the statism theory suggests that non-statist countries - in which political power has historically been legitimized by and concentrated in the citizenry (the people), and where there are few boundaries between the state and society - will generate higher levels of participation that statist countries where political authority is disconnected from the people and centralized.

This study holds that in order to adequately study civic engagement at the country- level, it is crucial to invoke Esping-Andersen’s (1990, 1999) welfare regime typology. A welfare regime (or ‘welfare world’) is a group of countries which behave in a similar way over time. They share an underlying logic which shapes the design of their systems of social provisions. This underlying logic is reinforced by the social, political and economic arrangements which can be traced back to a country’s historical legacy. Esping-Andersen identifies three welfare regimes: the liberal regime, consisting of Anglo-Saxon countries; the corporate-conservative regime, consisting of Western European countries; and the social democratic regime, consisting of Nordic countries. This approach is vital because it a) can be used to select country-cases in a systematic and logical way; b) can link citizenship status and SES and its impact on participation at the individual-level to the country-level, via the welfare state; c) it is more comprehensive than other comparative models, such as that of

Titmuss or Rothstein; d) it clearly highlights how different groups of states mould citizenship by reinforcing the citizen’s relationship to the state, the family, the market and

7

the community; e) unlike common single indicators used to measure welfare state inputs and outputs, regimes offer an explanation as to the content of welfare states, acknowledging the very different designs of social provisions and their varied outcomes, as well as the role states play in promoting these, and why. Esping-Andersen’s welfare regime typology is used to select countries for study, and test the three theories of state on associational life, each linking particular regime-based civic engagement outcomes to different dimensions of welfare states.

This study is path-breaking in several ways. First, it considers civic engagement as a dependent variable in a cross-national perspective – something that has rarely been accomplished. The lack of such research is surprising considering the vast amount of research on civic culture and social capital, and considering the great variation in levels of networking and trust that exist across countries. Second, it considers civic engagement from multiple levels of analysis, while also considering cross-level interaction. Such an approach is necessary if we are to fully grasp civic life, since the findings of this study show that both inequality and the state have significant effects on participation and trust. Third, it develops three theories of the state, building on the small set of varied studies that exists on comparative civic engagement. In light of the important impact of the state on associational life, creating these theories is not only valuable because it addresses the sheer lack of effective approaches, but also because it considers various state dimensions and can be applied across regime-types. Fourth, it not only underlines the need to consider citizenship in the study of associational life – membership and belonging in the community awarded through rights and entitlements – something that is scarcely done, but specifies the appropriate approach required in order to gain a more complete understanding of civic life.

8

This is particularly helpful in the pursuit to gain a more complete understanding of participation because a feminist citizenship approach, grounded in substantive equality, makes it impossible to overlook how difference and social and economic inequalities shape civic engagement.

Fifth, this study contributes to the existing civic engagement literature because it takes a broad approach to participation, not only including alternative modes of participation such as protesting and boycotting, but also including parent-, child- and care-related networks that are typically excluded from surveys and analyses. Adopting a methodology which includes measures for a variety of network-types is key to gaining a better understanding of the civic engagement of disadvantaged groups as these tend to favour unconventional forms of participation. In the case of women, including care-related groups is particularly important because these types of social ties are attached to women’s roles as primary caregivers – historically associated with the private sphere of the home and family.

As such, any participation taking place in the home or related to care work has been, and continues to be dismissed and disqualified as a form of civic or political participation. Not only is accounting for such networks important to shedding light on the way care work, combined with paid work, shapes women’s civic lives, but also to underscore that participation can take place in the home, and care-related networks do constitute a form of civic activity, building trust, a sense of community and fostering other social ties. Finally, this study adds to the existing civic engagement literature because it pays attention not only to how the state shapes civic engagement, but does so from a comprehensive regime-based perspective, linking the state to other institutions such as the family and market. A welfare regime analysis is valuable because it can help us see how different regime-based

9

institutional logics (e.g. promoting the private market, the state or equality) lead to specific policy designs which promote a certain citizenship ideal – such as reinforcing traditional gender roles. Moreover, such country-level factors have important ramifications at the individual-level, since social and economic conditions, which we know are shaped by the state, affect participation. Before providing an outline of the study, the following section offers a review of the civic engagement literature, providing a brief outline of some key works. The Dominant Civic Engagement Literature This study is informed by what is referred to as the ‘dominant civic engagement literature’ which includes the civic culture and social capital literatures. While their methodology and theories about causal mechanisms may vary, dominant civic engagement scholars agree that an actively engaged , as well as norms and values such as trust and reciprocity, are important to democratic strength and stability. According to the prevailing understanding of civic engagement, when individuals network with one another on a regular basis, they will grow to be more trusting of others – what is known as generalized trust (as opposed to interpersonal trust, referring to trusting individuals whom we already know); this will lead them to be more likely to enter into other networks, producing a virtuous circle where civic engagement is produced and reproduced.

Conversely, the experience of reciprocal contact and trust in others that is missing from inactive communities leads to a vicious cycle where isolation and distrust become a major challenge to future collective action. Social networks are considered important, even if they are apolitical. Both formal networks, such as those that require membership dues, have established an organizational structure, and tend to develop weak ties, as well as informal networks which tend to produce strong ties, are included in measures of civic engagement.

10

While certain organizations, such as the KKK, bring negative returns because they breed intolerance of others, tend to be highly exclusionary, and generally threaten social cohesion, overall, voluntary associations are considered beneficial to democracy because they bring individuals together and foster values such as trust and cooperation, required to facilitate other forms of civic and political engagement. They allow members to share information, discuss politics and develop political awareness, and learn civic skills, such as public speaking, organizing meetings, and writing letters (Granovetter, 1973; La Due Lake and

Huckfeldt, 1998; Erickson and Nosanchuck, 1990; Verba, Scholzman and Brady, 1995;

Putnam, 1993, 1995, 2000; McClurg, 2003; Zukin, Keeter, Andolina, Jenkins and Delli

Carpini, 2006).

Furthermore, norms of trust and cooperation are considered an important component of civic engagement because they are said to bond community members together, facilitating social exchanges by minimizing any problems of uncertainty or risk involved when interacting with others. The tendency of approaching trust as both a component and outcome of civic engagement has led to a debate about the exact relationship between civic networks and norms which has been aptly outlined elsewhere, and is not pursued further here (Adam and Rončević, 2003; Uslaner, 1999; Brehm and Rahn, 1997; Hooghe and Stolle, 2003).

Civic engagement studies often turn to Tocqueville’s Democracy in America (2000), which praises the thriving and diverse associational life in nineteenth century America.

Tocqueville is amazed to find that, compared to his native France, “Americans of all ages, all stations of life, and all types of dispositions…forever forming associations.”

(Tocqueville, 2000, p.513). He is particularly impressed with the amount and diversity of organizations in which people are involved which have no political objective, and how these

11

are understood as a powerful means of action (Tocqueville, 2000, p.513-4). For Tocqueville, networks of sociability act as an important social glue, pulling people into cooperative relationships which build important social norms that are invaluable to strong communities:

“[f]eelings and ideas are renewed, the heart enlarged, and the understanding developed only by the reciprocal action of men one upon another.” (Tocqueville, 2000, p.515) Thus, he suggests that a vibrant civil society is key to a healthy democracy because it builds civic skills and averts one of the potential dangers of freedom - social isolation and apathy.

Almond and Verba include measures for norms of trust and networking to account for democratic stability in The Civic Culture (Almond and Verba, 1965), considered “the first systematic attempt to explain democratic outcomes with cultural variables” (Laitkin,

1995, p.168). Drawing on public opinion data, their study seeks to explain the persistence of democracies by considering the attitudes of citizens in Mexico, Germany, Britain, Italy and the United States. They examine the political cultures of each nation by evaluating the orientations of citizens towards political institutions and outcomes, concluding that Britain and the United States demonstrate the most favourable political culture to ensure democratic stability. However, The Civic Culture (Almond and Verba, 1965) was critiqued by scholars for its methodological approach (for example, the use of public opinion surveys; the sample size used for the five countries under study; an over-representation of urban respondents, particularly in the Mexican case), its conceptualization of political culture and democracy

(for instance, the exact causal order of civic culture and democracy; a vague definition of political culture; and a strong liberal democratic bias), and its limited explanatory power

(Laitkin, 1995; Fuchs, 2007; Sabetti, 2007, p.343). Moreover, while the question of democratic persistence was compelling and timely in light of two world wars and the Cold

12

War, the collapse of communism called for the type of shift in emphasis that social capital delivered (Fuchs, 2007, p.172).

The dominant approach to social capital also consists of measures of participation in voluntary associations as well as norms such as trusts and reciprocity, but the emphasis is no longer on explaining what makes democracies last, but what makes democracies work

(Fuchs, 2007, p. 170). Social capital is different from much of the civic culture literature in that it conceptualizes networks as resources which can benefit individuals and the wider community. Countless scholars have relied on social capital over time, but three dominant social capital prevail: that of Bourdieu and Coleman, who conceive of social capital at the individual-, and that of Putnam, whose analyses take place at the regional- and national-levels of analysis. Bourdieu (Bourdieu, 1986, p.252) aims to uncover the ways in which agents appeal to social capital to advance their social standing within society.1 The value of his analyses lies in his consideration of power, and his recognition that the inequalities that persist in the civil society are reflected in the production of social capital (Bourdieu, 1986, p250). Yet his formulation has failed to generate much social capital research, at least in part because of its methodological fuzziness (Wall,

Ferrazzi and Schryer, 1998, p.307) .

While Bourdieu is concerned with social capital as a means to attaining economic capital within the context of social stratification, Coleman (Coleman, 1998, p.S100) turns to social capital to uncover how norms and sanctions in family and community associations

1 Bourdieu first introduced the notion of social capital in his book, Distinction, first published in French in 1979. The obligations and mutual recognition, Bourdieu contends, that accrue from these social connections arises from both material (e.g. money) and symbolic (e.g. prestige) exchanges by agents in social networks. In his analysis, Bourdieu identifies three forms of capital: a) economic capital which can instantly be transformed into money and institutionalized into property rights; b) cultural capital which may be transformed into economic capital and institutionalized into educational qualifications; and c) social capital which consists of social obligations that can be transformed into economic capital, and institutionalized into title of nobility.

13

might foster or constrain levels of human capital held by individuals. In examining the impact of social capital on human capital, he correlates the rate of high school dropouts amongst sophomores with low levels of social capital in families and communities, uncovering a positive relationship between social capital and human capital. Coleman has been critiqued for disconnecting social capital from any social and historical contexts, working with “concepts such as power and control, quite apart from interests, externalities, public goods, etc.” (Fine, 2001, p.75). Nevertheless, he is applauded for having elaborated on social capital’s mechanisms and components, and charting the course for the study of social capital at the macro level of analysis. As a result of his development of social capital, his work has inspired a slew of empirical studies, including Putnam’s.

Putnam’s method of analysis (Putnam, 1993, 1995, 2000) lies at the heart of the most recent social capital literature, and has renewed interest in this approach. His approach has cachet not only because he prescribes a remedy to alleviate social ills, but also because he suggests a methodology for measuring social capital at the regional- and national-levels of analysis. He defines social capital as “features of social life – networks, norms and trust – that enable participants to act together more effectively to pursue shared objectives.”

(Putnam, 1996). Social capital, according to Putnam, has the capacity to explain how membership in social networks can enhanced democracy. In Making Democracy Work

(1993), he links a of great civic engagement and trust in Northern Italy to better institutional performance in the region compared to Southern Italy, which he finds suffers from lower levels of trust, withdrawal from the community, and less responsive government.

In Bowling Alone (Putnam, 1995, 1996, 2000), Putnam studies American civic life over time and uncovers an alarming decline in social capital which he links to pressures of time and

14

money, sub-urbanization, television viewing, and generational changes. Putnam’s formulation of social capital is valuable because it offers a working methodological model for measuring social capital. While his approach has met with much critique (Paxton, 1999;

Wills, 2002; Skocpol, 1996; Winter, 2000; Foley and Edwards, 1999; Woolcock, 1998;

Portes, 1998; Portes and Landolt, 1996; Sabetti, 1996), Putnam’s work nevertheless continues to inspire a sizeable amount of research interested in further exploring the link between social capital and democracy. There are several reasons for the continued interest in

Putnam’s model, despite its critiques; one aspect of social capital that is both compelling and controversial is the characterization of social networks as resources which individuals and communities can tap into. Moreover, compared to The Civic Culture, social capital understands social networks in a relational way, as an aspect of “…human interaction as people confront concrete instances of coming to terms with their own environment. Almond and Verba’s formulation denies, or skips over, the processes by which people develop over time and through interactions the capacities proper for citizenship in particular communities, and in relation to the larger world” (Sabetti, 1996, p.352-3).

In light of this brief review of the dominant literature, civic engagement is defined and measured here as membership and participation in formal (i.e. belonging to a union) and informal (i.e. regular card night with friends) social networks which may be political or apolitical in nature. While it is considered and analyzed, social trust is approached more tentatively because of the limits that cross-national surveys present to measuring this norm.

Not only do the respondents of personal interviews conducted for this study seem confused by survey questions on trust, but large prevalent datasets themselves either do not include trust questions in every wave, focus on a single indicator for trust which throws into

15

question the reliability of results, or change trust questions over time, making surveys and waves incompatible with one another.

This study contributes to research on civic engagement by addressing two major weaknesses in the dominant literature. First, it proposes a more thorough evaluation of how inequalities affect participation. The civic culture and social capital literatures have been critiqued for their failure to adequately tie social, economic and political contexts to civic activity (Pateman, 1989, p.174; Lowndes, 2004, 2006; Bruegel, 2005). A lack of attention to gender dynamics has translated into a literature with an inherent male bias (Lowndes, 2006, p.215). Both Almond and Verba, as well as Putnam, have been challenged for their lack of attention to class- and status-based differences on civic engagement. The Civic Culture, for instance, is critiqued for its focus on the individual as agent, independent of economic and social contexts which otherwise structure participation. Pateman challenges Almond and

Verba on several counts, including their assumption that “…systematically structured inequalities appear as individual psychological and personal attributes that happen to be distributed in a particular way” (Pateman, 1989, p.174; Faulks, 2000, p.110). Thus, while

Britain and the US are deemed to represent the most ‘civic cultures,’ Pateman shows that such a conclusion can only be reached through a biased reading of the findings; one which is blind to the fact that the very values and behaviours they attribute to ‘the civic culture’ are actually divided along socioeconomic lines (Topper, 2005, p.199). Putnam is also accused of presuming that all members of the civil society are equal, paying inadequate attention to social and economic contexts, and the limited definitions of ‘civic’ participation, in shaping women’s associational life (O’Neill and Gidengil, 2005; Bezanson, 2006). Moreover,

Putnam (2000, p.358-63) uncovers a positive correlation between inequality and social

16

capital, suggesting that those who are already privileged will gain most from social capital.

This is supported by his findings that women living in American states that are more egalitarian fare better in social capital than those that live in states where women have lower status (Putnam, 1993, p.177). Yet, Putnam does not explore the relationship between equality and social capital further, and merely cautions: “[a] recognition of the importance of social capital in sustaining community life does not exempt us from the need to worry about how that ‘community’ is defined – who is inside and thus benefits from social capital and who is outside and does not” (Putnam, 2000, p.358). Other scholars have taken a more nuanced approached to inequality (Burns, Scholzman and Verba, 2001; Verba, Scholzman and Brady, 1995), highlighting that the uneven distribution of resources has an impact on participation. However, these studies remain largely centered on the American experience.

Yet, while inequalities exist across democracies, the way states choose to address them varies greatly.

Accordingly, the second weakness in the dominant civic engagement literature addressed by this study is its presumption that citizens are entirely autonomous agents, untouched by state policy or action. In their preoccupation with the individual, they presume that the qualities and dispositions of civic activity derive from outside the polity, and that the state and civil society are somehow mutually exclusive realms, downplaying the impact of institutional development on the growth of the associative sector and social values (Valelly,

1996; Skocpol, 1997; Rothstein, 2002; Worms, 2002). Consequently, the dominant civic engagement literature fails to recognize that the way states shape citizenship through the rights and entitlements of the welfare state, and the bearing this has on civic engagement.

17

A Brief Note on Methodology

This study uses a mixed methodological approach that involves both quantitative and qualitative analyses. Civic participation and norms are measured by conducting statistical analyses, most often using multiple waves of large comparative datasets including the World

Value Survey, and the International Social Survey Programme’s Role of Government IV and

Social Networks II surveys. While this study does not completely overlook the vote (see for instance figure 1.6 and chapter 3), it does not form a central part of the analysis due to data limitations. The World Values Survey, which includes the widest array of indicators for associational-types and political acts, does not include measures for the vote, while the

ISSP’s Role of Government IV only includes a measure for the vote. Therefore, informal networks with friends, colleagues and kin, formal networks in associations such as unions and religious groups, as well as political acts like joining a protest or boycott are at the heart of this research. This presents an excellent opportunity to not only pay closer attention to alternative forms of participation that tend to be understudied, but also better capture the involvement of those who avoid conventional forms of involvement. The impact of these is studied both as a whole, by building indices which incorporate all civic network and norm indicators into one variable, such as a ‘civic engagement index;’ but their effects are also studied independently, for instance, to focus on women’s involvement in kin networks or political organizations. Moreover, depending on the analysis, indices are sometimes examined in their scale format, and are sometimes recoded into dichotomous and polytomous dummy variables,2 allowing nominal-level variables to be treated as interval-

2 Dichotomous dummy variables (or dummies) take the values of 0 and 1. Polytomous dummies are the recoding of variables with more than two categories into a series of dummies. For example, welfare regimes

18

level variables for the purpose of certain statistical analyses, such as regressions. Indices, variables, and techniques such as analysis of variance (ANOVA) and ordinary least squares

(OLS) regression, as well as the datasets and countries included for study are all detailed in subsequent chapters. Moreover, unlike most civic engagement studies, this report examines networking and norms both at the individual- and country-levels of analysis. At the individual-level, the next three chapters focus on the case of gender, while over the course of the subsequent three chapters, a comparative country-level analsysis is conducted across

Western advanced democracies. Chapter six includes the cross-national statistical analysis, and makes use of hierarchical linear modelling techniques using HLM6. The hierarchical linear approach is useful when data and concepts operate on multiple levels, and can provide information on cross-level interactions, in this case, between individual-level and country- level factors.

In addition to quantitative analyses, this study draws on a vast amount of literature and document reviews. Multiple areas of literature are explored, such as: civic engagement, including the more recent analyses which challenge prevailing accounts; citizenship and welfare state, including from a feminist perspective; and literature and government document reviews of income security, family, and housing policies across three divergent welfare regimes, particularly in Canada, France and Sweden. Moreover, to gain a better understanding the effect of women’s labour market activity and family life on participation, as well as to fill the sizeable gap in surveys and literature regarding child- and care-related networks, 60 personal interviews were conducted with single mothers in the three welfare regimes of Canada, France and Sweden. This analysis is particularly valuable because it dummies can be created by recoding the ‘country’ variable (which has multiple categories) according to regime-types as such: liberal regime 1/all others 0; corporate-conservative regime 1/all others 0; social democratic regimes 1/all others 0.

19

deals with an area that is understudied and largely excluded from prevailing comparative datasets. As such, the interviews reveal new information about women’s networking habits, as well as the importance of child-, parent-, and care-related networks. Survey questions were both open- and close-ended, and the snowball method was used to locate respondents.

More details about the interview process are available in chapter three.

Outling of the Study

This dissertation is divided into two sections. Part I consists of three chapters which explore the individual-level factors that shape associational life. This section includes data from multiple countries, but focuses on women as a test case. The first chapter highlights the value of approaching civic engagement from a citizenship perspective that is grounded in substantive equality. Unlike the dominant approach to the study of associational life, such an approach makes it to understand how difference and social and economic factors shape participation and trust. Focusing on the case of women underscores this point, demonstrating that despite the great gains that women have made over time, traditional gender roles continue to pose a problem to their full inclusion into citizenship, and explains why a gender gap in participation endures. The dominant civic engagement literature is ill- equipped to address such issues because it is grounded in a combination of liberal and civic republican citizenship traditions which define inclusion into the community in ways that have historically excluded women.

Chapter two examines different theories about the enduring gender gap in participation. To fill the void left by the dominant civic engagement literature, a growing

20

number of scholars interested in class- and status-based groups have explored how inequality shapes civic life, and suggest substantive resource-, psychologically-, and agency- based explanaitions, as well as methodologically-based accounts for women’s associational disadvantage. Those who subscribe to a resource-based account suggest that individuals who suffer from class- or status-based inequalities will also tend to have fewer resources such as time, money, knowledge and skills necessary for building a vibrant associational life. Others prefer to focus on the effect of psychological factors, arguing that those who are not politically oriented, for instance because their mental energies are concentrated on care giving, will be less knowledgeable about politics and the community and, thus, less likely to become civically involved. Those who adhere to an agency-based account contend that, as we are products of our environment, those who tend to network with people who resemble them (network homophily), tend to have fewer opportunities to join or be recruited into other organizations. Finally, some scholars argue that women are in fact highly civically involved, but that their lower levels are the result of a methodological flaw in the conventional approach to measuring associational life which discounts women’s networks.

The findings support both the substantive and methodologically-based explanaitions, and confirm that social and economic factors linked to gender roles put women at an associational disadvantage. Specifically, a deficit in resources, a tendency to associate with others who are similar to one’s self, a lack of interest in and knowledge about politics, and the methodological issue - discounting certain types of social ties such as parent-, child- and care-related groups - all help explain women’s lower levels of participation compared to their male counterparts. However, despite the lingering gender gap in civic engagement, there is no evidence to support the notion that women’s primary caregiving duties

21

necessarily orient them away from politics and the community; in fact, women with children had higher levels of civic engagement than non-mothers.

In an effort to address the last point and shed light on forms of participation that are typically excluded from dominant civic engagement analyses, chapter three presents the findings of personal interviews aimed at exploring how single mothers network, as well as whether or are not they are civically involved and why. Sixty interviews with single mothers in Canada, France and Sweden, demonstrate that balancing paid work and family duties is challenging and stressful, but does not detract those who want to participate from becoming civically active. Although the interviews are unable to clarify civic engagement’s exact causal mechanisms and raise questions about the reliability of existing trust measures commonly found in surveys, they shed light on the impact that children, as well as parent-, child- or care-related associations have on overall civic life. The findings demonstrate the important contribution these networks make to women’s civic engagement. In light of this, parent-, care- and child-related networks should be understood as an inherent part of civic life, not as separate from it.

While Part I highlighted the impact that class- and status-based inequalities have on participation, focusing on gender, Part II explores how country-level factors shape civic engagement. This section draws on data from multiple countries, using Esping-Andersen’s welfare regime-types as case studies. The previous section illustrated how citizenship affects participation. Traditional gender roles expressed in the public-private divide had a direct bearing on women’s civic habits compared to men. However, it is important to recognize that the state plays a key role in shaping citizenship status and, as such, has a hand in shaping participation. For instance, states can design rights, policies and institutions in ways

22

that overlook, reinforce or alleviate inequalities. At the country-level, and outlined in subsequent chapters, a citizenship approach to civic engagement emphasizes the state’s role in shaping civic life, suggesting that inequalities are created and maintained. The way that states choose to interact with the family, market and community also has a lasting effect on civic life, for instance, by designing social provisions that either ignore, reinforce or strive to eliminate traditional gender roles and the public and private divide. From a comparative perspective, this suggests that some women may fare better than others by virtue of the types of entitlements and extensiveness of social provisions that are afforded to them by different states, and this may ultimately lead to higher amounts of participation at the country-level.

In other words, it is possible that comparative participation patterns can be explained by turning our analysis to the state. Drawing on a feminist citizenship perspective, as well as

Esping-Andersen’s welfare regime typology, family, income security and housing policies are examined across the three welfare regime-types of Canada, France and Sweden, to demonstrate the important role that states play in shaping women’s citizenship status. This analysis demonstrates that across policy areas, Sweden’s comprehensive provisions lead to the most egalitarian outcomes for women, followed by France. Canada has the least egalitarian welfare policies, often leaving women fending for themselves.

In light of the way that divergent welfare regimes can shape citizenship in different ways, chapter five explores how country-level factors affect participation. It begins by considering theories commonly used to explain cross-national variation in civic engagement levels. Yet, prevailing theories, including the religious legacy of a country, the degree of heterogeneity in the civil society, and the strength of family ties, are ultimately found to be unsatisfactory because they fail to account for the role of the state. Accordingly, two

23

approaches are examined: one that focuses on the state as a hindrance to civic life adopted by the dominant civic engagement literature, where strong, large states are said to crowd-out networking; and one that centers on the way states can encourage civic engagement.

However, studies examining the impact of the state on civic engagement are so rare that they cannot be considered a literature per se, and so diverse that they cannot be said to form a theory proper.

Since no adequate theory of the state exists which can explain variation in levels of trust and participation in a comparative context, chapter six develops and test three theories of the effect of the state on civic engagement, drawing on Esping-Andersen’s welfare regime typology and an analysis of the World Values Survey. Each theory focuses on a distinctive aspect of welfare regimes including state interventionism, egalitarianism, and political authority structure, and each predict specific regime-based civic engagement outcomes.

According to the first theory, states that are highly interventionist, for instance interfering in the market and family, will crowd-out civic involvement by addressing needs that were once met by combining with others in the community through voluntary work, with social provisions of the welfare state. Thus, non-interventionist states are ideal to foster vibrant communities. The second theory suggests that states which only provide meagre, minimal social provision and do little to address inequality and social stratification will make it difficult to produce the much-needed trust required to foster associational life. Thus, egalitarian states are ideal to foster vibrant communities. Finally, the third theory contends that ‘statist’ states, in which political authority is legitimized and concentrated in the central state, away and quite apart from civil society, will lead to a political system which does not place great value on and may in fact discourage collective action. Thus, ‘non-statist’ states

24

are ideal to foster vibrant communities. The findings show that interventionist, egalitarian and non-statist social democratic regimes lead in overall civic engagement levels, followed closely by the non-interventionist, non-egalitarian and non-statist liberal regimes. Western

European corporate-conservative regimes lag across all civic engagement indicators.

Ultimately, the findings suggest that there is some merit to the egalitarian theory. For instance, the individual-level analyses in Chapters 2 and 3 show that SES indicators have a significant effect on civic engagement, and the cross-level interaction analysis in Chapter 6 confirms this – even controlling for country-level factors. Indeed, the impact of SES on associational life is reduced in the social democratic countries, tempered by highly effective, egalitarian policies. However, the findings from this study also point to the validation of the statism theory, suggesting that countries in which political authority is located in civil society and legitimized by citizens tend to have consistently higher levels of civic engagement.

The final chapter concludes by providing a discussion of the findings. This study illustrates that inequality has a tremendous effect on civic engagement levels. However focusing on social and economic context alone is not enough to explain comparative levels of civic involvement. The way political authority is structured and legitimized – whether power has historically been located and legitimized in the state, removed from civil society, is also of crucial importance. This suggests that states need to pay attention to developing policies which aim to alleviate inequality and social stratification, striving to encourage a more equitable citizenship status in the community because reducing disparities can help foster greater participation across divergent class- and status-based groups. However, it also

25

indicates that in order to foster associational life, the state must create opportunities for citizens to participate in the first place.

26

PART I: INDIVIDUAL-LEVEL FACTORS

27

Chapter 1 All Things Being Equal: Gender, Citizenship and the Authors of an Indifferent Democracy

In their preoccupation with documenting what associational life holds for individuals and communities, dominant civic engagement studies have neglected to consider the ways in which trust and networks are shaped. To address this, Part I uses women as an individual- level case study, testing the effect of inequality on associational life. Using women’s gender gap in participation as a backdrop, this chapter highlights a key failure of the dominant civic engagement approach: a flawed understanding of citizenship that is rooted in formal equality. The findings confirm what many single-country studies over time have already shown (Conway 1991; Munch, McPherson, Smith-Lovin, 1997; Nancy, Scholzman and

Verba, 2001; Karp and Banducci, 2008; Inglehart and Norris, 2003; Norris and

Inglehart, 2006; Delli Carpini and Keeter 1996): despite the great advances women have made over time, unresolved substantive equality issues translate to lower levels of civic engagement compared to men. The dominant civic engagement literature has little to say on the matter because it is premised on a view of citizenship which ignores differences between those in the community that might cause some to be at a disadvantage. Since it is reproduced in participation, inequality has serious implications for the process and progress of democracy; it can, for instance, result in a discrepancy in the citizen-feedback that reaches the ear of decision-makers, leading to a skew in the types of issues and concerns that are likely to be addressed by the state (Verba, Scholzman and Brady, 1995). Equal participation is also important because, according to civic engagement research, it is said to build a virtuous circle; the more citizens are active, the more they are likely to develop civic skills,

28

spread productive values such as cooperation and trust, and be informed and knowledgeable about politics. This chapter calls for a shift away from this dominant civic engagement approach to embrace, instead, a feminist citizenship model that is premised on substantive equality. The feminist approach presents a more complete view of the citizen as it accounts for difference, and the way social and economic factors structure civic life.

The Impact of Inequality on Civic Engagement: Exploring the Gender Gap

Women serve as an ideal test case in the pursuit to better understand how class- and status-based inequalities shape associational life. Their long-documented systematic exclusion from participation in the community, ranging from involvement in formal politics to the paid labour market, has been linked to social, economic, cultural and political factors and, more broadly, tied to their unequal citizenship status (Mariën, 2008; Hughes and

Paxton, 2008; Ramirez, Soysal and Shanahan, 1997; Inglehart and Norris, 2003; Goetz,

2008, p.17-34; Kenworthy and Malami, 1999; Rule, 1994; McDonagh, 2002; Norris, 2002, p.94-102). However, over time, they have increasingly gained rights and entitlements, joined the paid labour market in large numbers, and have, for the most part, formally been extended full membership in civil society. While the advances women have made might explain why gender-based differences in civic engagement have narrowed over time, scholars largely drawing on single-country data have found that a consistant gap persists, and does so in favour of men (Conway 1991; Munch, McPherson, Smith-Lovin, 1997; Nancy, Scholzman and Verba, 2001; Karp and Banducci, 2008; Inglehart and Norris, 2003; Norris and

Inglehart, 2006; Delli Carpini and Keeter 1996).1 A preliminary analysis of gender

1 It is worth noting that some scholars suggest that gender gap in civic engagement is largely the result of bias, with men being more likely to over-report having political knowledge, for instance, because men are less likely to choose ‘don’t know’ on a survey, compared to women (Mondak and Andersen, 2004; Dolan, 2011).

29

differences in civic engagement based on data from the ISSP Social Networks II (2001), and

the Role of Government IV (2006), as well as the World Values Survey (2000) across over a

dozen advanced Western democracies confirms that a modest but consistent gender gap in

participation endures.2

Figure 1.1 Gender Differences in Civic Engagement

8% 7%

6%

4%

2%

0%

‐2% ‐2% ‐4% ‐3% ‐4% ‐4% ‐6% ‐6% ‐8% ‐7% Membership in a Volunteer work Informal networks Informal networks Political action Social Trust Political Trust group with friends with family

Civic Engagement Indicators

Sources: ISSP Social Networks II (2001); ISSP Role of Government IV (2006); World Values Survey (2000). Note: Z-test scores show the differences in proportions between men and women are all statistically significant at the .05 level.

Figure 1.1 shows the difference between men and women across seven civic

engagement indicators, including:3

However, these studies tend to be built on single-country (U.S.) case studies, and test other aspects of political orientation (political efficiacy or interest) or civic engagement (formal or informal networking and trust). Moreover, even when accounting for bias, their findings often fail to fully explain away gender differences (Mondak and Andersen 2004). 2 More specifically, ISSP Social Networks II (2001) analyses include 13 countries: Canada, Australia, Austria, Italy, Denmark, Finland, France, Germany, Norway, Great Britain, US, Spain, New Zealand. The ISSP Role of Government IV (2006) data includes 14 countries: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Portugal, Spain, Sweden, Great Canada, Australia, Denmark, Finland, France, E and W Germany, Netherlands, Norway, Sweden, Great Britain, US, Portugal, Spain, New Zealand. World Values Survey (2000) analyses include 14 countries: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Portugal, Spain, Sweden, Great Britain and the US. 3 For details about which countries are included in the analysis, see supra note #1.

30

1) A voluntary membership index which collapses fifteen voluntary organizational categories from the World Values Survey (2000), including social welfare for the elderly; religious group; education, arts and cultural activities; human rights; professional associations; political party; labour unions; youth work; local political actions; sports and recreation; conservation, the environment, ecology and animal rights; women’s group, peace movement, organization concerned with health, and other groups.4

2) A voluntary unpaid work index which measures activity, and collapses fifteen voluntary organizational categories from the World Values Survey (2000), including social welfare for the elderly; religious groups; education, arts and cultural activities; human rights; professional associations; political party; labour unions; youth work; local political actions; sports and recreation; conservation, the environment, ecology and animal rights; women’s group, peace movement, organization concerned with health, and other groups.5

3) An informal networking with friends index which collapses four network categories from the World Values Survey (2000), including spending leisure time with: friends; people from church; people from work; and people from arts and cultural groups.6

4) An informal networking with family (or kin) index which incorporates eight different networking categories from the ISSP Social Networks II survey (2001), including how often do you visit your: mother; father; brothers and sisters; aunts and uncles; nieces and nephews; cousins; parents-in-law; and brothers- and sisters-in-law.7

5) A political action index which collapses five categories from the World Values Survey (2000), including: signed a petition; joined a boycott; attended a lawful demonstration; attended an unofficial strike; and occupied a building or factory.8

6) A political trust index which incorporates six variables from the ISSP Role of Government IV survey (2006), including: trust in civil servants; public officials deal fairly with people like me; treatment by public officials depends on

4 The alpha score for the voluntary membership index is .67 and thus meets the reliability threshold (.70). The scaled index can be used for OLS regression, and a dummy version can be used in bivariate and elaboration analyses. 5 The alpha score for the unpaid work index is .66. Great Britain is not included in unpaid work analysis because it does not include the ‘other group’ category. 6 The Cronbach's alpha for the informal networks scale is .42. While this may be expected to decrease the strength of the findings presented above, scholars have noted that alpha threshold should be approached which caution (Cortina, 1993; Field, 2009; Osburn 2000). For instance, there is evidence that using an attenuated scale to generate a higher alpha can result in an overestimate of true correlation (Schmitt 1996). Accordingly, I rely on the scale as constructed, and note that its component parts closely reflect the concept they are meant to measure. 7 The alpha score for the informal networks with family index is .68. 8 The alpha score for the political action index is 6.0.

31

contacts; politicians are involved in corruption; public officials involved in corruption; public officials want bribe.9

7) A social trust dummy variable, using the commonly used indicator ‘do you think most people can be trusted’ from the World Values Survey (2000).

Figure 1.1 shows women’s gender gap in participation compared to men. It reports the difference, expressed in percentages, between men and women across seven civic engagement indicators. Negative coefficients denote women are less likely to be civically engaged than men, positive coefficients indicate women are more likely to be civically engaged than men, and a value of 0% means that men and women are equally civically involved. Testing the differences in proportions on dummy version of each indicator reveals that statistically significant gender differences exist across both networking and trust measures, with men leading in all categories except ‘informal networking with family.’ In some cases, differences between men and women outlined in figure 1.1 are almost imperceptible, such as women’s 2% disadvantage compared to men in levels of social trust.

While such differences may appear paltry, when translated into actual activity and norms, even seemingly small gaps translate into important gender differences in civic engagement.

For example, according to the latest Statistics Canada census (2006), there are approximately thirteen million voting-aged women in Canada; based on this, a 2% gender gap in trust in Canada alone is equivalent to just over a quarter of a million - about 260, 000

- fewer women than men saying they trust others. In addition to general trust, women are also 3% less likely to trust in politicians and civil servants than men. Women also face a disadvantage in the area of volunteering and informal networking with friends, where the gender gap is 4% less than men, membership in voluntary organizations with a 6% gender

9 The alpha score for the political trust index is .70.

32

gap from men, participation in political action where women face a 7% disadvantage compared to men, and informal networking with family where women lead by 7%. Figure

1.1 also suggests that the gender gap for women is present in both demanding and less difficult forms of civic engagement. For instance, fewer women hold a membership in a voluntary association - an act which often requires minimal involvement from members – leading to a 6% gender difference. On the other hand, the gender gap narrows with respect to doing volunteer work, with women being only 4% less likely than men to do unpaid work in a voluntary group, despite the fact that this requires a greater effort on behalf of members.

Thus, figure 1.1 shows that a gender gap persists with women continuing to display lower levels of civic engagement than men in both networking and norms of trust.

Delving deeper into gendered associational patterns, scholars have found that men and women tend to favour different types of networks. Women gravitate to informal family networks, welfare- and community-based groups as well as religious organizations, while men favour more instrumental, professional groups and sports and recreation-based associations (Moore, 1990; Munch, McPherson, Smith-Lovin, 1997; Popieliarz, 1999;

Burns, Scholzman and Verba, 2001; Inglehart and Norris, 2003; Norris and Inglehart, 2006;

Gidengil, Goodyear-Grant, Nevitte, Blais and Nadeau, 2006; Lowndes 2006; O’Neill, 2006;

Caiazza and Gault, 2006).

33

Figure 1.2 Gender Differences in Voluntary Membership Types

4% 3% 2% 2% 0% 2% 4% 6% ‐ ‐2% ‐2% 4% ‐ 6% ‐ 8% ‐ ‐8% ‐9% 10% ‐ Social, Welfare or Religious Group Leisure Organizations Employment‐related Politically‐inclined Politically‐inclined Other group Service‐related Organization Organizations Organizations (NO Organizations WOMEN’S GROUPS)

Gender Differences in Types of Voluntary Memberships

Source: World Values Survey (2000). Note: Differences in proportions z-test scores show the differences between men and women are all statistically significant at the .05 level.

In order to shed light on the degree to which men and women favour some associations over others, fifteen voluntary organizations were broken down by the type of group, categorized according to their general aim (Burns, Scholzman and Verba, 2001, p.74-75;

Salamon and Anheirer, 1997; Pestoff, 2009, p.34). As such, six voluntary organizational categories were created:

- Social, welfare and service-related organizations: including social welfare, health and youth work groups10 - Religious groups - Leisure organizations: including culture, music and art, as well as sports and recreation groups11

10 The alpha score for the 3 category ‘social, welfare and service-based’ organization index is .40 for both the membership volunteer indices.

34

- Employment-related organizations: including unions and professional associations12 - Politically-inclined organizations: including human rights, peace, women’s, environmental/ecology and animal rights groups, as well as political parties and local action13 - Other groups

Figures 1.2 and 1.3 show women’s gender gap in organizational involvement compared to men. It is the difference, expressed in percentages, between men and women across seven civic engagement voluntary membership and activity indicators. Negative coefficients denote women are less likely to be civically engaged than men, positive coefficients indicate women are more likely to be civically engaged than men, and a finding of 0% indicates that men and women are equally involved.

This analysis reveals that women are less likely than men to have a membership in a leisure-oriented organization, such as sports and recreation groups, by a -8% difference, and

5% less likely than men to volunteer for one. Fewer women belong to a work-related group, such as a professional association or a union, leading to a 9% gender difference from men, and women are 3% less likely than men to devote any unpaid work to such groups.

However, women have a statistically significant 4% advantage in membership and activity levels over men in organizations which focus on social, welfare and services, as well as a

3% advantage in membership and activity in religious associations.

11 The alpha score for the 2 category ‘leisure’ organization index is .34 for the membership index, and .24 for the volunteer index. 12 The alpha score for the 2 category ‘employment-related’ organization index is .23 for the membership index, and .27 for the volunteer index. 13 The alpha score for the 6 category ‘politically-inclined’ organization index is .50 for both the membership and volunteer indices.

35

Figure 1.3 Gender Differences in Voluntary Activity by Association Type

4% 3% 3% 3%

2% 1% 1% 1%

0%

‐1%

‐2% ‐2% ‐3% ‐3% ‐4%

‐5% ‐5% ‐6% Social, Welfare or Religious Group Leisure Organizations Employment‐related Politically‐inclined Politically‐inclined Other Group Service‐related Organization Organizations Organizations (NO Organizations WOMEN’S GROUPS)

Gender Differences in Activity Levels by Associational Type

Source: World Values Survey (2000). Note: Differences in proportaion z-test scores show the differences between men and women are all statistically significant at the .05 level.

In view of women’s historical exclusion from politics and the enduring gender gap in participation, moreover, women’s involvement in politically-oriented organizations is particularly interesting. According to figure 1.2, women have a 2% membership advantage over men in ‘politically inclined’ groups - associations with mandates that lead them to take a particular stand on a political issue. With respect to membership levels, this advantage seems to be purely due to the effect of women’s participation in women’s groups; when the category ‘women’s group’ is excluded from ‘politically-inclined groups,’ women’s membership levels in this category drops by 2% when compared to men (see figure 1.2).

However, women are more actively engaged in political groups, as figure 1.3 suggests,

36

leading by 1% over men – irrespective of ‘women’s groups.’ Thus, findings from this brief survey of gender differences in civic engagement reveal that, while women may hold fewer overall memberships in voluntary associations and may do less volunteer work than men overall (see figure 1.1), they are active in the networks they do belong to, and even beyond this, extending to ‘politically-inclined’ groups that do not include ‘women’s groups.’

Women’s involvement in faith-based associations is worth taking note of in light of the sizeable scholarship which posits that an active religious life, including participation in religious associations and regular attendance of religious services can have a spill-over effect, leading to greater participation in overall networks beyond the religious community

(Uslaner, 2002; Smidt, 1999; Lam, 2006; Norris and Inglehart, 2004; O’Neill, 2006; Caiazza and Gault, 2006). According to this view, religious activity is said to act as an associational catalyst, fostering participation by providing churchgoers with civic skills and values which can carry-over into other networking opportunities. In other words, those who are religiously active are likely to be more civically engaged overall. In addition to their higher levels of membership and unpaid work in religious organizations compared to men, already outlined in figures 1.2 and 1.3, 37% of women said they frequently attended religious services, while only 26% of men said they did so – an 11% statistically significant difference.

37

Figure 1.4 Membership in Voluntary Organizations by Gender, Controlling for Religious Attendance 80%

70%

60%

Organizations 50%

40% Voluntary

in % Men

30% % Women

20% Membership

%

10%

0% Less than once a year or Only on holidays Once a month Once a week or more never

Gender, Controlling for Religious Attendance

Source: World Values Survey (2000). Note: The relationship is statistically significant at the .05 level. Z-test scores show the differences between men and women are also statistically significant at the .05 level.

In view of this, and the posited ‘spill-over’ effect of religious activity into secular civic life, it follows that women should be more civically engaged than men.14 Yet, as figure 1.4 shows, looking at voluntary membership by gender, controlling for religious attendance, does not explain away gender differences in networking.

Unlike the bivariate analyses above (figures 1.1-1.3) which serve to describe gender differences in civic engagement, elaboration analysis goes a step further by testing the impact of an additional independent variable, in this case, religious attendance, on voluntary membership levels between men and women. Figure 1.4 shows that this analysis replicates

14 The World Values Survey asks respondents ‘how often do you attend religious services,’ and includes 8 values relating to frequency ranging from ‘more than once a week’ to ‘never – practically never.’ This variable is recoded into a dummy to measure for religious practice in bivariate and elaboration analyses.

38

the original bivariate result. In other words, controlling for religious attendance does not explain away gender differences. Both men and women who regularly attend religious services form a higher proportion of those who hold a membership in a voluntary organization, but men continue to lead in membership levels over women. Thus, any advantage in civic engagement that women may have gained from their higher religious attendance levels is lost when tested against male voluntary membership levels.

Figure 1.2 and 1.3 are drawn from data which asks respondents whether they belong to or are active in any voluntary organization from stated in a list. However, it is also possible to calculate the sum of organizations men and women belong to, in order to get a sense of membership and activity volume. Figure 1.5 confirms that women are more active than men, with a greater number of them saying they volunteer in five or more voluntary associations. However, this gender difference is minimal (1%) and not statistically significant. What is surprising, however, is that despite their lower overall membership and voluntary activity levels outlined in figured 1.1, women belong to a greater number of voluntary organizations than men. In fact, figure 1.5 shows that women are 3% more likely than men to hold a membership in five or more associations – a small, but nonetheless statistically significant difference.

39

Figure 1.5 Membership and Activity Volume by Gender

6% Volunteers in 5 or more organizations

5%

% Women % Men 22% Belong to 5 or more organizations

19%

0% 5% 10% 15% 20% 25%

Membership and Activity Volume

Source: World Values Survey (2000). Note: Z-test scores indicate the differences between men and women are statistically significant at the .05 level for voluntary membership volume, but not for volunteering volume.

Scholars have suggested that it is important to look beyond formal, organizational involvement when considering gendered patterns of participation because, women tend to favour alternative forms of civic engagement, such as joining a boycott or signing a petition

– categories that are included in the ‘political action’ index (Stolle and Micheletti, 2006;

Lowndes, 2004, p.55-56; Burns, Scholzman and Verba, 2001). Figure 1.1 has already established that women lag in overall levels of political action. However, taking a closer look at gender differences for various types of political action including voting behaviour, does uncover some interesting results, as figure 1.6 shows. For instance, the gender gap in participation disappears for signing a petition and voting – two activities that require little involvement from individuals. Moreover, the gender gap widens for certain activities that demand more work – such as the 8% lag women have compared to men in attending a

40

demonstration, or the 4% lag they experience in joining an unlawful strike. Nevertheless, there is no clear pattern here between gendered participation and the difficulty of a political act. Contrary to the literature for example (Stolle and Micheletti, 2006), women are 4% less likely than men to join a boycott - something that could qualify as requiring relatively little effort.

Figure 1.6 Political Action, by Gender

3% Occupied a building or factory 5%

5% Joined an unofficial strike 9%

24% Attended a lawful demonstration 32% % Women 13% % Men Joined a boycott 17%

60% Signed a petition 60%

84% Vote in last election 84%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

Sources: Voted in last election from ISSP Role of Government IV (2006). All others take from World Values Survey (2000). Note: Z-test scores show the differences between men and women are statistically significant at the .05 level.

While they may not agree on the exact causal relationship, dominant civic engagement scholars posit that there is a positive correlation between networks of sociability, and norms of trust. Individuals and societies that are highly active should also be highly trusting of others, as well as of their political system. In light of this, it follows that since women generally display lower levels of political action, voluntary membership and

41

activity levels compared to men, they should score lower on social and political trust measures. Studies suggest that this is indeed the case, revealing that women tend to have lower levels of trust than men (Arneil, 2006, Chapter 5; Lowndes, 2006, p.220; Norris and

Inglehart, 2006, p.82; Patterson, 1999, p.173-4). Others have found a positive correlation between high levels of trust in society and support for women’s equal rights (Uslaner, 2002, p.197), as well as a link between generalized trust and political trust (Schyns and Nuus,

2007). In other words, trust is an important, integral component of civic engagement. Figure

1.1 lends support for a gender gap in trust, with women being 2% less likely to trust others compared to men, and 3% less likely to trust politicians and public officials.

In sum, this analysis shows that while the gender gap in civic engagement may be modest, it is significant and consistent across various indicatars, with women being less trusting in others and in the political system, and having lower levels of participation in every category except informal family ties. Moreover, this preliminary analysis finds that women favour many of the types of organizations that are typically associated with them – beyond family networks, such as social and service groups, as well as religious organizations. However, contrary to existing research, alternative forms of political action, such as signing petitions, do not particularly stand out as distinctively ‘female’ forms of participation. On the other hand, women are surprisingly more likely than men to engage in politically-inclined associations. Moreover, while they may be less likely to join groups, the women who do join belong to a greater number of groups and are particularly active in those groups, compared to men.

The case of women and the lingering gender gap in participation not only suggests that inequalities have an undeniable effect on civic engagement, but also that they are not

42

randomly distributed. Inequality tends to disproportionately plague certain identifiable groups, disadvantaged by virtue of their status- and class-based differences. These differences have also been used to determine who is a citizen – that is, who can get rights and entitlements, and who cannot. For instance, gender, socio-economic status, race and ethnicity have historically been used to withhold political, civil and social rights. Feminist scholars have documented the many ways this has affected women’s status, including: challenges to labour market participation and owning property; an absence of or limits on the civil rights of women in ; political rights such as the right to run for office and vote; and restrictions from a majority of fraternal societies which either excluded them entirely, or admitted women on a limited basis as the of existing male members

(Fraisse, 1995; Melby, Ravn and Wetterberg, 2009; Skocpol, 1999; Wikander, Kessler-

Harris and Lewis, 1995; Nord, 1994, p.827-9; Bock, 2002; Bermeo and Nord, 2000; Arneil,

2006; Koven and Michel, 1993; Hagemann, Michel, and Budde, 2008). Yet, as the following section will show, the dominant civic engagement literature does not have much to say on the issue of gender differences or inequality in general.

Authors of an Indifferent Democracy The dominant civic engagement scholars offer little explanation for women’s persistent disadvantage in civic engagement. Despite their preoccupation with the health of democracy, they have consistently failed to account for the effects of inequality on participation. Yet, paying close attention to social disparity is vital when conducting any inquiry into the health of democracy, for as Verba, Schlozman and Brady assert:

Democracy rests on the notion that the needs and preferences of no individual should rank higher than those of any other. This principle undergrids the concept of one person, one vote as well as its corollary, equality of political voice. Thus any system

43

that denies equal participation rights violates a fundamental principle of democracy (Verba, Schlozman and Brady, 1995, p.2-3).

Such a view of democracy, however, has not always prevailed. In fact, we find in

Tocqueville an understanding of democracy that champions gender inequality. Democracy in America (2000), written in the 1830s - at a time when civic and political activity was the privilege of white males - Tocqueville highlights and even praises inequality as a valuable tool to maintain order in society. At this time, he famously asserts: “Americans of all ages, all stations in life, and all types of dispositions are forever forming associations…” and goes on to make the case for a connection between the art of associating, democracy and a distorted version of ‘equality’ that is based explicitly on the exclusion of women

(Tocqueville, 2000, p.513). Tocqueville does not downplay women’s inequality, but defends traditional gender roles in which, “[a]lmost all the men… either enter politics or practice some calling, whereas limited incomes oblige the wives to stay at home and watch in person very closely over the details of domestic economy…” (Tocqueville, 2000, p.598). He describes American women’s status, such as their exclusion from politics and never being

“… allowed to leave the quiet sphere of domestic duties…” (Tocqueville, 2000, p.601). He praises the fact that American women “…seem to take pride in the free relinquishment of their will…” and that “…one never hears an adulterous noisily proclaiming the rights of women…,” (Tocqueville, 2000, p.602). Based on this portrayal of American women as accommodating, compliant ‘inferiors’ (Tocqueville, II, Ch.12, p.600), and their reluctance to disturb the ‘natural order’ of things by demanding equality, Tocqueville concludes that,

“…if anyone asks me what I think the chief cause of the extraordinary prosperity and growing power of this nation, I should answer that it is due to the superiority of their women” (Tocqueville, 2000, p.603). For Tocqueville, gender inequality is favourable

44

because, as the rise of democracy dismantles historic allegiances, it is an effective means to secure at least one long-standing way to order the social world. (Tocqueville, 2000, p.601;

Boryczka, 2009, p.282). Not only does he overlook women’s civic engagement, but he also fails to spot the contradiction in extolling the virtues of a political system that prided itself on being democratic while excluding over half of its citizens (women and non-whites).

Women’s limited rights and the implications this has on their ability to participate in the decision-making process is of no consequence to Tocqueville’s definition of democracy; his democratic America is built on a presumption of inequality.

Despite their exclusion from conventional forms of political participation, however, women were active around the time of Tocqueville’s American visit in the 1830s, particularly on abolition and temperance issues (Ryan, 1992, p.14-5). Yet during this time, there was still little consensus among women that they should seek equal rights, and many espoused maternalist views which exalted their traditional role as wives and mothers (Orloff,

2006, p.237; Koven and Michel, 1993). Thus, it might be tempting to assume that, because

Tocqueville was writing in the early nineteenth century, he was but a product of his time; that later civic engagement scholars – who studied the effect of the civil society on democracy after universal suffrage, two World wars, and the steady increase in women’s mobilization efforts – would take such inequalities more seriously. As it turns out, however, it would be a mistake to make such an assumption.

In the early 1960s, when The Civic Culture was published, the second wave of the women’s movement was on the verge of making its mark. In the post-war years, women had won many rights, but the traditional remained the norm, where women were primarily homemakers, and therefore remained relatively unequal to men (Morgan, 2006).

45

Against this backdrop, Almond and Verba (1965) identify a lingering gender gap in participation, despite advances in formal equality through the extension of suffrage to women, but fail to analyze it further, or differentiate this from more substantive equality outcomes. Unlike Tocqueville, Almond and Verba do not highlight women’s unequal status as a favourable strategy to maintain social order, but nor do they deplore it and its implications for democracy. Instead, they accept it, failing to probe further when asserting that, “…women differ from men in their political behaviour only in being somewhat more frequently apathetic, parochial, conservative and sensitive to the personality, emotional, and esthetic aspect of political life and electoral campaigns” (Almond and Verba, 1965, p.325).

In fact, as Pateman underscores, Almond and Verba interpret their own findings in a biased way that skips over their discovery of a link between socio-economic status and gender on the one hand, and political activity and orientations on the other (Pateman, 1989). They downplay the differences uncovered between citizens within political cultures championed as ‘civic,’ such as Britain and the United States – discounting important divergences between those who are actively engaged and politically oriented, and those that are not. Yet, since there exist such a sharp, systematic, division within these civic cultures – and one that is not random, it is misleading to refer to the civic culture – as though it characterizes the community on a uniform basis (Pateman, 1989, p.156-8) Consequently, Pateman asserts:

That the orientations of the civic culture are found mostly among individuals from higher SES backgrounds, and that it is also these individuals who are likely to belong to voluntary associations and to be political active, is not a finding that empirical theorists of democracy can take for granted. Such findings suggest that political apathy is… [a] socially structured and maintained phenomenon; what might be called a non-participation syndrome exists. It is precisely this syndrome that needs to be explained; why lower SES – and female sex – lack of civic orientations and political inactivity tend to be associated (Pateman, 1989, p.156-7).

46

Since most draw on Tocqueville and Almond and Verba, it is perhaps not surprising to find that dominant social capital scholars have also tended to be uninterested in the implications of gender differences and exclusion on democracy. Bourdieu’s class-based approach stands as an exception to this and has been praised by some feminist researchers for its attention to the role of history and culture in reproducing inequality, and the linking of social capital with other forms of capital (Bourdieu, 1984, 1986, 2001; Arneil, 2006). Yet, despite such advantages, others caution Bourdieu’s social capital approach continues to pose significant methodological problems, presents gender as secondary to class, and ultimately requires significant assessment and amendment before it can effectively be claimed by feminist scholars (Hughes and Blaxter, 2007, p.122; Lovell, 2000; Le Hir, 2000).

In his earlier Italian study, Making Democracy Work (1993), Putnam hints at a positive link between political equality and social capital, declaring that, “[e]quality is an essential feature of the civic community” (Putnam, 1993, p.105). However, his Italian study does not pay adequate attention to women’s unequal status, and how this might affect their participation. Later, in his diagnosis of American civic decline, in Bowling Alone (2000),

Putnam does test for the effects of gender, race and class, and confirms that a positive correlation exists between equality and participation (Putnam, 2000, p.358-9). Accordingly, he concludes, “[c]ommunity and equality are mutually reinforcing, not mutually incompatible. Social capital and … equality moved in tandem through most of the twentieth century… American states with the highest levels of social capital are precisely the states most characterized by economic and civic equality…” (Putnam, 2000, p.358-59; Caiazza and Putnam, 2002). Despite these findings, however, Putnam does not explore this relationship further because class- and status-based differences fail to provide the answer to

47

his central question: what is causing the overall decline in social capital in America? In other words, the vast inequalities in participation that were the norm in the heyday of American social capital, roughly from the Progressive Era to the 1960s, are immaterial; taking this baseline of ‘equality’ for granted, what counts for Putnam is not the cause or effect or even the narrowing of enduring inequalities - but that the decline in social capital happens across the board. For instance, the 5% advantage that whites have over blacks in participation in local politics between 1974-1994 is quickly brushed aside to highlight that both groups have suffered equally in terms of a withdrawal from the community over this period of time

(Putnam 2000, p.186). Thus, Putnam concludes:

Civic disengagement appears to be an equal opportunity affliction… The trends are down among women and down among men… down among renters and down among homeowners, down in blank ghettos and down in white suburbs… down among the affluent and down among the impoverished…down among full-time workers and down among homemakers… (Putnam 2000, p.185).

Furthermore, Putnam considers gender when looking at the possible effects of the

‘pressures of time and money’ on social capital in his America study, finding that women network less than men, and that women who work part-time have the highest rate of civic activity. In other words, Putnam finds that full-time work hinders both formal and informal networking (Putnam, 2000, p.201). Yet, he does not expand on how these pressures of time and money are gendered. He does not elaborate on women’s dual roles as caregivers and workers, how this shapes the availability of resources at their disposal and how, together, these factors affect women’s networking; nor does he consider how the state might relieve this through policies such as childcare, family and maternity benefits. Rather, gender is approached as an individual characteristic that ‘happens’ in society – not as a category that

48

is constructed and maintained institutionally, and has social and economic ramifications which affect participation.

Putnam is not alone is linking women’s entrance into the paid labour market to lower levels of participation. Coleman, and Fukuyama raise similar concerns, with conservative overtones, linking low or declining levels of civic engagement with the weakening of the traditional nuclear family and tight-knit communities triggered in large part by women’s entrance into paid work (Fukuyama, 1999, p.117-120; Coleman, 1998, p.118). According to

Coleman, the increasing breakdown of families and communities signals the decline of both social and human capital for successive generations to come (Coleman, 1998, p.118).

On the other hand, looking at Great Britain, Hall points to the way women’s networking more than doubled at a time when women were entering the labour market – between 1959-1990, with a rise of 27%. In light of this he argues, “[s]ocial capital has been sustained in Britain largely by virtue of the increasing participation of women in the community” (Hall, 2002, p.37. Lowndes, 2000, p.533). Similarly, scholars studying gender differences in participation in America have found that working women are, overall, more active than homemakers (Scholzman, 2000; Burns, Scholzman and Verba, 2001). Such inconsistencies in the relationship between women’s labour market activity and civic activity hint at the possibility that women’s dual roles as paid and unpaid workers may act as a ‘double-edged sword,’ offering women greater opportunities to join in different kinds of networks, while also draining them of the resources required to invest in their associational lives (such as time).

The dominant civic engagement literature has opted to overlook the implications of gender-based inequality on participation and politics. Instead, they either accept the

49

inequality as favourable, as Tocqueville does; view it as natural, as Almond and Verba seem to in The Civic Culture; or downplay its effects, treating it like an individual trait that can be controlled for – as Putnam and the prevailing social capital approach does. Ultimately, this scholarship presents an image of civil society as primarily being made up of individuals who each have the equal capacity, will, and ability to join a social network because they are detached from, and undeterred by, any social, economic or political context; those who choose to network will produce better communities whose togetherness will support democracy, and those who choose to withdraw will punish their communities with a myriad of social ills that will erode democracy. This snapshot of idyllic life in the civil society presented by the dominant civic engagement literature suggests that, above all else,

‘democracy’ is the sum of the community’s actively engaged parts. Exclusion, and the enduring inequalities in resources, rights or opportunities that structures everyday life - often reinforced by institutions - do not factor into this definition of democracy.

Thus, something is missing from the civic engagement equation: a proper conceptualization of citizenship. Citizenship accounts for inclusion or exclusion in the community which can help shed light the relationship that citizens have with each other, as well as with the institutional context they are embedded in. The following section presents a brief sketch of the liberal and civic republican citizenship traditions which are at the heart of prevailing civic engagement approaches, and outlines some of their main lacuna as identified by feminist scholars concerned with citizenship and participatory democracy.

Rather than utterly abandon classic understandings of citizenship, some feminist scholars have worked to make these more inclusive by demonstrating how they are built upon and reinforce the wider exclusion of women in society. They do this by unravelling the myth of

50

the citizen as an abstract (genderless, classless, raceless etc.), rational agent – an understanding that presents a false universalism, downplays the importance of difference and particularities in the status and performance of citizenship, and obscures the historic exclusion of groups, such as women. Moreover, feminist scholars highlight the exclusionary effect of the artificial divide separating the ‘public’ and ‘private’ into neat spheres that endure in classic citizenship traditions; a demarcation which delineates the parameters for citizenship rights and obligations and which, as the next section shows, has resulted in women’s differential citizenship status.

LOCATING ‘THE CITIZEN’ IN CIVIC ENGAGEMENT’S DEMOCRACY This section proposes that the pursuit for a proper understanding of associational life must begin with a reconceptualization of citizenship. Citizenship is a status bestowed on citizens, providing membership in the community through rights and entitlements that are not available to non-members. Citizenship is also a performative practice since, in return for this membership status, citizens are expected to fulfill certain obligations, including participation (Lister 1997a, 1997b 2003). In Western advanced democracies, citizenship is linked to equality in the sense that it is expected that all individuals in the community have the same rights, entitlements, capacities and duties. This understanding of equality, reflected in both the liberal and civic republican citizenship traditions that are examined below, is one of formal equality which presumes that all individuals are the same and should be treated in the same way. However, this section demonstrates that a citizenship model, grounded in formal equality, is not suitable because it overlooks how class- and status-based inequalities affect participatory democracy. When citizenship is rooted in formal equality, it overlooks

51

differences such as gender, race and class, as well as the social and economic conditions of individuals, which can pose limits on the enjoyment of rights and entitlements, capacities and the ability to fulfill civic duties.

This section reviews the liberal and civic republican traditions of citizenship, borrowed by the dominant civic engagement literature. Using women as a case study, it proposes that a feminist understanding of citizenship is required if we seek to further our understanding of civic engagement. The former is rooted in substantive equality which, unlike formal equality, is premised on the understanding that individuals should be treated differently to make them equals. Drawing on a citizenship model grounded in substantive equality can account for the social and economic conditions of individuals which translate into the class- and status-based inequalities – all of which are overlooked by dominant civic engagement studies. Such an approach can account for the fact that, while women were initially formally excluded from full citizenship, their status as citizens remains precarious even after the extension of rights and entitlements due to lingering social and economic inequality, as subsequent chapters confirm. It is a lack of substantive inequality which explains women’s enduring lower levels of civic engagement.

Marshall, who developed a highly influential citizenship model, differentiates between three different layers of citizenship rights, including: 1) civil rights such as those concerning individual freedoms and associated with the institutions of legal justice like the courts, said to arrive in the eighteenth century; 2) political rights such as those concerning democratic participation and associated with democratic institutions like Parliament, said to arrive in the nineteenth century, and; 3) social rights such as those concerning a guarantee a minimum standard of welfare and related to welfare state institutions, said to arrive in the

52

twentieth century (King and Waldron, 1988; Marshall, 1992). While theories about citizenship obligations are much less developed than those of rights, one school of thought suggests obligations can be understood as a form of contract or consent to be governed. Not far behind a right of citizenship is a corresponding duty, because “…rights require obligations for their fulfillment, since no right may exist without an obligation to help the right exist…” (Janoski, 1998, p. 53; Marshall, 1992). For instance, citizens have a right to a fair trial, but are obliged to serve on a jury when called on to do so (Janoski, 1998, p.53).

Citizenship obligations can range from paying taxes and voting, to being tolerant of others and obeying laws, to being called for duty by the state under conscription.

Citizenship approaches can be distinguished based on how they balance rights and duties. Two main traditions have been identified, including: liberalism, which conceptualizes citizenship as status and emphasizes rights and the individual; and civic republicanism, which conceptualizes citizenship as a performative practice, and focuses on obligations and the interest of the wider community (Lister, 1997a, p.7; Kidd, 2002, p.329).

Others have added a third tradition, that of expansive democracy (or the social democratic approach), which represents a ‘middle-ground’ between the liberal and civic republican tradition. Expansive democratic citizenship rejects liberal individualism in favour of an approach that emphasizes social relations not only within the community, but also between the community and institutions. Moreover, it opts for a balance between civic status and practice (Janoski, 1998; Bussemaker and Voet, 1998; Siim, 2000; Lister, et.al, 2007, p.24).

The dominant civic engagement literature reflects elements of both the liberal and civic republican citizenship traditions, but as it does not adequately conceptualize citizenship, it draws on a weak amalgam, combining elements, as well as trappings, of each.

53

The liberal individual involved in networks is the launching point for scholars who have contributed to the prevailing dominant civic engagement. According to this view, citizens of a highly active community are not understood to be ‘selfless saints;’ they do not join voluntary organizations because they have risen above their self-interested individualistic tendencies (Putnam, 1993, p.88). In fact, individualism has a central place in dominant civic engagement literature. For instance, Tocqueville understood it as an important check on the potential oppression and tyranny of pre-modern modes of social organization. He recognized that, prior to democracy, feudal society was not working in the interest of the common good.

Under such conditions, “…the world was under the control of a few rich and powerful men… It gratified them to make out that it is a glorious thing to forget oneself and that one should do good without self-interest, as God himself does” (Tocqueville, 2000, p.525).

Thus, individualism represented the new way of ordering society, under the conditions brought about by democracy. Over a century later, Coleman’s approach to social capital, which starts from rational agency, not only takes individualism for granted but, values self- interest as an explanatory factor, driving individual behaviour (Coleman, 1998). Putnam, who builds on Coleman, contends: “[t]he dichotomy between self-interest and altruism can easily be overdrawn, for no mortal, and no successful society, can renounce the powerful motivation of self-interest. Citizens in the civic community are not required to be altruists”

(Putnam, 1993, p.88).

Yet, civic engagement scholars recognize that, when taken to the extreme, individualism also carries with it inherent dangers. Specifically, the very individual freedom that is a component of democracy, Tocqueville contended, is also a threat to its very survival. With the old feudal, hierarchical bonds broken, there is little in democratic society

54

keeping free, autonomous, ‘liberal individuals’ from turning inwards, away from the community, to tend to their personal interests; the ensuing social isolation would inevitably result in apathy and anomie, and ultimately weaken democracy (Tocqueville, 2000, p.524-

28). As such, the practical, civic republican citizenship tradition is called upon, emphasizing social networks and the common good, and equating civic virtue with the norms of trust and cooperation that emanate from ties with other members of the community (Herreros, 2000).

Thus, Tocqueville states, “when no firm and lasting ties any longer unite men, it is impossible to obtain the cooperation of any great number of them unless you can persuade every man whose help is required that he serves his private interests by voluntarily uniting his efforts to those of all the others” (Tocqueville, 2000, p.517). According to Putnam, the inherent dangers of increased individualization are evident in dwindling levels of social capital in the United States brought on, to a significant extent, by the uses and effects of technological advances such as television, turning citizens into passive, lethargic, distrusting loners. (Putnam, 2000, Chapter 13).

Thus, for dominant civic engagement scholars, norms of trust and reciprocity stemming from civic activity, and borrowed from the civic republican citizenship tradition, are said to help resolve social traps stemming from self-interested behaviour. For scholars like Coleman, social capital provides a cogent solution to dilemmas of collective action such as the prisoner’s dilemma or free riding; it explains how otherwise self-maximizing individuals choose to cooperate - even when it may not be in their immediate interest to do so (Field, 2008, p.24). An active civil society is more likely to resolve dilemmas of collective action because members of cooperative relationships benefit from a greater level of trust and reciprocity which can facilitate communication, and reduce the level of risk and

55

uncertainty involved in social exchanges (Coleman, 1998; Putnam 1993, p. 177-8). In this light, associational life through participation in intermediary organizations is an effective way to marry the liberal individual’s ‘self-interest,’ without compromising the common good that lies at the core of the civic republican citizenship conception; the result is something Tocqueville refers to as ‘enlightened self-interest’ (Tocqueville, 2000, p.524-28).

According to Putnam, enlightened self-interest is less myopic and more public-spirited than the average asocial, egoistic brand of self-interest (Putnam, 1993, p.87-88). In this context, the practical side of citizenship, via social networks, is championed for its collective benefits, equating civic virtue with the trust and cooperation that emanates from ties with other members of the community (Herreros, 2000).

Yet, the liberal and civic republican citizenship traditions tend to understand citizenship as a universal experience, taking place in the public sphere, absent particularities.

They view everyone as an equal citizen with equal rights, entitlements and capacities to participate. In other words, these approaches to citizenship are grounded in formal equality, requiring difference be transcended. As the following section demonstrates, when difference cannot be overlooked, individuals are demoted from full-citizenship; in the case of women, they are relegated to the private sphere of the home and domestic life. Thus, in drawing on these citizenship approaches without question, the dominant civic engagement literature also adopts their presumptions. In light of the limitations of these traditional approaches to citizenship, some feminist scholars have broadened the concept to consider the influence of difference and inequality on the citizenship ideal and its practice. The result is an approach to participatory democracy grounded in substantive equality, accounting for the realities of civic life as being socially and economically structured, where the boundaries of the public

56

and private are understood as Man-made constructs and, hence, capable of change. The following section links central assumptions of the dominant civic engagement literature to the liberal and civic republican citizenship traditions. Drawing on feminist scholarship, it highlights how these approaches are premised on a narrow, gendered understanding of the citizen.

1. Liberal Citizenship: A Gendered Reality-Check The liberal citizenship tradition conceptualizes society as consisting of abstract

(absent class or status-differences), autonomous individuals who possess equal rights. This approach tends to understand power in a uni-dimensional way, originating from individual capacity of self-determining rational agents (Faulks, 2000, p. 57-8). Yet, in practice, civic behaviour does not merely spring from individual capacity, but is tied to citizenship status, and shaped by social and economic factors that constrain self-determination. The rights attached to citizenship, for example, have never been equally or automatically granted to those in the community, as women’s historical struggle to gain full rights illustrates. For instance, while the development of citizenship status in France is often described as ‘active,’ demanding an extension of citizenship to all through the French Revolution, this interpretation overlooks the fact that women were excluded as potential citizens. French women were granted the right to vote only after a long, protracted struggle of their own, finally achieving suffrage in 1945. In contrast, Britain’s development of citizenship, often described as more ‘passive,’ negates the rallying effort that went towards extending political citizenship rights to women (Turner, 1990; Siim, 2000).

As with civil and political rights, social citizenship rights were also not automatically granted to all individuals equally, as the liberal citizenship tradition implies. Historically, living in and contributing to the community was not enough to warrant social citizenship

57

rights if the claimant was a woman. For example, while the British Beveridge Report (1942) offered one of the most progressive welfare programs in the mid-twentieth century, it constructed married women as a wholly distinct and separate class (Class III) of contributors

(Beveridge Report, 1942, paragraphs 108, 117). This Class III status not only erected great barriers to married women’s employment, but also resulted in a loss of any social security contributions they had made prior to marriage, rendering them entirely dependent upon their husband’s contributory benefits (Beveridge Report, 1942, paragraphs 339).

Moreover, since social rights are generally tied to labour market participation, women continue to face difficulties with regards to claiming them. The difficulties that arise from the link between paid work and social rights is reflected in workfare policies in countries like Canada and the US which make benefits conditional upon labour market activity, even for single mothers of small children who have no access to childcare (Gurstein and Vilches, 2009; Vosko, 2002; Mink, 2002). In light of such policies, scholars identify the presence of a two-tiered welfare state in which the more meagre, precarious social assistance programs serve a predominantly female clientele, while the more generous contributory social insurance targets a predominantly male clientele (Sainsbury, 1993; Orloff, 1993).

Alternatively, women are often granted social rights on the basis of fulfilling their traditional gender roles. For instance, when unemployment levels skyrocketed in France in the 1970s – a country lauded for its support for women’s employment, a childrearing policy (the allocation parentale d’education, now the prestation d’accueil du jeune enfant) was put in place to discourage women from entering the labour market, fighting the unemployment of men by enticing women back into the home for a period of up to three years (Jenson and

Sineau, 2001, p.100).

58

Thus, contrary to the highly abstract and neutral approach to citizenship that is implied in prevailing civic engagement analyses which presume that participation is dependent on the innate capabilities of individual citizens, a feminist citizenship perspective demonstrates the extent to which social and economic context matters to equality. Indeed, inequalities in the civil society are likely to be reproduced in civic engagement (DeFilippis,

2002, p.783). Over time and across space, women’s status as rights bearers and claimants has been greatly determined by traditional gender roles that have been imposed upon them and reinforced in policies. This is reflected, for instance, in the endurance of the industrial- era ‘family wage’ which presumes a male-breadwinner and female caregiver and continues to inform the logic of many welfare states, as well as certain tax schemes. Therefore, the case of women illustrates that being a good citizen and doing civic things requires more than sheer individual willpower – as the liberal approach to citizenship suggests. Rather, citizens are bound by a wide range of affiliations and collectivities which influence their goals, constrain their abilities, and shape their capacities as citizens. A citizen’s location in the social matrix inevitably affects the resources and opportunities available to them (Loury,

2002, p,102). Thus, as Loury, contends:

…[the] context within which individual maturation occurs strongly conditions what otherwise equally competent individuals can achieve. This implies that absolute equality of opportunity, where an individual’s chance to succeed depends only on his or her innate capabilities, is an ideal that cannot be achieved…. An individual’s social origin has an obvious and important effect on the amount of resources that is ultimately invested in his or her development” (Loury, 1977, p.176).

Furthermore, the liberal citizenship tradition’s appeal to rational agency has long been used as a template for the qualities of the ideal citizen and political agent in the public sphere, and pitted against women’s traditional roles as wives and caregivers in the private

59

sphere (Meagher and DiQuinzio, 2005, p.3; Hobson and Lister, 2002, p.24-5; Lister, 2003, p.70-1). As such,

This distinction locates political action in a public sphere of abstract reasoning and the advancement of independently determined interests, while locating the activities of meeting human material and psychological needs and taking care of dependent persons in a private sphere where emotion and the maintenance of interdependence are paramount. The public/private distinction is thus crucial to maintaining the structure of gender because it supports the exclusion of women from the public sphere that also follows from abstract individualism and essential motherhood (Meagher and DiQuinzio, 2005, p.3).

Such a view is reflected by Tocqueville, who understands American women’s station in life as centered in the private sphere of the home – outside of the realm of politics. He states:

It is hard to explain the place filled by political concerns in the life of an American. To take a hand in the government of society and to talk about it is his most important business and, so to say, the only pleasure he knows. That is obvious even in the most trivial habits of his life; even the women often go to public meetings and forget household cares while they listen to political speeches. For them clubs to some extent take the place of theatres (Tocqueville, 2000, p.243).

Tocqueville thus locates women in the private sphere and, as such, presents them as merely passive spectators, rather than active citizens, despite early women’s mobilization not only in America, but across Western advanced democracies.

Hence, the conventional approach to liberal citizenship is rooted in formal equality, where differences are overlooked and all are presumed to have equal rights, entitlements and capacities. This section has shown that such an approach to citizenship reinforces the public- private divide, conceptualizing the ‘private sphere’ of the home and the ‘public sphere’ of politics and civil society as mutually exclusive. It is a very real divide in the sense that it has concrete effects on the citizenship of women, but it is not a natural boundary. As feminist scholars and activists have long argued, the public and private are not separate spheres; the dividing line between them is but a political construct.

60

2. Civic Republican Citizenship: A Gendered Reality-Check

Just as with the liberal citizenship tradition, civic republican citizenship presents a false universalism that overlooks differences and inequalities; a community of equals working for the common good. According to this view, active citizenship takes place in the public space, which is open and ‘common to all,’ and where citizens meet as equals, on a face-to-face basis, to participate in deliberation and decision-making. The civic republican citizenship tradition dictates that individuals are not true citizens unless they have a direct hand in shaping the political outcome of the community, and must remain active for as long as they intend to be a member of the community. In other words, one is not born, nor ever becomes a citizen; citizenship is a continual process, a work-in-progress (Johansson and

Hvinden, 2007, p.39-40; Snyder, 1999, p.3). This is the central idea of Rousseau’s social contract where, as Heater argues, “[t]he people are consequently in turn citizens and subjects: citizens while formulating the General Will; subjects when obeying the implications of those decisions. But in both capacities they are truly free, free from any arbitrary authority. Nor must this participation be distorted by gross inequalities between citizens…” (Heater, 2004, p.69)

Yet gross inequalities abound, and exclusion from this social contract has never been arbitrary. The exclusion of women from the sphere where citizenship is won, through participation and deliberation in the public sphere, is diametrically opposed to their role in the private sphere; a point that has long been made by feminist scholars, and that serves to highlight that inequalities are not random - but are created and maintained (Pateman, 1988;

Lister et.al, 2007, p.30-1). For example, the various incarnations of the civic republican citizen ideal are characterized by what are commonly regarded as male attributes. This is

61

reflected in the historically masculine character of ‘citizen-soldiers’ promoted by theorists such as Machiavelli and Rousseau, representing both military service in defence of the community, and civic participation (Snyder, 1999, p.1; O’Ferrall, 2001, p.127-33; Snyder,

2004, p.213; Mettler, 2005, p.3; Ruth Lister et.al, 2007, p.20). However, as Snyder asserts, the citizen-solider ideal is built-in contradiction to everything that represents ‘femininity,’

“fusing armed masculinity onto citizenship” (Snyder, 2004, p.227). On the other hand, when women are included into the community, it is often based on typically feminine characteristics, such as their roles as mothers. Thus, while the idyllic citizen-soldier would bear arms to protect the community, the citizen-mother would bear children and raise the community’s future citizens. As Pateman notes,

Their [women’s] political duty (like their exclusion from citizenship) derives from their differences from men, notably their capacity for motherhood. The 19th century doctrine of republican motherhood provides an illustration of the multiple layers of meaning of motherhood as a political status…They were to be the subordinate companions of citizens, but with their own political task; they were to be republican mothers (Pateman, 1992, p.16-7).

Echoes of the ‘citizen-mother’ are found in modern welfare state policies from early mother’s pensions to more recent family and welfare benefits which reward women by offering social citizenship rights for those who perform the citizenship duty of mothering.

Another dimension of idyllic civic republican life which gained prominence with the development of the welfare state, and which has posed difficulties for women, is the obligation to engage in paid work (Lister, 2003, p.19-20). According to this view, the citizen-worker participates in the paid labour market and, as such, contributes to the public coffers towards the building and welfare of the community (Sarvasy, 1997, p.57). Since the late twentieth century to the present, social rights have increasingly become tied to labour market participation, whether it is through social security benefits, or activation and

62

workfare policies. This brings to light the degree to which citizenship is constructed in a way that excludes women since, having long faced limits to labour market participation, women are bound to have a harder time securing adequate social rights. In advanced

Western democracies, for instance, such as Britain, Germany, the United States, and the

Netherlands, women were restricted from entering the paid labour market until as recently as the mid-twentieth century. Marriage bars, mainly adopted in female-dominated areas such as teaching, clerical work, nursing and civil service positions, forced women to relinquish their jobs upon marriage and disqualified married women from being hired (Allen, 2005, p.221).

Moreover, instead of accounting for context and difference, the civic republican approach adheres to a view of citizenship that is rooted in formal equality. It strives to transcend the particularities of members of the community to focus on an impartial, universalist, ‘common good.’ Elevating the citizen as ‘public actor’ is the civic republican’s answer to the individualism inherent in the liberal citizenship tradition, “in which individual citizens are reduced to atomized, passive bearers of rights whose freedom consists in being able to pursue their individual interests. The reclaiming of active, collective politics as the essence of citizenship is pivotal to contemporary civic republicanism…” (Lister, 1997b, p.32). Thus, like the liberal citizenship model, the civic republican citizenship tradition creates a divide between ‘the public space’ in which participation, deliberation and decision- making takes place, in contrast to the private space, where identities and private life unravel.

In the civic republican tradition, the public space must remain a universal plane where all citizens are equal and homogenous and, as such, individuals can only become citizens when they learn to ‘privatize’ their particularities and differences (gender, race, class, )

(Laborde, 2004, p.61). In fact, as Turner and Hamilton assert, since the very essence of the

63

civic republican citizen is participation in the public sphere, “[t]he term ‘private citizen’ is…an oxymoron.” (Oldfield, 2003, p.192).

That the dominant civic engagement literature underestimates the importance of rights and resources as prerequisites to fulfilling the duties that the civic republican tradition demands of ‘good citizens’ is reflected in an example by Almond and Verba. A lack of appreciation for the way citizenship status, and the broader social and economic context, structure civic life is crystallized in The Civic Culture (1965). While they acknowledge the difficulties involved with balancing competing duties and private and public obligations, necessary for being a virtuous citizen – Almond and Verba fail to link this to inequality, or how such difficulties have serious gender implications. This is particularly the case since the

‘failed citizen’ in question is a woman whose private duties as mother have interfered with her public civic duties. Almond and Verba state:

The good citizen does not equal the good man. No zealous advocate of good citizenship would argue that political participation ought to be pursued to the neglect of all other obligations… Nor would one want his civic activity to be at the expense of his private obligations. Surely the lady described by Riesman who left her screaming children locked in their room while she attended a meeting of a neighbourhood improvement association does not represent the ideal toward which the advocates of good citizenship are striving. There will, of course, always be conflicts between the demands of different roles, but the obligations of one role do not replace those of another (Almond and Verba, 1965, p.120).

This exemplifies how the dominant civic engagement literature appeals to the civic republican citizenship tradition, without taking note of the presumptions it makes about a citizen’s social status and access to basic resources, such as free time to engage in associations, or money to pay for a babysitter. Instead, Almond and Verba merely underline the importance of civic and political participation while utterly ignoring the implications of the inequality they casually describe.

64

Ultimately, striving for the civic republican ‘common good’ is an exclusionary process, requiring the surrender of differences and overlooking inequality (Lister, 1997b, p.33). As such, it is not the ideal path to participatory democracy. It begs the question: in a heterogenous society characterized by inequality, who determines the content of ‘the common good’? Under conditions of inequality, a ‘common good’ approach can serve to reinforce exclusion since, in silencing difference, the interests of those who are already disadvantaged are likely to be subordinated to more powerful groups (Lister, 1997b, p.33).

Young asserts,

Assuming a discussion situation in which participants are differentiated by social position or culture, and where some groups have greater symbolic or material privilege than others . . . definitions of the common good are likely to express the interests and perspectives of the dominant groups in generalized terms. The less privileged are asked to put aside the expression of their experience . . . (Young, 2000, p.43).

Therefore, because they draw from classic liberal and civic republican citizenship traditions, the prevailing civic engagement literature does not recognize how socially situated individuals really are, and how citizenship status and practice are intertwined. As such, they fail not only to account for women’s limited citizenship status, but also how this may restrict women’s citizenship practice.

Considering women’s citizenship status over time raises two points; first, that citizenship is not fixed or random, but based on political choices and cultural preferences; as such, citizenship can change over time and space. Second, the work of early women reformers who affected change across advanced Western states without having the full complement of citizenship rights demonstrates that participatory democracy does not solely or primarily stem from social cohesion, but also from the struggle for equality itself

(Skcopol and Fiorina, 1999, p.14-5). However, there is an important link between the rights

65

and entitlements that accompany citizenship status, participation and democracy which deserves the attention of dominant civic engagement scholars. Influenced by both the classic liberal and civic republican citizenship traditions, dominant civic engagement scholars presume that participation springs from individual capacity, unimpeded by social or economic context. Moreover, they suggest that the disengagement that can result from individuals pursuing their interests can be allayed, and collective action problems overcome, by appealing to civic republican virtues of active participation, holding the community together with norms of trust and cooperation. Yet, members of the civil society are far from equal in status, and this is reflected in their ability, capacity and readiness to participate. In their classic forms, both citizenship traditions adopted by the prevailing civic engagement literature overlook women’s inequality. They falsely assume a universal citizenship experience which necessarily requires particularities such as class- and status-based differences be ignored, and equate civic and political life with the public sphere, in opposition to the private sphere.

Feminist scholars have shown that a comprehensive approach to citizenship which surpasses the limits of both the liberal and republican traditions by recognizing civil society’s diversity is not only preferable, but it is possible as well. Contrary to the false universalism which characterizes the classic approaches to citizenship, Young contends

“[p]eople can aim at agreement in the sense of being open to changing their positions as a result of discussion, however, without acceding to the claim that there is a single set of interests and order of goods to which they can all agree . . . agreement . . . can leave intact differences of affiliation and perspective, and even give them prominence in discussion”

(Young, 2000, p43-44). In order to better understand participation in the community,

66

prevailing civic engagement approaches can benefit from an approach to citizenship that considers whether citizens have the resources necessary to participate in the first place. Yet, the active, collective, participatory approach to citizenship should not be abandoned because, as Young highlights, the civil society offers a voice for the excluded, provides goods and services, and produces social innovation. The civil society is,

[A]n important site for the consolidation and expression of social perspectives. Organization and agitation in the public spheres of civil society, furthermore, are among the best methods of maintaining connections between representatives and constituents, and insisting that representatives be accountable. We deepen democracy when we encourage the flourishing of associations that people form according to whatever interests, opinions, and perspectives they find important. Strong, autonomous, and plural activities of civic associations offer individuals and social groups maximum opportunity in their own diversity to be represented in public life (Young, 2000, p.153).

Accordingly, civic engagement must be approached in a way that accounts for both citizenship status and practice. By drawing on feminist scholarship, the liberal and the civic republican approaches to citizenship may be reconciled. According to Lister, such a synthesis can:

[D]raw on both the liberal formulation of free and equal rights-bearing citizens and the richer republican conceptualization of active political participation and civic engagement (but based on a radical, pluralist reframing of the ‘common good’). A reading of the literature suggests that, while the rights and participatory approaches to citizenship remain conceptually different, they do not necessarily have to conflict; instead, they can be seen as mutually supportive, even if tensions between them remain. (Lister, 1997b, p.34).

Of course, as subsequent chapters show, a more comprehensive approach to citizenship also requires a broadened understanding of what constitutes ‘civic activity,’ for instance by considering the impact of unpaid care work on civic engagement, and certain forms of care or child-related networks as civic engagement.

67

In light of the centrality of civic engagement to democracy, and the many benefits individual and collective benefits that accrue from active communities, paying attention to the factors that shape associational life is vital. This chapter has shown that a gender gap in participation and trust exists which puts women at a disadvantage. Yet the dominant civic engagement literature does not recognize such inequalities. Rather, because it begins from a narrow understanding of citizenship, one that is grounded in formal equality, it presumes that all in the community have equal rights and entitlements as well as the same capacity to participate. Instead, a feminist approach to citizenship, rooted in substantive equality, is proposed because it accounts for difference as well as the social and economic conditions which constrain the civic life of disadvantaged groups, such as women. The following chapter takes a more empirical look at the civic life of men and women in order to better understand the extent of gender differences, and visits key explanations for gender gaps in participation.

68

Chapter 2 Exploring the Roots of Gendered Civic Life

The previous chapter drew on the case of women to highlight the need for an approach to civic engagement which begins from an understanding of citizenship that accounts for differences between individuals in the community, and pays attention to how social and economic conditions affect associational involvement. Using this as a foundation, the present chapter reviews and tests theories which seek to explain women’s lower levels of networking and trust compared to men. The findings support both the substantive and methodologically-based explanations for the gender gap in participation, and confirm that social and economic factors linked to gender roles put women at an associational disadvantage. Specifically, a deficit in resources, a tendency to associate with others who are similar to one’s self, a lack of interest in and knowledge about politics, and the methodological issue arising from a tendency to overlook certain types of social ties such as parent-, child- and care-related groups - all help explain women’s lower levels of participation compared to their male counterparts. However, despite the lingering gender gap in civic engagement, there is no evidence to support the notion that women’s primary caregiving duties orient them away from politics and the community; in fact, women with children had higher levels of civic engagement than non-mothers.

Women in Western advanced democracies have made great advances over time, gaining, for the most part, social, political and economic citizenship rights. In light of this, why is it that a gender gap in participation and trust endures? The dominant civic engagement literature holds no answers to this because their narrow understanding of citizenship, grounded in formal equality, does not lead them to ask such questions in the first

69

place. Their lack of consideration for difference and social and economic conditions, has led to a scholarship that presumes individual behaviour and norms are entirely self-determined, and that citizens exist in a conflict- and constraint-free world. To fill this void, a literature has developed by scholars interested in gender, race and class that studies variation in civic engagement across various groups in society, and underscores the connection between inequality and participation. This scholarship recognizes that disparity is not arbitrary, but disproportionately affects those who are socially and economically disadvantaged. Such an approach sheds light on the way gender roles are perpetuated in society, and highlights how they are reinforced by institutions, as will be shown in subsequent chapters. Informed by this alternative literature, four theories offering explanations for the gender gap in civic engagement are examined, drawing once again on the ISSP Social Networks II (2001), and

Role of Government IV (2006) surveys, as well as the World Values Survey (2000), across over a dozen countries.1 Although each gender gap theory is different, they all illuminate the intricate relationship between women’s citizenship status and participation, and underscore how socially and economically structured associational life actually is.

RESOURCE-BASED ACCOUNTS

The first theory for the gender gap in participation focuses on the way social cleavages such as gender, race or class are linked to the uneven distribution of resources such as time, money, knowledge and skills which facilitate civic involvement, and explores

1 The ISSP Social Networks II (2001) data includes 13 countries: Canada, Australia, Austria, Italy, Denmark, Finland, France, Germany, Norway, Great Britain, US, Spain, New Zealand. The ISSP Role of Government IV (2006) data includes 14 countries: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Portugal, Spain, Sweden, Great Canada, Australia, Denmark, Finland, France, E and W Germany, Netherlands, Norway, Sweden, Great Britain, US, Portugal, Spain, New Zealand. The World Values Survey (2000) data includes 14 countries: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Portugal, Spain, Sweden, Great Britain and the US.

70

how this shapes opportunities, capacities and willingness to be active (Norris and Inglehart,

2006 p.74; Inglehart and Norris, 2003, p.88-89; Burns, Scholzman and Verba, 2001, p.33).

For example, women may be less engaged in the community because they are struggling to juggle their paid and unpaid work duties, leaving little time or energy for anything else – such as joining a club. According to this approach, women network less than men because they can’t (Inglehart and Norris, 2003; Verba, Scholzman and Brady, 1995; Burns,

Scholzman and Sidney Verba, 2001).

Figure 2.1 Gender-Differences in Socio-Economic Status 2%

0%

‐2%

‐4% ‐3% ‐4% ‐4% ‐6%

‐8%

‐10%

‐12%

‐14%

‐16% ‐15% High Income Employed Post‐secondary Degree Upper/Middle‐Class Socio‐economic Status

Source: World Values Survey, (2000). Note: Z-test scores show the differences between men and women are all statistically significant at the .05 level.

Figure 2.1 examines gender differences for women across a number of common socio-economic indicators – all of which turn out to be statistically significant. Negative coefficients denote that women have a lower socio-economic status than men, positive coefficients indicate that women have a higher socio-economic status than men, and 0%

71

means men and women are equally involved. Figure 2.1 shows that fewer women hold a post-secondary degree or diploma compared to men, resulting in a gender difference of 3%.

Fewer women also belong to the higher/middle income bracket, as well as the upper/middle class compared to their male counterparts leading to 4% gender differences. Finally, figure

2.1 indicates that fewer women are employed, resulting in a 15% gender difference. Thus, the findings confirm what scholars have already shown elsewhere; that women find themselves continue to face socio-economic disadvantage compared to men.

Scholars have found support for the notion that socio-economic disadvantage translates to an associational disadvantage because the ‘costs’ of participation are too high; civic and political activity require basic resources such as time and money that are in short supply, particularly for women (Das, 2004, p.31; Verba, Scholzman and Henry E. Brady,

1995; Burns, Scholzman and Verba, 2001). This is reflected, for instance, in studies which reveal that the decline of participation in America is disproportionately borne by the lower class (Wuthnow, 2002, p.81-2; Schlozman, Brady and Sidney Verba, 1997, p.2-3). An insufficiency of resources can also shape the sorts of networks and voluntary activities women choose to engage in. Scholars such as Lowndes, for instance, argue that women who lack certain resources may be forced to decline from engaging in more active positions of responsibility within the associations they join (Lowndes, 2004; Verba, Scholzman and

Brady, 1995, p286). Adkins makes a similar connection between resources and participation, arguing, “… in contrast to men, women do not usually belong to the kinds of networks that bring economic advantage and… female networks generally command fewer economic resources and rely more frequently on time and non-monetized labour exchanges” (Adkins,

2005, p.200). Scholars also suggest that networks may be compromised if members do not

72

have access to adequate resources because this can lead them to default on their commitments to other members in the network, rendering norms of reciprocity obsolete. An extreme, but very real example is useful in illustrating this point. Based on interviews conducted in two Indian rural areas in Orissa, located in Eastern India’s poverty belt, Das exposes the high level activity that transpires between poor and politically disempowered villagers. This analysis reveals that the very poor cannot effectively benefit from their social networks and when they do, it is only on a temporary basis, to ‘get by’ (Das, 2004, p.34).

Networks between the very poor serve as stop-gap measures in times of dire need. Those who are so poor they cannot spare any rice for hungry neighbours will find it impossible to get help from those neighbours when they, in turn, run out of food on another day (Das,

2004, p.34-5). Those who are too resource-poor to contribute equally to a group or are unable to reciprocate in the network fail to build trust. In such cases, networks become a last-ditch source of short-term assistance, and production of social capital, for instance, can be thwarted. Scholars have arrived at a similar line of reasoning in the case of women’s networks in advanced democracies (Bruegel, 2005; Lowndes, 2004). For instance, Bruegel notes:

[Social capital] provides differential access to hierarchies of influence much in the way differential ownership of financial capital does. So the impact of women’s community organization tends to get restricted to system maintenance, keeping heads above water, rather than effecting change in the underlying circumstances of the community, partly because women tend to get represented by men the higher ‘up’ the system goes (Bruegel, 2005, p.9).

In their study on participation in America, Burns, Scholzman and Verba confirm that the intersection of gender, race and class has, over the span of one’s life, an important cumulative effect on participation. They find that resources, such as money and skills beginning in childhood and accruing over time serve to explain differences between men

73

and women, as well as between different types of women, based on race and class differences. Different exposure to politics in the home growing up and educational backgrounds, lead to different positions in the workplace and in other institutions, such as voluntary, political and religious organizations; these lead to differences in income levels, the acquisition of civic skills, and opportunities for recruitment into a multitude of networks of socialization (Burns, Scholzman and Verba, 2001, p.274-306).

Figure 2.2 Voluntary Membership by Gender, Controlling for Education Level 90% (%)

in 80% Organizations

70% Membership

Volountary 60%

50%

40% Men Women

30%

20%

10%

0% Incompleted Completed Incomplete Complete Some university University with elementary elementary secondary secondary without degree degree

Gender, and Controlling for Level of Education

Source: World Values Survey, (2000). Note: The relationship is statistically significant at the .05 level, except at for the values: incomplete elementary, incomplete university and graduated from a post-secondary institution.

According to this view, those with fewer resources may, in turn, reap fewer benefits from their networks, whether this is in terms of having access to information, developing

74

knowledge or skills, affecting change in their community, or being able to effectively voice their preferences or interests to decision-makers. This is why civic engagement is often said to exist in a vicious or virtuous circle; compared to those in the lower socioeconomic strata - where many women find themselves, those with greater resources are not only wealthier and better educated, but they are also more engaged in types of activities that help them learn new information and gain skills. This is, at least in part, reflected in the findings by Hall,

Putnam and Scholzman, who suggest that the growing number of women that are active in the labour market are more civically engaged than homemakers, attributing the difference to the advances women have made in educational attainment over time (Hall, 1999;

Schlozman, 2000; Putnam, 2000).

Gender differences in voluntary membership can be examined more closely with the use of elaboration analysis which can tell us about the effect of a control variable on the civic engagement of men and women. It is immediately evident, from the direction of the lines in figures 2.2 and 2.3 that those with a lower level of education and socio-economic status suffer from lower levels of organizational involvement. In fact, a similar trend is present in an analysis elaborating on the relationship between gender and voluntary activity levels, not included here. The analysis in figure 2.2 specifies that, among those who have completed elementary school, and attended or completed secondary school, gender is a good predictor of membership in voluntary organizations, with men continuing to lead in this area. However, gender does not explain organizational involvement for those who have not completed primary school, or have attended or graduated from a post-secondary institution.

Here, differences between men and women narrow and become statistically insignificant. In other words, irrespective of gender, those who have little education, and those who have

75

higher education will have lower and higher levels of networking respectively. Moreover, the results outlined in figure 2.3 indicate that gender is a good predictor of voluntary membership levels among all socio-economic ranks – except the higher ‘upper/middle-class’ level, associated with higher levels of networking, where gender differences narrow and become statistically insignificant. Thus, this analysis not only demonstrates the impact of socio-economic factors on civic engagement, but also highlights the implications this has in terms of differences between men and women, as well as between different types of women

(higher educated and higher income versus lower educated and lower income), based on

SES and educational achievement.

Figure 2.3 Voluntary Membership by Gender, Controlling for Socio-Economic Status 70% (%) 60%

in

Organizations 50%

Membership

Voluntary 40%

Men 30% Women

20%

10%

0% Manual Workers, Middle, Manual Workers Middle No Manual Upper/Upper Middle Class Unskilled, Unemployed Workers

Gender, and Controlling for Socio‐economic Status

Source: World Values Survey, (2000). Note: The relationship is statistically significant at the .05 level, except for the value: Upper/middle class level.

76

In sum, the findings highlight the impact of the intersection of gender and class. It shows that organizational involvement increases sharply as education and SES rises – two areas in which, on average, women lag behind men as indicated in figures 2.1 - 2.3. Indeed, considering statistical significance indicates that gender is a good predictor of membership in voluntary organizations for those who have completed elementary, secondary and post- secondary school, and for manual and unskilled workers, middle manual workers, middle non-manual workers, and the unemployed. Irrespective of gender, however, those with little education demonstrate the lowest levels of participation, while those with higher education and SES benefit from the highest levels of organizational involvement.

In addition to the effects of education and SES, Burns, Scholzman and Verba find

that the effect of cumulative advantages also explain significant differences in civic

activity between white, African American and Latina women in America (Burns,

Scholzman and Verba, 2001, p.274-306). They assert,

Our most general finding is that groups differ less in what they do with the participatory factors than in whether they get them in the first place. We have seen that disparities in participation... [in the] United States have their roots in social processes, many of which are based in the non-political institutions of adult life, that structure differently the experiences of groups defined by gender and by race or ethnicity. These processes leave in their wake individual and group differences in the factors that foster activity and a net advantage to Anglo-White men. Once members of different groups have acquired participatory factors, there is more similarity than difference in their ability to translate these factors into political activity. Our analysis has thus shown what it is about being non-male or non-Anglo-White, or both, that results in disparities among groups in political participation (Burns, Scholzman and Verba, 2001, p.303-4).

According to these findings, and the resource-based theory’s logic, race or ethnicity should have a negative effect on gendered participation because of the cumulative disadvantages related to being a non-white woman (Burns, Scholzman, Verba, 2001). Yet,

77

when the impact of race is tested on gendered civic engagement, we get unexpected results.

Figure 2.4 shows the results of an elaboration on the relationship between gender and

voluntary membership, controlling for race. This particular analysis only includes Canada

and the US because these are the only Western advanced democracies which participated in

the civic engagement questions which also happen to have answered the ethnicity question

in the survey. Furthermore, since the large number of ethnic categories have too few cases,

these have been recoded into a race dummy variable (e.g. white/non-white).

Figure 2.4 Voluntary Membership by Gender, Controlling for Ethnicity

Non‐white/Other 88% 82%

Male Female

White/Causasian 81% 78%

0% 20% 40% 60% 80% 100% 120% 140% 160% 180%

% of Voluntary Membership

Source: World Values Survey, (2000), Canada and US only. Note: The relationship is statistically significant at the .05 level for the Whites category, but is not statistically significant for the Non- whites group.

A bivariate analysis of gender and voluntary membership in Canada and the US (not

included here), reveals that 82% of men say they belong an organization, compared to 78%

78

of women, a 4% statistically significant difference.2 When controlling for race, we see in figure 2.4 that, overall, non-white respondents are more involved in voluntary organizations than their white counterparts, for both men and women. Irrespective of their race, men continue to have the highest levels of voluntary membership. Moreover, this analysis reveals a specification; while gender appears to be a good predictor among whites, with almost no change between men (81%) and women (78%) from the original relationship before any controls were added, the relationship becomes statistically insignificant for non-white respondents. In other words, gender does not explain differences in participation when non- white individuals are involved. Yet how can it be that non-white respondents generally network more than white respondents to the point that it renders gender insignificant? Such findings contradict the literature and the expectations of the resource-based theory which contend that, since inequality is a cumulative process, those who are non-white and female should have withstood the greatest amount of disadvantage over their life, resulting in the lowest socio-economic status, and should therefore network the least (Verba, Schlozman and

Brady, 1995; Burns, Schlozman and Brady 2001).

Drawing on a qualitative, ethnographic approach to the study of racial of ethnic communities, however, reveals a different and more nuanced civic engagement story. As

Haynes and Hernandez (2008) show, highly vibrant communities can form and thrive under pressure, oppression and segregation. Their study of the Black community in Runyon

Heights in Yonkers, New York demonstrates that exclusion and disadvantage can combine to act as a catalyst for social solidarity and collective action. Haynes and Hernandez (2008) chronicle the physical (racially-biased mortgage practices which led to all-black residential

2 The significance of differences is measured using z-test scores (.001 level).

79

neighbourhoods), social (networking along racial lines) and psychological (stigmas and stereotypes) exclusion of the Black members of Runyon Heights over time. According to

Haynes and Hernandez (2008), social isolation led Black community members to invent a social world that was theirs and triggered the creation of a highly motivated, cooperative and active Black community, including church groups, social and leisure clubs, as well as civic and political organizations (p.68). One resident who grew up in Runyon Heights described the community:

...[I]t had a link I think through our morals and our values again because you know it was kind of a place – where everybody knows everybody else as a child. Everybody else, at least back when I was growing up, is your supervisor or guardian and you were a little more restrained about doing things because you always kind of had eyes on you all the time. And I’d like to feel that in other communities that might not have been the case.... (Haynes and Hernandez, 2008, p.69)

The account of Runyon Heights suggests that it should not be considered an anomaly to find higher levels of networking by non-whites over whites, as shown in figure 2.4, since Black communities, for example, have long been highly active. However, Haynes and Hernandez’s

(2008) study also underscores the importance of resources required for participation and effective mobilization. In the case of Runyon Heights, they reveal that irrespective of the amount of bonding-type networking between Black community members, organizations were only truly effective at promoting the needs and interests of the community when they forged bridging ties with white community members and leaders – where the resources were located (Haynes and Hernandez, 2008).

80

Figure 2.5 Education Level for Men and Women, Controlling for Race

80%

70%

60%

50%

40% White Male White Female 30% Non‐White Male Non‐White Female 20%

10%

0% Incomplete Complete Incomplete Complete Incomplete Completed Elementary Elementary Secondary Secondary Post‐SecondaryPost‐Secondary

Education Level

Source: World Values Survey, 2000, Canada and US. Note: The relationship is only significant for white male and female (at the .05 level).

Figures 2.5 and 2.6 show the results of analyses on the same countries for the same wave of the World Values Survey as the analysis in figure 2.4. These particular analyses, however, examine the relationship between gender, income and education, controlling for race. Based on the literature which highlights the social and economic disadvantaged experienced by women and non-white groups (Verba, Schlozman and Brady, 1995; Burns,

Schlozman and Brady 2001; Adkins, 2005; Haynes and Hernandez, 2008), we should expect to see higher levels of education and income for white men. For education, men and women generally follow a similar pattern, irrespective or race, with the majority saying they have completed post-secondary education. In this category, however, non-white males and

81

females score higher than white males and females, contrary to what might be expected from the literature. When it comes to income levels, the downward slopes for non-white male and female are more in line with the literature, as is the downward slope for white women. Here, white men lead in high income levels, followed by white women, non-white males, with non-white women coming last. What is particularly telling from this analysis, however, is that gender turns out to only be a good predictor of income and education levels for whites, becoming statistically insignificant for non-whites. Thus, these findings show that a complex relationship between gender, race and class exists which is not always captured in simple gender analyses. In some cases, considering race can render gender inadequate in explaining differences between individuals in terms of socio-economic status.

Figure 2.6 Income Level for Men and Women, Controlling for Race 50%

45%

40%

35%

30%

White Male 25% White Female

20% Non‐White Male Non‐White Female 15%

10%

5%

0% Low Medium High

Income Level

Source: World Values Survey, 2000, Canada and US. Note: The relationship is only significant for whites (at the .05 level).

82

This section has found support for the resource-based account of the gender gap in participation. Women’s socio-economic disadvantage translates into a deficit in many of the resources needed to participate. As feminist scholars suggest, it seems that women’s enduring role as primary caregivers coupled with labour market activity can tap them of time and energy, pulling them away from civic activity in a way that is different from men - especially, when young children are involved (Burns, Scholzman and Verba, 2001; Moore,

1990; Munch, McPherson and Smith-Lovin, 1997). This last point is further explored in the subsequent section. POLITICAL ORIENTATION ACCOUNTS

The second theory used to explain gender differences in networking focuses on the psychological dimension, and the way values and attitudes shape behaviour. This approach suggests that people join networks because they are politically oriented or knowledgeable, predisposing them to participate (Norris and Inglehart, 2006, p.74; Inglehart and Norris,

2003, p.91-92). For instance, women who are mothers of young children may be less involved in the community because their role as primary caregiver shifts their focus or mental energy away from politics (Burns, Scholzman and Verba, 2001, p.7). According to this view, women network less than men because they won’t (Inglehart and Norris, 2003;

Verba, Scholzman and Brady, 1995; Burns, Scholzman and Verba, 2001). It is worth mentioning that some studies suggest that gender gap in political knowledge is largely the result of bias, with men over-reporting knowledge, for instance, because they are less likely to choose ‘don’t know’ on a survey, compared to women (Mondak and Andersen, 2004;

Dolan, 2011). However, even when accounting for bias, such findings often fail to fully explain away gender differences.

83

For example, analyses of gendered networking patterns have shown that the opportunities and constraints stemming from differing positions in the work force, as well as in marital roles and parenthood, can explain variations in the associational life of men and women (Moore, 1990, p.727. Munch, McPherson and Smith-Lovin, 1997). Scholars have found that men and women’s networking patterns change significantly at certain points in time, such as with the presence of small children (Munch, McPherson and Smith-Lovin,

1997). Childrearing, particularly with preschool aged children, can work to constrain women’s ability to engage in associational life by diminishing the number of people with whom they interacted (Munch, McPherson and Smith-Lovin, 1997, p.509). While parenting affects the networking habits of both men and women, it seems to affect women’s associational life more deeply and for a longer period than men, most likely because of women’s role as primary caregivers. Thus, women’s roles as mothers, “…and society’s definition of childrearing as a female activity place women and men in different structural positions with respect to the flow of information and other resources in social networks. This difference may account for many of the disparities between men’s and women’s outcomes over the life course” (Munch, McPherson and Smith-Lovin, 1997, p.518). As primary caregivers, women may concentrate all their attention on family matters, orienting them away from political or civic matters. As Burns, Scholzman and Verba, state, “raising children so absorbs available mental energy that mothers, especially of toddlers under foot, are too preoccupied at home to pay attention to politics” (Burns, Scholzman and Verba,

2001, p.7).

However, not all scholars subscribe to this view. Herd and Meyer as well as

Lowndes, for instance, suggest that just as care work can act as a hindrance to women’s

84

associational life, it can also act as a catalyst. According to this view, play-dates, parent’s groups such as the PTA, and children’s activities can lead parents to forge new, lasting, reciprocal networks (Lowndes, 2006; Herd and Meyer, 2002). The effect of family status on voluntary membership levels for men and women is tested in an elaboration analysis, and the results outlined in figure 2.7. As expected, this analysis shows that, even controlling for family circumstance, men continue to hold the higher scores for voluntary membership levels. However, we also see an increase in organizational involvement when controls are added, with a 4% advantage for men, and a 5% advantage for women with children compared to without any controls. Among the greatest and statistically significant gender differences, married women without children are 7% less likely than married men without children to belong to a voluntary organization, and unmarried women without children are

6% less likely than unmarried men without children to belong to voluntary organization.

Overall, those with the highest score for holding memberships in any organization are married men (66%) and married women (62%) with children. The second highest score for men comes from those that are married without children (63%), while the second highest score for women comes from those that are unmarried with children (59%). This indicates that, contrary to what might be expected based on the work of Munch, McPherson, Smith-

Loving, and more in line with what scholars who argue for an approach to civic activity that includes measures for child-related networks propose, such as Lowndes and Herd and Meyer

– women with children appear to be more involved in associational life than women without children. Thus, as Figure 2.7 demonstrates, differences are not only present between men and women, but also between women with and without children.

85

Figure 2.7 Voluntary Membership by Gender, Controlling for Family Status 70% 66% 62% 62% 63% 59% 60% 60% 57% 54%

50%

40%

% of Men who Belong to any Voluntary 30% Organization % of Women who Belong to any Voluntary Organization 20%

10%

0% Unmarried with Unmarried without Married with Children Married without Children Children Children

Family Status

Source: World Values Survey , (2000). Note: The relationship is significant, except for the value: Married with children.

Moreover, figure 2.8 shows the results of an analysis of various indicators for political orientation by gender, including measures for interest in politics, and political efficacy. Fewer women follow politics in the news compared to men, resulting in a 9% gender gap, and fewer women feel they are well informed, leading in an 11% gap with men.

In addition, fewer women say they have a good understanding of political issues and are interested in politics compared to men, resulting in 15% gender differences. Thus, these findings reveal that sizeable and statistically significant gender gaps in political orientation exist – in accordance with the orientation theory.

86

Figure 2.8 Political Orientation by Gender

33% I'm no less informed than other people 44%

44% Has a Good Understanding of Political Issues 59%

23% Average Citizen Influences Politics 24%

30% % Women People Like Me Have a Say in Politics 31% % Men

38% Interest in Politics 53%

74% How Often Follows Politics in News 83%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

% of Political Orientation by Gender

Source: Interest in Politics and How Often Follow Politics in News from World Values Survey,(2000). All others from ISSP Role of Government IV, (2006). The 14 countries in the analysis include: Australia, Canada, Denmark, Finland, France, Germany, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Great Britain, United States. Note: Z-test scores show the differences between men and women are statistically significant at the .05 level, with the exception of average citizen influences politics, and people like me have a say.

Interestingly, gender differences narrow and become insignificant with political efficacy

indicators, suggesting that both men and women seem to believe that the average citizen

influences and has a say in politics. Thus, perhaps gender differences in civic engagement

are best explained by indicators measuring political interest, rather than beliefs about

efficacy.

That women tend to score consistently lower in political orientation measures has

been interpreted as a possible explanation for the gender gap in participation because if

women are less knowledgeable about and interested in politics, they will be less likely to

want to participate (Burns, Scholzman and Verba, 2001). Thus, despite some confusion

about the exact causal order, there appears to be consensus that some degree of

87

psychological involvement in politics is tied to participation (Verba, Scholzman and Brady,

1995, p.344).

Figure 2.9 Voluntary Membership by Gender, Controlling for Interest in Politics 80%

70%

60%

Organizations 50%

40% Voluntary % Men in

30% % Women

20% Membership

%

10%

0% Not at all interested Not very interested Somewhat interested Very interested

Gender, Controlling for Interest in Politics

Source: World Values Survey, (2000). 13 Countries include Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Portugal, Spain, Great Britain and the US. Note: The Cramer’s V indicates that the relationship is statistically significant at the .05 level, except for the values: Not very interested, and somewhat interested.

Looking at the direction of the lines in figures 2.9 and 2.10, one can instantly see that, for both men and women, membership levels generally rise as interest in politics and following politics in the news increases. Figure 2.9 demonstrates that gender is a good predictor for voluntary membership for those that are either ‘not at all interested’ or ‘very interested’ in politics, with men leading in associational involvement in both categories. Men and women’s voluntary membership levels converge among those who said they were not very interested in politics, and are varied among those who are somewhat interested in politics, where women take a slight but statistically insignificant lead.

88

Figure 2.10 Voluntary Membership by Gender, Controlling Following Politics in the News 70%

60%

50% Organizations

40% Voluntary

in % Men 30% % Women

20% Membership

%

10%

0% Never Less often Once or twice a week Daily‐several times a Gender, Controlling for How Often Follow Politics in the News week

Sources: World Values Survey, (2000). 14 Countries include Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Portugal, Spain, Sweden, Great Britain and the US. Note: The relationship is statistically significant at the .05 level for the value ‘Daily-several times a week’ – all other values are statistically insignificant.

As figure 2.10 highlights, among those who follow politics in the news several times a

week, gender proves to be a good predictor of organizational involvement. Here, fewer

women who say they follow politics in the news frequently belong to a voluntary group

compared to men, resulting in a 4% gender difference. Men and women’s participation

levels converge among those who never follow politics in the news, or do so less often or

only once or twice a week.

Thus, the findings show that organizational involvement increases consistently as

interest and knowledge of politics rises – areas in which, on average, women tend to lag

behind men, as figure 2.8 demonstrates. Indeed, gender proves to be a good predictor of

89

voluntary membership levels for those who are not at all interested and very interested in politics, as well as for those who follow politics in the news on a daily or weekly basis.

Ultimately, the findings indicate that women are generally less plugged into politics than men, and this appears to contribute to gendered patterns of civic engagement.

AGENCY ACCOUNTS

The third theory looking to explain differences in gendered networking suggests that the social world that individuals belong to has a deep effect on their behaviour. In particular, it begins from the premise that people are pulled into networks through their kin or friends who invite them to join (Burns, Scholzman and Verba, 2001, p.33). Thus, women may be less likely to be involved in voluntary associations, for example, because they are surrounded by family and friends who do not tend to belong to diverse groups and, in turn, do not recruit them as members. According to this view, women network less than men because nobody asked them (Burns, Scholzman and Verba, 2001, p.33).

One cross-national civic engagement study, for example, found that informal networking with family, more common among women, is negatively correlated with associational membership, while informal time spent with co-workers, more common among men, is positively correlated with membership in formal associations (Norris and

Inglehart, 2006). In other words, kin and close-family ties, favoured by women, tend to lead to limited involvement in other organizations. Findings from several single-country studies supports this description of women’s network homophily (Munch, McPherson and Lynn

Smith-Lovin, 1997; Popielarz 1999; McPherson, Smith-Lovin and James M. Cook, 2001;

Gidengil, Elizabeth et.al, 2006). As figure 2.11 shows, women do lead over men in informal networks with family, with a 7% statistically significant advantage.

90

Figure 2.11 Frequent Networking with Kin and Extended family 88% 87%

86% kin 84% with

family

82% Men frequently

extended Women

visit

and 80%

who 80%

%

78%

76% Men Women

Source: ISSP Social Networks II, (2001). Note: Z-test scores show the differences between men and women are statistically significant at the .05 level.

Moreover, figure 2.12 lends support to what Norris, Inglehart, and others have found – that those who spend less time with family tend to be more civically engaged, in this case, with higher levels of voluntary membership for both men and women alike. Gender only appears to be a good predictor for those who frequently network, moreover, with men showing a 8% statistically significant lead over women. However, the relationship becomes insignificant when controlling for those who rarely or do not visit with family, suggesting that, in such cases, gender is not a good predictor for voluntary membership levels.

Thus, findings show that women lead in familial ties, a form of networking that is associated with lower levels of civic engagement. Tight, close-knit relationships, such as family ties, that do not bridge or cut-across other types of groups and memberships, are said to offer women narrow experiences. Such restricted, bonding-type of networks, for instance,

91

are found to limit the possibility for significant returns such as information or learning new skills.

Figure 2.12 Membership in Voluntary Organizations by Gender, Controlling for Frequent Visits with Kin and Extended family

82% 81%

80%

78% 77% Organizations

76%

74% 73% Voluntary Men in

72% 71% Women

70% Membership

% 68%

66% Rarely or never visits with kin and extended family Frequently visits with kin and extended family

Gender, and Controlling for Frequent visits with Relatives

Sources: ISSP Social Networks II, (2001).3 Note: The relationship is statistically significant at the .05 level for the value ‘Frequently visits with family’ but not significant for ‘Rarely visits with family.’

The consequences of this have been underscored by a Canadian study that uncovered a relationship between women’s preference for tight-knit, homogenous networks, and their likelihood to know less about politics than their male counterparts (Gidengil et.al, 2006, p.264-65). Moreover, in another study of gendered participation in Canada, Erickson explains:

3 Since the question of informal family networks in the World Values Survey is only included for the surveys of three Western advanced democracies (Canada, Spain and the US), I turn to the ISSP Social Networks II (2001) data, which has extensive family networks indicators for 12 countries (Australia, Germany, Great Britain, the US, Austria, Italy, New Zealand, Canada, Spain, France, Denmark and Finland). As such, a different voluntary membership index is built and used here. The voluntary membership index built from the ISSP (2001) data includes 7 voluntary membership categories: political party, trade union or professional association, church or religious organization, sports group, charitable organization, neighbourhood group, and the category ‘other.’ The alpha score for the index is .66. The scaled index was recoded into a dummy.

92

Men have more strategic locations in social structure on the whole: they are better placed to meet many others, and to enter their networks. Men seem to be more strategically located in many social spheres, since involvement in most of them adds slightly more to diversity of contacts to men than to diversity of contacts to women. Being so pervasive, the male advantage entering networks would be hard to alter. And being so pervasive, it has a strong cumulative effect: both men and women have more diverse ties to men than to women (Erickson, 2004, p.48).

In addition, Popielarz warns that one consequence of this gender-based homophily is that it can perpetuate segmentation in women’s mobilization. Women’s groups, she asserts, may find it increasingly difficult to recruit members that reflect different age cohorts, social classes, races and ethnicities, as such individuals do not tend to be integrated in voluntary organizations in the same way (Popielarz, 1999, p.245-6). Popielarz suggests that such cleavages can affect what issues are – and, more importantly, are not - at the forefront of the women’s movement’s agenda.

METHODOLOGICAL ACCOUNTS

Finally, the fourth explanation for inequality in participation between men and women suggests that women are active, but no one is paying attention. This approach tends to highlight modes of networking and activity that are typically discounted in the dominant civic engagement literature. This view proposes that women may seem less active in the community – but this is only because their preferred affiliations are deemed to fall outside of what ‘counts’ as civic activity (Stolle and Micheletti, 2006, p..64; Lowndes, 2006).

According to this approach, women appear to network less than men because their networking is not being counted.

Feminist scholars argue that the gender-bias inherent in the liberal and civic republican citizenship traditions is reflected in civic engagement methodology. The liberal individualism present in prevailing civic engagement studies, for instance, tends to reduce

93

categories of difference into independent variables - fixed and unchanging, and ill-equipped to capture the complexities of identity because they are founded on the false assumption of equal agency (Junn, 2007, p.130; García Bedolla and Scola, 2006; Bruegel, 2005, p.8-9). As a result, dominant civic engagement approaches fail to accurately capture the importance of identity to social relations, theorizing agents as abstract, rational actors, and epitomizing the masculine actor which feminist theorists have long lamented (Adkins, 2005, p.198). Bruegel asserts:

… [T]he associations, organizations and the development of trust are modelled as outcomes of rational individual decisions, albeit in situations of limited information and acknowledged inter-dependence and potential altruism…. Putnam’s work, tends to be written as the aggregation of genderless [as well as raceless and classless] individuals in some geographical space or in some civic association. Thus, the power that white men may gain…through the exclusive white male clubs, the Elks, the moose, Lions and Knights of Columbus – that he takes as exemplars of social capital is ignored… (Bruegel, 2005, p.9).

Just as with the liberal citizenship tradition’s methodological individualism, feminists have highlighted the methodological consequences of the dominant civic engagement literature’s civic republican tendencies. Since the dominant civic engagement scholarship values those social arrangements which foster reciprocity, cooperation, and social cohesion in the ‘public space,’ in line with the civic republican citizenship tradition, it has tended to discount alternative forms of participation that are aimed at broad social change. In doing so, it has excluded a form of participation favoured by many disadvantaged groups, including women. The significance of the practical aspects of citizenship, not only as a source of strength, support, and solidarity in the struggle for equality, but also more broadly, as a basic expression of human agency, is recognized by feminist scholars (Lister,

1997a, 1997b, 2003). Dietz, for instance, proves to be an ardent proponent of an approach to citizenship which emphasizes participation, arguing that it is only: “‘through active

94

engagement as citizens in the public world’ and in the recognition of the activity of citizenship as itself a value that feminists will ‘be able to claim a truly liberatory politics as their own’” (Lister, 1997, p.33). However, in light of the historical limits on their civil and political rights, women have tended to gravitate towards alternative forms of participation to promote equality and influence public policy. Yet, as feminist scholars have highlighted, although such modes of participation are typically aimed at influencing decision-makers or policy change, and tend to mobilize citizens in the public realm, they are often disqualified from the list of acceptable social networks that ‘count’ as civic engagement because their aim to challenge societal norms threatens social cohesion. Accordingly, Mayer maintains:

This blind spot in the social capital perspective – the avoidance of adversarial movements – certainly has to do with the ambivalence of protest milieus which represent, on the one hand, social networks shaped by trust, cooperation and shared values, allowing them to unfold high levels of action, i.e. to mobilize ‘social capital,’ while on the other hand, towards the ‘outside’ world they articulate conflict and use disruptive action repertoires, thus challenging relations of trust and reciprocity (with other social groups) and ‘social capital.’ Movement groups thus pose, unlike bowling clubs or choral societies, the question ‘social capital for whom and to what end?’ (as well as ‘inclusion for whom?’) and thereby problematize a concept of civil society free of conflict and neutral about different interests. Not all movements fit into the peculiar understanding of a (harmonious) relationship between civil society and the state inherent in the social capital perspective (Mayer, 2003, p.118).

In light of the notion that alternative forms of participation are overlooked by the dominant civic engagement literature because they are not seen as building trust or social cohesion, figure 2.13 and 2.14 test whether potentially adversarial groups erode social trust. Figure

2.13 examines whether those who say they belong to a politically-inclined group have lower levels of trust in others, while figure 2.14 examines whether those who partake in various forms of political action have less trust in others.

95

Figure 2.13 The Effect of Involvement in Politically-Inclined Groups on Social Trust 50% 45% 45%

40%

35%

30% 27% 25% Trust 25% Social 20%

15%

10%

5% 3%

0% Don't belong Belong to political group Belong to a nonpolitical group Belong to both types of groups

Membership in Any Political and Non‐Political Group

Source: World Values Survey, (2000). Note: The relationships are significant at the .01 level.

While all forms of political action found in the survey are included here, two forms of action stand out as being the most potentially adversarial of all, namely ‘occupying a building or factory’ and ‘joining a demonstration.’ The results for figure 2.13 show that while the majority of respondents actually don’t think most people can be trusted, those who belong to non-political groups have the highest levels of social trust, while those who belong to political groups have the lowest levels of trust. Thus, the findings lend support to the notion that potentially adversarial, political-motivated groups erode social trust.

The result of an analysis on the impact of political action on social trust, outlined in figure 2.14, reveals that aside from signing a petition, most respondents exhibit low levels of trust in others. Although signing petitions – a political act which can be considered relatively

96

non-adversarial - is related with high levels of social trust, it is difficult to discern a clear pattern about how forms of political action shape this norm. On the one hand, a political activity which may be considered relatively non-adversarial such as boycotting is associated with low levels of trust; and striking, which usually involves working alongside a collectivity such as a union, is found to have no effect on trust. On the other hand, other forms of political action which would intuitively be considered much more individualistic and adversarial in nature, either have no significant effect on trust, as is the case with occupying a building or factory, or are associated with moderate levels of trust, such as participating in a protest.

Figure 2.14 The Effect of Political Action on Social Trust

80% 71% 70%

60%

50% Trust 40% 34% Social 30% 22% 20%

10% 7% 4%

0% Have Occupied a Have Demonstrated Have Joined a Strike Have Joined a Boycott Have Signed a Petition Building or Factory

Political Action

Source: World Values Survey, (2000). Note: The relationships are significant at the .01 level, except for striking and occupying a building or factory. While more recent surveys have started to include voluntary activities such as local political action groups, signing petitions, and joining boycotts, they continue to overlook family- and care-related networks. This is perhaps not surprising in light of the dominant

97

civic engagement literature’s civic republican and liberal citizenship roots which apply the public-private divide, locating civic virtue and politics outside the home, care work and family. Although he points to the significance of family dinners in fostering social capital,

Putnam largely overlooks the family unit in his analyses. For instance, his Italian study is based on data in which 73% of local associations are sports clubs – popular among men, whereas just 1% are concerned with health and welfare – groups favoured by women. In a similar vein, Hall’s analysis of Britain elaborates on the popularity of pub attendance, but only offers a footnote on the rise in time spent on caregiving (Hall, 1999; Lowndes, 2006, p.215). In fact, the normative family structure, along with care or child-related networks, and their gendered implications, are never scrutinized. Women continue to be disadvantaged by a political and economic system which undervalues unpaid care work overwhelmingly performed by women, largely taking place in a context which privileges the traditional family unit where women are the ‘dependents.’

Others link dwindling levels of civic activity to a decline of , specifically with ‘deterioration of the family’ (Herd and Meyer, 2002, p.668). While both Coleman and

Bourdieu consider the family unit as important locations for the production of social capital, their approaches idealize the patriarchal family (Field, 2005, p.24-6; Edwards, Franklin and

Holland, 2003). Coleman’s work, for instance, has conservative implications since he attributes high levels of human capital generated by high levels social capital to nuclear families and tight-knit communities. According to this view, the increasing breakdown of families and communities signals the decline of both social and human capital for successive generations to come (Coleman, p.118). Ultimately, although the realm of family is not

98

completely overlooked, the care work disproportionately shouldered by women goes unnoticed in standard civic engagement analysis. Herd and Meyer insist,

[C]are work, performed mostly by women, has long provided the necessary support for other family members to maximize their own civic engagement. Families provide the basic necessities required to develop, and exercise, the rights and demands of citizenship… By not addressing or examining unpaid care work and how it affects civic engagement, the present-day debate about civic engagement is incomplete (Herd and Meyer, 2002, p.672).

‘Care work,’ according to Herd and Meyer, fits Putnam’s criteria for civic engagement, which include civic activity of a voluntary, unpaid, altruistic, nature, unattached to state functions, which breeds trust, reciprocity and social ties (Herd and Meyer, 2002, p.674).

Moreover, McClain asserts, “[f]amilies, as part of the ‘basic structure’ of society, have the important task of ensuring the ‘nurturing and development’ of children who, as citizens, must ‘have a sense of justice and the political virtues that support political and social institutions” (McClain, 2001, p.1624). Thus, the care work taking place in the private realm of the family acts as a training ground for citizenship by fostering human capital (the skills and capabilities that enable a person to act in new ways) in the next generation of citizens

(Coleman, 1998, p.S100; Paxton, 1999, p.100).

This failure to adequately account for the family unit and care or child-related networks may have a significant impact on associational life – not only in the forms of networks that women tend to belong to, but also in the forms of networks that ‘count’ as civic engagement. As Lowndes asserts:

Why are child-care networks not considered in analyses of social capital? The most likely explanation seems to lie with the enduring influence of the public/private split in the consideration of politics in general, and citizenship in particular… the cluster of activities, values, ways of thinking and ways of doing things which have long been associated with women are all conceived as outside the political world of citizenship and largely irrelevant to it (Lowndes, 2000, p.535).

99

What also goes unmentioned is the role that family and caring for children has played in catapulting women into civic activity and politics. Feminist scholars have documented how their traditional gender roles as mothers and wives led nineteenth century early women reformers in North America, as well as Western and Northern Europe, to join and form charities and voluntary associations, stimulated political awareness, developed civic skills, and were propelled into political life, promoting issues such as women’s rights and maternal health and playing a central role in the development of the welfare state

(Koven and Michel, 1990, 1993; Bock and Thane, 1994). To this day, scholars highlight the way women’s concern for the health and well-being of their family has led them to become civically involved, turning what began as informal kitchen table gatherings into formal associations, with working campaigns (Lister, 2005, p.22-3; Herd and Meyer, 2002, p.672-3;

Herda-Rapp, 2000). Women’s network activity may thus, in actuality, remain largely

‘undocumented.’ Important aspects of women’s civic participation is rendered invisible, overlooked by civic engagement scholars because it takes place inside the home, or is related to caregiving – realms not commonly understood or considered to have any significant bearing on associational life. This underscores the need to broaden the scope of what is defined as ‘civic engagement’ to include – or at the very least explore - sources of involvement in both the public and private sphere.

Interest is growing for a broadening of the understanding and measuring of

‘participation,’ such as the literature on ‘small democracy’ and political consumerism – outside of larger or more formal types of associations. For instance, Danish ‘everyday makers’ may not be engaged in ‘big democracy’ politics through traditional channels of power or formal organizations; their goal is less about government responsiveness, and more

100

about resolving the very real, tangible problems in their daily lives through interaction with other individuals and institutions (Bang and Sørensen, 1998, p.2). This view challenges the civic engagement approaches such as Putnam’s, whose “…spherical insulation of the political from the social tends to blind us to the democratic and life-political potential of engaging oneself with others for the political governance of everyday life… by perceiving unity, homogeneity and consensus as the constituting elements of civil society, the focus is removed from its political aspects: the handling of differences, diversity and dispute” (Bang and Sørensen, 1998, p.2; Bang, 2005; Siim, 2004).

Stolle and Micheletti make a similar argument in defence of expanding what ‘counts’ as civic engagement to include political consumerism: seemingly small but political acts that are manifested in the choices citizens make on a daily basis. Their study highlights that women tend to express themselves through acts such as political consumerism. However, surveys often miss such acts because their measurements are too narrow, or disregard these altogether. For example, the World Values Survey, used here and in the political consumerism study by Stolle and Micheletti,4 reveals that women do not boycott more than men (see figure 2.5), although over time, there has been a reduction in the gender gap. Stolle and Micheletti demonstrate the effect that different phrasings of interview questions have on findings, showing that while 17% of women interviewed in the twenty-one countries included in the 2002 European Social Survey say they have participated in a boycott, a more sizeable 26% of women also say they chose to purchase a product for some underlying ethical or political reason – also an act which qualifies as boycotting (Stolle and Micheletti,

2006, p.56-7). This demonstrates the difficulties encountered in trying to quantify what turns

4 Stolle and Micheletti (2006) actually draw from the European Social Survey (2002), but this dataset is used in the making of the integrated World Values Surveys.

101

out to be highly complex political acts in daily life. Ultimately, however, boycotting questions are rarely included in large cross-national surveys, and therefore miss an area of political action engaged in by women. In light of this, scholars concerned with gender differences in participation fault the dominant civic engagement literature’s preoccupation with formal organizations and overlooking “…forms of civic engagement related to the everyday lives of women and are very much immersed in daily activities like shopping…

[yet] political consumerism allows women to work on political issues in a way they find comfortable and natural from the perspective of their interests and roles in society” (Stolle and Micheletti, 2006, p.65).

Thus far, the findings point to a positive correlation between SES and participation.

This is exemplified in the case of women whose lower SES compared to men is tied to lower levels of civic engagement, in accordance with the resource-based theory. However, focusing on resources alone is not always enough, particularly when considering the intersection of gender, race and class. For example, the findings show that race explains away gender differences for non-white individuals when looking at key SES indicators, education and income. Moreover, despite the fact that non-white respondents have lower levels of income than their white counterparts, they display higher levels of education, as well as civic engagement – a finding which contradicts the literature, and for which a resource-based account alone cannot explain. What the findings suggest is that while inequality might make it difficult to participate or may reduce one’s networking opportunities in some contexts, it is also possible that it creates opportunities in others; in this case, the struggle for equality may propel non-white respondents into action. On the other hand, the case of Runyon Heights shows that networking alone is not always enough

102

when those ties are disproportionately homogenous and bonding. For civic engagement to be most effective, citizens require resources as well as bridging networks across communities

(Haynes and Hernandez, 2008; Green, Tigges, and Brown, 1995; DeFilippis, 2001; de Souza

Briggs, 2001).

The case of Runyon Heights also highlights the way in which space and limited resources are interconnected, and impede on the associational life of disadvantaged groups such as Black community members and women. Differential networking opportunity structures abound in relation to race, for example, particularly when combined with space

(or place). As the case of Runyon Heights shows, racial residential segregation is not only related to less frequent interracial (bridging) ties with other groups, but also limited resources (de Souza Briggs, 2001, p.10-13, 31). Together, these factors invite social ills such as crime, and generally work to cut non-white communities off from the fruitful contacts and opportunities which are beneficial to a healthy civic life:

Inequality is often spatial in America … we have structured unequal access and spatially biased decision-making into the way we run the welfare state and universal- access public services … Jobs, not just houses, have sprawled to the horizon, further isolating the disadvantaged from ladders to opportunity. Crime and its traumas are spatially concentrated. Space turns out to act as a signal of worth … For at least some employers, realtors, and others, ghetto is at once a place and a kind of person … We “are” where we live (de Souza Briggs, 2001, p.58-9).

Thus, the interaction between race and space leads to a ‘geography of opportunity,’ where physical segregation translates into a segregated associational life. In the case of women, space has also played an important role in structuring networking opportunities, although in this case the ‘space’ is not a specific geographic area but the private sphere of the home, tied to traditional gender roles. Over time, women’s location, work and social ties have been associated with domesticity, in opposition to the public sphere where civic norms and

103

behaviours are said to count. Despite women’s increased participation in the labour market and a history of activism, their roles as caregivers continues to place them at a social and economic disadvantage, leaving them with limited resources, and care-related associations continue to be discounted as a valid form of civic activity.

Empirical support was also found for the orientation account, with women being less likely than men to follow politics in the news, feeling they are less informed and have a worse understanding of politics, and being generally less interested in politics than their male counterparts. According to the orientation theory, it is women’s lack of interest in politics which translates to a lack of participation. However, the argument that women’s roles as primary caregivers and the presence of children likely shifts women’s attention away from politics was not confirmed. Specifically, women with children had the highest levels of civic engagement than those without children, and gender becomes completely insignificant in the case of those who are married with children. Women’s high level of kin- based networking, moreover, lends support for the agency account which contends that women are less active because the people they surround themselves with do not invite them to join other groups. Finally, the methodological explanation for gender differences in civic engagement suggests that certain types of organizations that women favour, including family and care-related groups, are missing from surveys. This is confirmed here by the sheer absence of such group-types in the ISSP and World Values Surveys. As such, a later chapter will outline the findings of personal interviews with sixty mothers across three

Western advanced democracies, delving into the role of family, children and care work, and how these shape their associational lives. Furthermore, an analysis of various forms of participation which may qualify as adversarial showed that politically-inclined groups, in

104

which women are very active, are associated with significantly lower levels of social trust.

The potential danger of groups associated with low-trust is that they can encourage intolerance of others outside the group, and erode the general social trust that is so necessary for a vibrant civil society. Associations such as the KKK, premised on prejudice and exclusion, can understandably be considered to have negative effects on social cohesion and civic life more broadly. However, the notion that political parties, as well as women’s, human rights, peace, and environmental groups, which do not promote intolerance but give individuals an outlet to voice their concerns and shape decision-making, could be unfavourable for democracy highlights the puzzling, sometimes counter-intuitive relationship that civic engagement theory posits between participation and trust.

Another analysis, of which the findings can be found in table 2.1, is conducted to test for the impact of various indicators that are said to shape gendered networking patterns, as outlined above. Ordinary least squares (OLS) regression offers a powerful way to take control variables into account. It allows researchers to test how independent variables explain change in the dependent variable. In regression analysis, Y is a function of X; that is, one unit of change in the independent variable (X) produces a one-unit change in the dependent variable (Y). In a multiple regression like the ones found in table 2.1, however, one unit of change in dependent variable Y is caused by one unit of change in independent variable X, controlling for the effect of all other independent variables in the model.

Thus, table 2.1 displays the results of three separate multiple OLS regressions with civic engagement as the dependent variable, testing for the impact of a number of independent variables that are known or expected to have an impact on associational life, as discussed above as well as in Chapter 1. Model 1 is run on both men and women, and serves

105

as a baseline to compare the effects of the other two models; model 2 is a regression which

only includes men, while model 3 is exclusive to women. The purpose of running separate

regressions is to distinguish and better account for the effect of gender. The dependent

variable, the civic engagement index, is a six-point scale which includes the volunteering,

membership, informal networking with friends, and political action index dummies, as well

as a dummy indicator for social trust – ‘most people can be trusted.5

Table 2.1 Predicting the Causes of Gender Differences in Civic Engagement: OLS Regression Results

CIVIC ENGAGEMENT

Variable Model 1 Model 2 Model 3 Gender Male Female B SE B Beta B SE B Beta B SE B Beta Structural /Resource- based Accounts Level of Education .18*** .01 .22 .21** .02 .22 .20** .02 .21 Income .12*** .02 .08 .12** .03 .10 .08** .03 .06 Employment Status .08* .03 .03 .19** .06 .07 .31** .06 .11 Socio-economic Status .06*** .02 .05 .01 .03 .01 .13** .03 .09 Family Status: Unmarried with Children .09 .05 .04 .14 .09 .05 .24** .09 .07 Married with Children -.02 .08 -.00 -.04 .14 -.01 .08 .15 .01 Married without Children .04 .04 .02 .09 .08 .03 .02 .07 .01 Gender -.08*** .03 -.03 ------

Cultural/Political Orientation Accounts Interest in Politics .25*** .02 .20 .34** .03 .25 .26** .03 .18 Follow Politics in the .04** .02 .04 .06 .03 .04 .05 .03 .04 News

. Religious Attendance 28*** .03 .11 .27** .06 .09 .43** .06 .15

R² .19 .23 .22 Sample Size 5491 2177 2006 Source: World Values Survey, (2000). Note: Unstandardized coefficients with standard errors and beta values. *** p > .000; ** p > .01; * P> .05. ** Significant at the .01 level; * Significant at the .05 level. Tolerance measures all above .20 indicating there is no collinearity problem.

For the structural account, independent variables in table 2.1 include a six-point SES scale, a

three-point income scale, an employment dummy, a gender dummy that is only included in

5 For more details on the indicators included in the civic engagement index, see chapter 2. The civic engagement index is built from World Values Survey indicators. This survey does not include measures for the vote. The alpha score for the 6-point civic engagement index is .51.

106

model 1, and a ‘family status’ set of four polytomous dummies, with ‘unmarried without children’ serving as the reference category. For cultural factors, independent variables include indicators a four-point scale for both following the news and interest in politics.

Finally, a religious attendance dummy variable is also tested for in light of its importance, highlighted in previous studies, and because women seem to favour religious organizations.

Indicators for the agency account of gender differences in participation are not included in this regression model because the World Values Survey question on networking with family is not included in a majority of the advanced Western democracies that are examined here.6

Moreover, the methodological account of the gender gap is more adequately explored in chapter 7, with the use of personal interviews.

The findings in table 2.1 indicate that, overall, structural and cultural factors matter, as does religious attendance. According to model 1, education, employment and income have significant and positive effects on civic engagement while the negative coefficient for gender indicates that women have lower civic engagement than men. Moreover, cultural factors matter as well, with both ‘interest in politics’ and ‘how often do you follow politics’ both having positive and significant impacts on civic engagement. Finally, religious attendance also has a positive and significant effect on the dependent variable. Looking at beta values, we can gauge the effect of the different factors, controlling for the whole model.

Here, we find that education and interest in politics explain the greatest amount of variation in civic engagement levels when controlling for all factors. Using this as a baseline, we can turn to models 2 and 3 to examine the specific effect of men and women on associational life. For structural factors, we see that education, income and employment status continue to

6 Spending time with family was only included in the Canadian, American and Spanish surveys out of all advanced Western democratic countries.

107

have important effects for men and women when they are considered separately, although for women, socio-economic status also has a significant and positive effect, while this is not the case for men. Furthermore, while family status also has no effect on men, or in the baseline model for that matter, when considering the case of women alone, it appears that being unmarried with children has the strongest, positive and significant effect compared to the other family status categories – which are all insignificant. In terms of the effect of cultural factors, when both genders are considered separately, interest in politics is the only indicator which continues to have a significant and positive effect on civic engagement.

Finally, religious attendance remains positive and significant for both men and women. The beta values reveal that interest in politics and education explains the greatest amount of variation on the dependent for men and women. Thus, this analysis suggests that not only does gender explain real differences in civic activity, but that these are tied to social and economic context such as family and SES, which shape men and women’s participation in distinctive ways.

Inequalities in citizenship are reflected in participation levels because, unlike the assumptions made by the dominant civic engagement literature, individuals in the community do not all equally enjoy of rights and entitlements nor do they have the equal capacity to participate. Instead, one’s location in the social matrix is tied to the resources and opportunities that are available for them to develop civic skills, make new and diverse ties, and build trust. The case of women demonstrates that, despite the advances in formal equality they have achieved via rights and entitlements, enduring social and economic disadvantage tied to gender roles, leads to lower levels of civic engagement compared to

108

men. This chapter has reviewed four theories seeking to explain women’s gender gap in participation. It showed that limited resources such as time and money, a preference for homogenous tight-knit ties, and low levels of interest in politics all contribute to women’s gender gap in participation. There also appears to be strong evidence in support of the fourth theory which suggests that gender differences in participation can be attributed to a methodological oversight, overlooking child-related networks in surveys and prevailing civic engagement analyses. However, the very lack of data on such networks makes it difficult to draw any further conclusions. To address this issue, the next chapter turns to a more in-depth analysis of women’s networking habits, described by women in their own words. In addition to providing information about how work and family life shape their associational lives, the data reaped from this analysis offers a better understanding about the value of child-, parenting-, and care-related groups, and the costs of discounting such networks.

109

Chapter 3 Count Me In: Taking Stock of the Networking Habits of Single Mothers

This chapter seeks to learn about women’s civic lives - from the women themselves. It aims to find out whether and why single mothers are or are not civically engaged, what types of networks they join, and which factors shape their networking habits. While survey data such as that used in the previous chapters provides a snapshot of women’s associational life, it omits certain types of networks related to children and parenting, and skips over a great deal of information which can shed light on how inequality and socio-economic context shape networking. This chapter uses personal interviews with a sample of sixty single mothers in

Canada, France and Sweden to fill the gaps left by survey data analysis, and test theories on the gender gap in participation outlined in the preceding chapter. The findings lend credence to the civic engagement theory which posits that networks are linked to higher trust levels, with single mothers in Canada and Sweden being both more active and trusting than those in

France. Moreover, familial bonds, when not accompanied by a range of other types of informal and formal networks, are found to be linked with lower levels of trust and networking. The results also show that balancing work and caregiving leaves lone mothers with a shortage of the time, money and energy required to participate, lending support for the resource-based account of the gender gap in participation outlined in the previous chapter. While the presence of young children proves to be a source of stress, often forcing respondents to tailor their civic activity around carework, they also act as a catalyst to networking. In fact, there is no support here for the psychological, orientation-based account for gendered participation; children and care work are not found to shift respondents’ mental energies away from civic engagement. The findings also lends credence to the

110

methodological account of the gender gap in participation, illustrating the importance of parent-, child- and care-related groups so often discounted by dominant civic engagement studies and surveys. These types of networks not only form a vital part of the civic life of single mothers, but they help build trust and reciprocity, foster a sense of community, and facilitate other forms of networking.

METHODOLOGY

Between the fall of 2004 and 2005, interviews were conducted with a sample of sixty single mothers in three of the Western advanced democracies included in the analyses from previous chapters: Canada, France and Sweden. Countries were chosen to reflect a diversity of contexts, rather than focus on most-similar cases in order to avoid bias. Countries were selected based on Esping-Andersen’s welfare regime typology (1990, 1999) which is elaborated on in Part II. For the purpose of this chapter, it suffices to note that this approach distinguishes between groups of countries which differ in terms of their historical legacy, their approach to citizenship and the relationship between the state, market, family and community. According to this approach, Canada, France and Sweden represent three distinct country cases.

The term ‘single mother’ here encompasses: mothers who are separated (never married) or divorced, mothers who are widowed, and single women who have adopted. In the quest to uncover the impact of inequality on civic activity, focusing on a disadvantaged, understudied group like single mothers is useful and necessary. Across Western advanced democracies, the nuclear family continues to represent the ideal family type in each country.

As such, stereotypes about lone mother-headed households abound, stigmatizing them as lesser mothers and workers. Respondents confided, for instance, that they felt they were seen

111

by others as ‘social cases,’ ‘victims worthy of pity,’ ‘easy,’ ‘addicts or criminals,’ ‘always absent from work,’ and ‘unreliable employees’ because of their primary loyalty to their families rather than their jobs. Moreover, dual roles as paid workers and childcare providers meant single mothers lacked key resources, such as money, time and energy that are necessary for participation (Herd and Meyer, 2002; Das, 2004). Workplace-related anxieties are not only rooted in a worry about promotions and job security, but ultimately entail concerns about being able to earn enough money to maintain a household and support dependents. Thus, employment stress is necessarily linked to the financial stress that single mothers face, namely having to provide for their family, often on one modest income. In light of these pressures and limited resources, it would seem that single mothers represent a good test case to learn more about individual-level factors shaping civic engagement.

The snowball method was employed to locate the sixty respondents. Random sampling was not an appropriate recruiting method in this case because rather than trying to establish statistical significance and generalize to the broader population, the goal here was to uncover the experiences of a discrete, insular group: single mothers. Moreover, a central purpose of the interviews was not only to understand how single mothers network in each country, but whether and why they are or are not civically active. Accordingly, recruiting respondents directly from clubs or associations would have been unsuitable, overlooking those who are inactive in favour of joiners. In each country, the snowballs were generally started by appealing to personal contacts, asking friends, family, colleagues, and my thesis committee if they could refer any single mothers who might be interested in participating in the study. In the case of Sweden, I also recruited respondents by contacting Swedish organizations located in Canada and asking them to post a notice with their members about

112

the study and the need for single mother participants in Sweden. At the end of each interview, respondents were asked to offer the contact information of any single mothers they knew that might be willing to participate. To minimize skewed findings (for instance, the ‘small world’/‘six degrees of separation’ phenomenon and selection bias), rather than continue the process with a single, original snowball, several different snowballs were started at the same time. In addition, the ‘spin-off’ from each separate snowball was limited so that respondents only passed on the names of one to two prospective recruits. To protect their privacy, respondents are identified by the first letter of the country they represent, as well as by a number (for example, C20 is the twentieth respondent from Canada). Recorded interviews were transcribed, alphabetized by last name, and numbered accordingly.

Respondents were offered the choice of being interviewed over the telephone or in person, leading to both approaches being used relatively equally. All respondents were provided with a letter outlining the purposes and parameters of the interviews, relevant contact information if they had any questions or concerns, and requesting permission to record their words whether in person or over the telephone; all respondents agreed to be recorded. Moreover, all respondents signed a letter of agreement that also indicated that their identity would be concealed and their privacy protected. Letters and approval forms were translated and provided in the respondents’ native tongue (English in the case of Ontario

Canada, French, and Swedish) prior to interviews, and the process met with the approval of the University of Toronto Ethics Review Committee.

The interviews included both closed and open questions, and were designed to gather information on: demographics, informal and formal networking, political action, generalized trust, confidence in major institutions, and care work (See Appendix for the Interview

113

Schedule). In some cases, World Value Survey questions were used as a starting point for questions, only to be further developed by additional questions – for instance, in the case of networking and care work. In other cases the exact World Value Survey questions were used

- for instance, to measure levels of generalized trust, and confidence in major institutions.

The latter approach allowed for both the testing of respondents, as well as the effectiveness of World Value Survey questions.

RESPONDENT PROFILE

The majority of respondents interviewed in Canada, France and Sweden had young children, ranging from infants, to those attending grade school. There were a few in each country that had older children, including teenagers and, in a couple of instances, some

‘adult children’ (18 years and older, still living at home). Studies on associational life have underscored the importance of accounting for the age of children to the networking patterns of parents. Over time, as children grow, care demands which cost single mothers time, energy and money, shift and change. This, in turns shapes the associational opportunities of single mothers in terms of the type and amount of networks they join (Munch, McPherson,

Smith-Lovin, 1997, p.514, 517).

Child’s age did appear to have an impact on the networking patterns of interview respondents. Work-related ties weakened as single mothers of infants took maternity leave, and some admitted that it was challenging to remain engaged in their regular formal groups.

This is not to suggest that respondents did not have active social ties, but during this period, networks revolved much more around their children, and involved family or other bonding- type networks, such as visits with close friends in the home. As children became old enough

114

to go to day care, have play dates, sleepovers, and begin to attend school, respondents became progressively more engaged in the community, and while some, if not most, of their social ties continue to revolve around their children, networks became increasingly bridging, associating with a more diverse group of people. Finally, when children could be left on their own in the home, and childcare was less of a barrier to participation, this signalled another change in the amount and types of networks single mothers engaged in.

In addition to the age of the children, the number of children also depletes the limited stores of time, energy or money that single mothers have to draw on, leaving less of the resources required for civic engagement. Several interviewees suggested that the number of children in their household was a source of stress, threatening their ability to network. In

Canada and Sweden, 5% of respondents had more than two children while a sizeable 20% of

French respondents had three or more children. The latter is in line with previous studies

(Esping-Andersen 1990, 1999) which suggest that France and other countries with traditional, authoritarian, Catholic histories tend to reward larger families.

Respondents had a mix of incomes, ranging from low ($27,000 a year or less after taxes) to high income ($70,000 or more a year after taxes). Canada had the highest number of respondents (25%) reporting high incomes compared to France (5%) and Sweden (0%).

These findings suggest that the more generous the welfare state, the smaller the salary.

However, the more ambitious welfare states of France and Sweden provide a wide array of benefits and services that are either not available to Canadian mothers, or which they have to purchase from the private sector (see, for instance, chapter 4).

While the majority of Swedish and French respondents interviewed worked 35-40 hours a week, only half of Canadian respondents did so. However, the differences between

115

these countries and Canada are surprising. Specifically, 95% of French single mothers interviewed worked full-time, including one who participated in a work-training program.

Just 5% of French mothers interviewed said they were on maternity leave. By contrast, 85% of Swedish respondents said they worked full-time, of which two respondents worked flexible work weeks (working longer hours on the week they do not have visitation, and fewer hours on the weeks they have their children). Moreover, 10% of Swedish respondents were out of the labour market and on medical leave. By contrast, only half of all Canadian respondents said they worked full-time, while 20% worked part-time. 30% said they were not active in the labour market, with 15% of these being on parental leave, 10% being unemployed (with one on social assistance and one on her last week of employment insurance, and about to start social assistance), and 5% being students. Differences in income and labour market activity found here may be the outcome of living in different welfare states. For instance, the greater number of higher incomes reported by Canadian respondents could be explained by the lower taxes on income in Canada compared to Sweden and France, and the lower levels of full time work by Canadian respondents could be the result of the lack of support available to help women reconcile work and family life. It is outside the scope of personal interviews to determine how welfare states shape participation, however such questions are explored in subsequent chapters at the country-level of analysis.

Overall, the mix of respondents recruited offer an excellent test-case to study the impact of work and family on associational life, and shed some light on the nature of child- or care-related networks. Although this is a qualitative study, and the sample size and methodology to recruit respondents clearly does not meet the requirements for statistical significance, percentages are often used here to give a better sense of how single mothers

116

responded to the various questions posed. The following sections outline interview findings regarding key civic engagement indicators including formal and informal networking as well as levels of trust, but also examine less conventional measures such as child-related social ties.

THE CIVIC ENGAGEMENT OF SINGLE MOTHERS: VOLUNTARY ORGANIZATIONS

Single mothers’ formal networking, including memberships in clubs or involvement in organizations, is under-explored. As pointed out in preceding chapters, when civic engagement analyses do consider gender, they are limited to drawing conclusions based on the list of voluntary group categories included in surveys, such as in the World Values

Survey. The findings from such research, confirmed in chapters 1 and 2, suggest that women tend to gravitate towards what may be referred to as ‘expressive’ community or welfare networks, over the more ‘instrumental’ political and professional-type groups. Yet little can be ascertained from previous analyses about how the juggling act many women perform between paid and unpaid work shapes their associational lives, nor do they include indicators designed to measure participation in child or care-related groups. Respondents may very well report membership in such associations, but these go unnoticed – subsumed under broader categories like ‘other,’ ‘welfare,’ ‘community,’ or educational’ associations. This section sheds light on these underreported areas revealing that a strong desire to be civically engaged keeps respondents very involved, despite the real constraints involved in balancing work and family life – reflected in limited resources.

117

When asked about their formal networking, interview respondents in both Canada and Sweden said they were very engaged, with 90% of Canadian and 80% of Swedish respondents saying they belong to some type of association. French respondents had the lowest levels of networking, with just over half (55%) saying they belonged to an organization. Additionally across countries, single mothers with affiliations tended to be engaged in multiple groups, as outlined in figure 3.1.

Figure 3.1 The Voluntary Membership Habits of 60 Single Mothers in Canada, France and Sweden

10% Charity 0 15% 15% Gym/fitness 25% 45% 35% Parenting/care group 0 25% 15% PTA/school 10% 15% 0 Religious Groups 10% 15% SWEDEN 35% Culture/arts/leisure 30% FRANCE 30% CANADA 65% Union 5% 4% 10% Professional 5% 45% 5% Political party/group 0 20% 30% Other 20% 25%

0% 10% 20% 30% 40% 50% 60% 70%

Membership in Voluntary Groups (%)

Source: Sample of 60 personal interviews conducted in 2004-5.

118

In Canada, the most popular type of groups were related to fitness or sports (45%), professional organizations (45%), and cultural groups (30%). In Sweden, the most favoured group by far was unions (65%), followed by parenting groups (35%) and cultural organizations (35%). None of the Swedish respondents said they belonged to any religious group. By way of contrast, cultural, arts and leisure groups were the most popular among

French respondents (30%), followed by sports and fitness-related groups (25%). Some in

France belonged to school and child-related groups, but none belonged to political or interest groups, parenting organizations, or charities. Therefore, overall, interviewees in Sweden and

Canada gravitated not only towards expressive groups – as expected, but also, contrary to the findings in preceding chapters, to instrumental networks related to their employment.

Respondents in France, on the other hand, favoured expressive associations.

Having young children led many respondents into child, parent or care-related formal networks including the PTA or school committees, parenting and youth groups, camp volunteering, regular activity in formalized playgroups run by community centres or school boards and, in some cases, groups related to an illness or condition that affected their child.

Yet, while many single mothers belonged to these types of networks, child- or care-based associational ties did not define their formal associational life. In other words, it was very rare that a respondent uniquely belonged to a child, parenting or care-related group. Rather, those who were members typically also belonged to other groups unrelated to the care of their children.

The tension between work and family life leaves respondents with a short supply of those resources which facilitate civic engagement. Interviewees across all three countries said that childcare demands often hindered the development or maintenance of their

119

associational lives – particularly with respect to formal networks. In the case of single mothers, many of whom are active in the labour market, there is no partner to schedule childcare duty with during those times when daycare services are unavailable, or pool resources together to pay for a babysitter or nanny, participation in formal groups proves to be especially challenging. According to the resource-based account of the gender gap in participation outlined in the previous chapter, civic engagement requires an investment of time and energy – something that is scarce when trying to balance paid work and the responsibilities of primary caregiving. As Herd and Meyer contend, “…there are only 24 hours in the day. For women, more so than for men, unpaid care work leaves less time and fewer financial resources for participation in politics, social movements, civic or voluntary groups, or other traditional forms of civic activity” (Herd and Meyer, 2002, p.670). Support for the resource-based account of the gender gap in participation is found in respondent statements, linking constraints on their civic lives with the limited time and money (which typically manifests itself as a need for money to hire childcare) associated with having and caring after children.

For example, in Canada, the combination of a particularly strong interest in being active on the one hand, and a lack of support for mothers on the other hand – not only in terms of state services and benefits, but also in terms of uninvolved fathers, may explain why

Canadian respondents expressed the greatest concern about childcare, with 40% saying it posed a significant challenge to networking. Canadian respondents allude to this by talking about their struggle to juggle work and family responsibilities. In the instances when they directly indicate that limited resources contributed to their networking habits, the key words used (time, money, energy/ feeling tired) are marked in bold:

120

C2: I have sort of a [single parents’] network – but it’s hard to keep that up more and that’s the one that sort of goes. We have regular meetings that are optional, but I’ll often not go if I’m behind at work or I have to get childcare. That ends up getting dropped… I feel good about belonging to that group. But I have understood that you just can’t go all the time. If your child gets sick, you have to scrap all your plans and maybe go the next month.

C3: I don’t belong... Things like that are difficult because of childcare… If it was just me and [my child] was staying home and someone else was cooking him dinner, I probably would do those things. But I can’t and often it’s inappropriate for me to bring [my child when going to a group meeting].

C6: Sometimes I need childcare – that’s always been a problem because babysitting costs money, and I can’t afford it. I should find a local babysitter, but it’s a money issue.

C7: I’ve thought of volunteering and I’d really like to, but what it comes down to is a childcare issue. If it’s a Saturday and I can bring them [my kids] with me, I’d probably do it – but there would have to be something in it for them – because, when they’re not in daycare, they’re with me. Every now and then, they’ll go to their dad’s, but 95% of the time, they’re with me.

C9: I’m not active in anything. I don’t have time… I wouldn’t mind going to a book club, but… I’m either doing stuff with my kid, or I’m working or I’m trying to work, and then when I do have down time I just want to be alone. I know a lot of people make a lot of time for these things and I don’t’ know how they do it.

Canadian single mothers were not the only ones eager to discuss the challenges that care work presents to civic engagement. Even in France and Sweden, which have more state support for single mothers along with well-developed and accessible early childhood education systems, one quarter of respondents in each country said that their roles as care providers impeded their associational lives. Respondents linked having and caring for children with limited resources in the form of time, money and even energy. Again, while in some instances, respondents allude to this by talking about their efforts to reconcile work and family life, when they directly indicate that limited resources contributed to their networking habits, the key words used (time, money, energy/ feeling tired) are marked in bold:

121

F1: I do feel like I’m missing something by not belonging to a group: time for myself. I have no time. Outside of work, I spend all of my spare time with my kids.

F10: No – I don’t belong to anything. I don’t have the time to do that. To do these things you have to have some free time. I work a lot.

F12: Between work and being a mom, I don’t have spare time to join a group.

F19 : I don’t belong to a union or any other sort of group. I used to be into sports before having my kids, but now I’m really raising my kids on my own... so I don’t really have any time. I don’t have a moment to myself.

S4: I do not belong to any organization... there is no time because of the children.

S7: I would like to do voluntary work actually – I think I’ve been working too much. If I had time I would join groups.

S8: I don’t have the time to join [a group]; I have such a high rent for my apartment that I have to rent a room for tourists so when I come home [from work] I have to clean the beds at home for guests. I don’t really have the time. I’m tired when I get home.

S14: I wish I could [belong to a group] but I have no time. I have too much work and I travel a lot between the UK and Finland. So I couldn’t commit to anything.

S15: I feel tired. A lot of decisions are made by me – so I get to decide how to raise my child. But when your kid goes to bed, you are in prison. You can’t go anywhere.

Thus, single mothers contend that their dual roles as primary caregiver and paid worker translate into a resource deficit which make it difficult to socialize.

Moreover, support for the resource-based account of gendered participation also becomes apparent when considering participation in conjunction with income levels. This relationship is not as simple as suggesting that those who are of lower income do not network since 74% of all low-income respondents (earning $27,000 or less after taxes) belong to at least one group. Rather, if we turn our attention to respondents who are self- confessed non-joiners across Canada, France and Sweden, belonging to no organization, we find that more than half (53%) are in the low-income bracket. Thus, while being of low-

122

income does not mean one is fated to a life of isolation and the of mistrust of others, it does suggest that, for a sizeable number, socio-economic disadvantages pose significant challenges to leading an active associational life.

In light of the burdens and challenges that they face, one might have assumed that single mothers avoid networking altogether – at least until such a time as their children are old enough to be left unsupervised. However, as figure 3.1 shows, most interviewees in

Canada and Sweden, as well as many in France, were keen to be active in formal networks.

In fact, a desire to join and interact with others was the principal reason given by respondents about why they belonged to formal groups. Out of the sample of sixty respondents, 64% said they maintained social ties to interact with others, meet new people or get some emotional support– for instance, via or mothers’ groups. 6% joined groups that nurtured their religious beliefs, such as church or groups, while 30% participate in associations which promoted a particular vision of life or society, such as human rights or environmental organizations. When asked to explain why they joined an organization and what sort of benefits they felt they gained from their involvement in groups, respondents overwhelmingly said that they reaped a sense of belonging. This is reflected in the following statements:

C2: I feel that I get support from others by belonging to these groups. That’s the major reason and benefit.

C3: Belonging to these groups makes me feel like I belong. It gives me a sense of belonging, a sense of community.

C8: Belonging to a group gives me a sense of community.

C13: It feels good to go to coop meetings and I feel like I’m part of a community. I like going.

F2 : The benefit I get from my memberships is a sense of belonging to society.

123

F11 : The benefits of belonging to a group is a social, relational benefit; meeting others and socializing is a relational benefit.

S6: It feels good to belong in group – belong to something – it brings meaning to your life. You share things in that group – you have things in common. There’s a feeling of happiness when you gain something – when you’ve done something – and also to share it – it’s absolutely fantastic. We’d done this… really really big demonstration… I was there organizing seminars and things like that and it was really a fantastic thing to do... It was a fantastic feeling to be among so many people.

S10: Belonging to a group makes me feel good because I know that I am not alone in this situation; someone is standing by my side.

S17: I joined to fit in somewhere and feel belonging.

S18: How does belonging to these groups make me feel? I feel like I belong.

In sum, respondents claimed that the presence of young children was a source of anxiety for respondents because it constrained their ability to conduct an active associational life - at least in part by draining respondents of much-needed resources to participate. Yet children were also said to act as catalysts, propelling a majority of respondents into groups.

Canadian and Swedish single mothers were most active in formal networks while those in

France had the lowest levels of membership in formal organizations. Although the presence of young children may pose a challenge to joining formal groups, as can limited resources, single mothers nevertheless often found ways to remain engaged. As the following section shows, many turned to their informal networks to facilitate participation.

THE CIVIC ENGAGEMENT OF SINGLE MOTHERS: INFORMAL NETWORKS

As figure 3.2 shows, single mothers were highly active in informal social ties – networks with friends, family, neighbours and others in the community that do not require a membership card or the payment of dues. In Canada, France and Sweden, women were

124

involved in networks with kin, other parents, people from the clubs they belong to, childhood friends, as well as with work colleagues. Familial ties were the most common informal network that respondents included in the list of people they socialize with on a regular basis, with 90% of French and Swedish respondents saying they have close ties to family, and 70% saying so in Canada.

Figure 3.2 The Informal Networks of 60 Single Mothers in Canada, France and Sweden

55% Neighbours 35% 40% 75% Work colleagues 65% 35% 45% Childhood/school 40% 10% SWEDEN 15% From groups/clubs 25% FRANCE 15% CANADA 15% Other parents 15% 50% 90% Network with family 90% 70%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Informal Networks (%)

Source: Sample of 60 personal interviews conducted in 2004-5.

In terms of non-familial ties, Canadians led in networks with other parents (50%), while both

French and Swedish respondents said they favoured networks with work colleagues (65% in

France, and 75% in Sweden). Moreover, in Canada, 60% of respondents said they could count on five or more people, while 70% said so in France and 65% in Sweden. Thus, while they may lag in organizational involvement, French interviewees lead in informal networking volume.

125

Respondents across all three countries said they had frequent (daily to weekly) contact with friends and/or family. In Canada, 65% of single mothers interviewed said they had weekly contact with their informal networks, while in France, 75% said they had daily contact with informal ties, with just half of the respondents in Sweden (50%) saying they had weekly contact. Thus, a large part of the pervasiveness of informal activity compared to formal groups is that such networks are an integral part of life – seeking pleasure and support in one’s day-to-day life from peers, family, friends and the community. Furthermore, informal networks do not generally pose the same dilemma about childcare that formal networks do; it is more accepted to have one’s child with them when meeting friends or family, than when attending a feminist book club, union meeting, or pottery class.

The challenges that come from having to balance paid and unpaid work have forced single mothers to become ingenious with finding alternative means of care. In Canada,

France and Sweden, single mothers indicated that they rely on informal networks to meet their care needs – although in different ways. In Canada, 69% of respondents whose children were still too young to be left home alone made some reference to relying on their informal networks in the caring of their children when necessary. For instance:

C3: You have to be creative with your work hours and working around school… There’s only one parent so there’s only one person staying home – it’s not like I can say “well, I stayed home last time so it’s your turn.” It’s just me. So you have to be sort of creative on how to do that… There’s a lot more juggling because there’s just one person. Socializing – something I think I last did four years ago – I can’t afford a babysitter. So there’s a mad scuffle to ask if my child can do a sleepover. So my child will sleep over at someone’s house. I’ve found most people to be very supportive around that.

C8: When I was in hospital, people came out of the woodwork and there were actually more people offering to take care of her [the respondent’s daughter] than I needed…When I was recovering, I went to the country to someone’s house and [one friend] picked me up and brought me back to Toronto.

126

In the case of Canada, while 70% said that family members were among those with whom they have frequent interaction with and consider part of their social web, the interviews reveal that respondents tended to call on informal networks with other parents more so than any other group for support. This suggests that although it is important to tally how many respondents network with a certain group-type, the story that those numbers cannot tell is how these relationships differ; who they casually socialize with, for instance, versus who they rely on.

Many informal networks involve the children themselves – such as the children’s friends’ parents, daycare centres, schools, or children’s activities. As a result, much of those activities tend to originate from, and revolve around children. In the case of Canada, 60% said that their informal activity revolves around their children (compared to 5% in France and 30% in Sweden), and half said their informal ties consisted of other parents (compared to

15% in France and Sweden). The striking differences between Canada on the one hand, and

Sweden and France on the other, could be the result of a lack of services and supports available to help Canadian women balance work and family life. Such a lack of social provisions would force Canadian women to fend for themselves, spending more time caregiving and incorporating this into their networking, leading them to spend more time with other parents in social ties that revolve around their children. Such factors, however, are explored in greater detail in subsequent chapters which conduct country-level analyses.

The importance of networks that revolve around children and other parents is underlined by several Canadian respondents:

C2: I network with other parents… mostly through the school or the daycare – meeting parents through school functions, or picking up or dropping off my kid. And some are single parents also looking for play dates etc., so many have become friends… The activities do tend to revolve around our children. We’ll do something

127

after picking up the kids from daycare, or have dinner about once a month. I hardly ever phone or email anyone. I mostly run into people when I’m picking up my kid. We arrange get-togethers with our kids…

C16: My circle of friends is my lifeline – they are my kid’s friends’ parents. One woman I met in a new mothers group, and one has become a really good friend of mine. Her son had a second birthday party, and I met other people there. Those other people have now become close friends – this is three years ago. And from there, I have a really good solid group and support. I’m ‘the’ single mother of that group... We’ll take our kids skating together or one will take the kids. I like to take as many boys over on Sunday mornings because it’s a long morning so it helps and the parents like it… My networks revolve around my child 100%.

Thus, informal networks, and child-related networks in particular, are an important form of participation for Canadian single mothers in this sample. Relying on a system of other parents appears to be a very effective safety net for them. Moreover, child-centered networks often encouraged interaction with people from diverse ethnic and economic backgrounds, as well as family types. Therefore, the informal network of Canadian respondents is surprisingly bridging.

C9: I don’t really hang around other single moms – I suppose a lot of that is because [my son] goes to… a French school and people come from all over the city to take their kids there and it’s very affluent – so it’s a school where people have really made a conscious decision to send their kid there.

C14: For her to feel like she doesn’t have a father – that’s something I worry about. So I make sure to have other children around her - different types of families. When we go to the drop-in, she’s the only one that’s from a single parent family, she’s the only black child and she’s the only one that’s been adopted. I’m comfortable with it now though partly thanks to having joined this women’s group.

In addition, child-related networks with other parents can be a very effective way of building the trust and reciprocity so integral to associational life. In such networks, children act as the prerequisite for reciprocity, allowing parents to take turns caring for each other’s kids. Thus, the exchange inherent in child-related networks is rooted in the understanding that each parent can be trusted with the other’s kids, and that this will lead to a give-and-take

128

(i.e. you take my kid this time, and I’ll take you kid next time). Several respondents highlighted this aptly:

C6: I have a group at school that’s not formal – it’s lots of school moms. It’s just some of the mothers of the kids that [my son] knows. It’s sort of a support group and [my son] can go and play. If I’m late, I can count of one of the moms.

C16: There are two households involved [in my main informal network], and now two more couples are coming into the group… we take turns and I take my share of carework – “OK - tomorrow it’s my turn to take the kids.”

A foundation of trust is required for child-related informal ties, just as it is with formal organizational involvement, because it binds members together in a reciprocal arrangement and facilitates other networking opportunities. As such, for a child-related network to be effective, a respondent would have to feel other the parents involved could return the childcare favour in the future. If the other parent is unable to follow through and reciprocate, trust will be broken, and they will likely be unable to maintain that network. While the importance of trust in building networks is not at the forefront of the minds of respondents, and is not explicitly mentioned when describing their social web, it is apparent in a few statements. For example:

C3: I do have two siblings who are very supportive, but I’m just more likely to ask a friend because I want to help someone who has a child – because then I have something to bank on – you take my kid, then I’ll take yours next time. Since I don’t have a partner to take care of my child, I have to go through this whole creative thing just to go to yoga class.

C2: There’s about five people I can rely on… they’re all parents… mostly married couples. There is only one who is single. She’s separated and they have joint custody… I know a lot of single parents, but I find that it’s actually hard for them to help each other out... So I don’t really rely on them.

Moreover, of the 10% of Canadian respondents who said they did not belong to any organizations and had informal ties revolving around family, one admitted that she had

129

trouble building reciprocal relationships because her friends and family felt her children were too unruly, making it difficult to build reciprocity.

C7: For me, I don’t really have a strong social network. Usually, every now and then, I can depend on my mom. But, it’s not regularly, and it’s usually just for when I go to school. If it’s for something personal, she doesn’t want to babysit… I only have one friend that has a kid that’s around my son’s age, and one time she watched him and he was too excited and it became a problem. So I don’t even put her through that anymore… It’s very rare that I go out socializing. My kids are a handful… it’s tiring and it’s not easy.

In this instance, a lack of reciprocity from others (due to unruly children) and a sense of embarrassment from the respondent quickly put an end to an already limited social networking system, leading to further isolation. This also shows that, while the vast majority of Canadian respondents rely on family and friends for support in the care of their children with great success, in a minority of cases relying on informal networks for care can be highly precarious. This sentiment is echoed by another Canadian respondent

C16: The two girls that were renting rooms at my place took off without paying the rent and they used to take care of my son. When I don’t get screwed like that, I’m fine. But when I get screwed it can become a problem… I used to have these two Guatemalan sisters staying with me and it was great, and they used to babysit - no problem. But then they left and two other women moved in. They were not good babysitters and they just took off.

As already indicated in figure 3.2, the majority of Canadian respondents socialize with an array of informal networks that include friends, work colleagues, neighbours and, in some cases, family. Yet, it is undeniable that, irrespective of income and education, networking with other parents is at the core of informal social ties for those single mothers.

Despite its drawbacks, it seems to serve them well, providing them with a source of support and stability in their day-to-day life. In addition, it can secure them some much needed time and energy to engage – whether by joining an association, or socializing with parents while

130

their children play. Their informal networks, moreover, keep them highly engaged and trusting through countless reciprocal exchanges with other parents.

Since France has a highly effective, sophisticated early childhood care system, it would seem that childcare would pose less of a hindrance to the associational lives of French respondents than it does, for instance, for Canadians – who lack such social provisions.

However, since networking often takes place outside of the workday – when daycare facilities are closed, even a comprehensive daycare system is unable to ease women’s unpaid caring burden towards facilitating civic engagement. Thus, like their Canadian counterparts,

French respondents often turned to informal networks to meet their childcare needs. While

Canadian respondents generally called on other parents for both socializing and as a reliable source of childcare, the French tended to depend on friends for socializing on the one hand, and calling on family to meet their childcare needs on the other. As such, although many

French respondents said they were active with friends, particularly childhood friends and work colleagues (see figure 3.2), they are unique in the degree of reliance they place on family ties. However, in order to understand the importance of family to the French case, one must consider informal networks alongside formal networks because it is the low levels of

French organizational involvement that elevates family networks in importance. The strength of French family ties is reflected in the following responses:

F18: I mostly network with work colleagues. I also still have contact with my ex- brother-in-law and my ex-husband. I spend time with my family too – my parents and my sister, as well as my ex-husband’s family. I don’t really know my neighbours yet because we just moved here. And I have one good friend that I’ve know for 25 years and another one I’ve known for 20 years… I see my parents and my sister every single week.

F19: It’s with my family that I network with above all. My parents, and my sister and brother – but it’s mostly with my parents. I have a few friends from childhood. My

131

work colleagues aren’t really my friends – more like acquaintances. I don’t have much contact with my neighbours. So it’s really mostly my family.

Furthermore, 63% of French respondents whose children were too young to be left on their own said they relied on family to meet their childcare needs when necessary. For example:

F3: I’m very close to my family – my parents, my brothers, and also several cousins… I can easily ask my parents or my ex-husband if I need childcare.

F4: I spend my time with my family – mainly with my mother and sister… when I need childcare, mostly on weekends when I go out, my mother watches my son.

F20: If I’m in a bind and I really need childcare, I turn to my sister or parents. But it’s much more than just babysitting. For them, it’s a visit at the grandparents’.

In addition, while only 5% of French respondents said their networks revolve around their children, and 20% indicated that only some did, three quarters asserted that their associational life was not built around their kids. Rather, by dividing their free time between family and friends, French respondents said they developed their own informal ties. This is captured in the following statements :

F9: My networks do not revolve around my kids. It does happen from time to time that I’ll have my daughters’ friends over… but I don’t really see other parents…

F17: No – my networks do not revolve around my children - even when they were much younger. When you work, you don’t really have the time to get to know the parents of other kids. I would pick them up from daycare and I would never run into other parents on a regular basis.

Organizing their networking so that they socialize with friends and family, and rely on family for childcare, seems to give French respondents the ability to not only participate in activities with their kids in certain networks, but also take part in more adult forms of socializing when they are without their children. Thus, while 30% of French respondents and

40% of Swedish respondents mentioned going to a pub or a nightclub as a common social

132

activity they engaged in with their friends and family, not one single Canadian respondent engaged in this.

Building an associational life around family and old friends is common across

Canada, France and Sweden. Yet, while these ties may prove to be highly reliable and effective, centering networking habits on them is likely to be limiting in terms of the types of the civic skills women can develop, the norms of generalized trust that can be built, the information gathered and the opening of other networking opportunities their memberships can reap (Munch, McPherson, Smith-Lovin, 1997; Popielarz, 1999; McPherson, Smith-

Lovin, and Cook, 2001; Gidengil et.al, 2006). Fostering close networks around others who resemble us, or with whom we share a common history or background is engaging in bonding networks (Uslaner, 1999). These types of associations, which tend to be inward looking, may only be able to generate particularized, rather than general social trust; they also tend to do very little to develop tolerance of others that are different than us - a common answer by French respondents who say they are not ‘joiners’ and mostly spend their free time with family. The following statements exemplify how family networking might be linked to a lack of confidence in those outside one’s immediate family or circle:

F14: I network with my family. This includes my mother and my aunt, but also the family of one of my children’s father. I don’t really socialize much [outside my family] because I think people are hypocrites. I prefer to be left alone… So I’m closest with two people… I spend a lot of time with my mother and my aunt.

F13: My family is my network – I only have my mother with me now. So I mostly rely on her. I don’t have any contact with my neighbours. With my friends, it’s always just about enjoying yourself and having fun... so when I need to confide in someone about my worries or troubles, I prefer to go to my mother.

Rather than networking with family like the French, or other parents like the

Canadians, the majority of Swedish respondents have a rich tapestry of networks that include

133

both friends and kin. Thus, every Swedish single mother interviewed claimed to network with friends, 90% also said they frequently interact with family members (see figure 3.2), and a vast majority said they belong to voluntary organizations (see figure 3.1). Respondents asserted:

S11: The whole world is my social network. Even people in restaurants etc. because my ex-husband is a musician, so we used to take the kids to the clubs etc. and everyone would help us…I also have a good relationship with my family – especially my sister and mother. I also have friends from childhood and school, as well as from work. And I get along well with my neighbours – especially one in particular.

S20: I have very good contact with my neighbours. I still have contact with friends who live further away but who don’t have kids – but it’s not in my everyday life – these are friends that I knew from traveling or through other friends etc. My family – they live in a little town 350km away – my mother comes to see me quite a lot and my son visits with her. Also my father, brother and sister – are all interested. But it’s not like I can always depend on them since they live out of town. We do weekly phone calls. I’m also in contact with work colleagues – like three or four people that I see when I don’t have my son.

While 30% of Swedish single mothers said that their networks revolve around their kids, and

20% said only some do, half of all Swedish respondents said that their social activities are not linked to their children, insisting that they maintain their own separate networks.

S7: My networks are still the same as before I had kids. So, no – I wouldn’t say that all my networks revolve around my children.

S13: Yes to some extent, my networks revolve around my son’s networks. But I have the same people around me as before I had my son, so it’s not all about him. So I’d say, some - but definitely not all.

Thus, a majority of Swedish respondents have found a way to participate in children’s activities, while still pursuing their own interests - all within their informal networking.

Moreover, according to the literature, since their social ties are so numerous and diverse, there is reason to believe that Swedish interviewees tend to develop bridging rather than bonding ties. Scholars suggest that interacting with such a wide-ranging group of people –

134

many of whom are different from oneself - is ideal for forging generalized trust. The findings here support this, with the Swedes and, to a lesser extent, Canadian respondents having the greatest mix of ties and correspondingly higher levels of trust compared to the French who have the least diverse ties and the lowest levels of trust.

Yet, how do Swedish single mothers manage to work, for the most part, full time, remain highly engaged in formal networks, and still have the time and energy to maintain such a vibrant, social life? After all, as in the case of France, Sweden’s effective and affordable early childhood care and learning system does little to directly facilitate participation since most facilities close shortly after regular business hours – when networking typically takes place. The answer lies in the their informal networks, drawing on them as an effective solution to the enduring childcare issue that, as it turns out, plagues the associational life of single mothers across all three countries. As in the case of Canada and

France, there is an important and intricate link between formal and informal networking, with informal networks not only acting as social ties in themselves, but also facilitating more formal types of activities.

While Canadian respondents relied mostly on other parents, and the French on family ties to meet their care needs, many Swedes depended on a ‘formalized’ version of an informal network in the form of shared custody of the children. In fact, over half of the

Swedish respondents shared custody with the father of their children. Characteristic of

Swedish egalitarianism, there is a concerted effort to have both parents involved in care work which has resulted in a system that encourages both the mother and father to share custody equally. This approach, commonly referred to by respondents as ‘50-50,’ is extremely popular, with 75% of Swedish respondents who are divorced or separated, and have a claim

135

to shared custody, saying they have such an arrangement with the father of their children. It seems to allow Swedish respondents the ability to live two separate lives: one with their children, and one without – when their children are with their father. Those who did not take part in 50-50 tended to turn to other types of informal networks for support – mostly friends, rather than family. However, the majority of Swedish respondents had 50-50, and they highlighted how this arrangement has largely eliminated their childcare problems. For instance:

S17: I’m not the primary care giver – I have 50-50. In Sweden, unless you petition the court for something else, when you separate – it’s 50-50… I never need childcare because of 50-50.

S19: Being a single mother is quite tough – it’s very intensive. But it’s not too bad because of 50-50 with the father… I never need childcare because of 50-50. I plan around it.

In addition to helping ease the childcare burden that single mothers would otherwise disproportionately shoulder, the 50-50 arrangement opens doors to interesting networking opportunities. It allows Swedish single mothers to engage in certain networks when they do not have their children, and shift their activities from more adult forms of socializing (going to pubs or groups) to those that revolve more around kids’ activities on those weeks when they do have their children with them. Respondents highlighted this during interviews:

S5: It can be tiring to be a single mother… Now I have 50-50 custody with the father, so I have the kids half the time and then they both go with my youngest daughter’s father. So that’s good because I get more time for myself… Sometimes I have the kids, and sometimes I’m without the kids… so based on this, I do certain activities like dinner with friends, or going for a beer or a coffee. And other times, we’ll go to the playground…It’s very rare that I need childcare because of 50-50. I just plan accordingly.

S13: I feel pretty good as a single mother – my son’s father is taking care of our son 50% of the time…. I feel pretty good about it actually… I mostly network with friends … sometimes we go out together to the park or take a trip somewhere or go to the pub on the weekend when we don’t have our kids…I don’t need childcare very often.

136

I plan my life so that I do certain things when I don’t have my son. So when my son is with his father, that’s when I do my activities… and if I do need childcare for whatever reason, I just ask the father.

In addition, when coupling the 50-50 shared custody arrangement with flexible work schedules - available to parents until their children are eight years old – this creates a situation that is favourable to freeing single mothers of some of the stresses or burdens they might otherwise feel. For instance:

S20: I work 80%. So I work 6 hours 4 days/week when I have my kid, because I have 50-50 custody with the father. And I work 9 hours 5 days/week when I don’t have my kid… It’s very flexible - and it’s legislated by the state.

While the approach most relied upon by Swedish respondents may be a reliable source of childcare, it nevertheless has its drawbacks – most of which are associated with shared custody in general. For instance:

S5: I have 50-50... so that’s good because I get more time for myself [but]....In Sweden there has been a lot of talk about not letting the father in 50-50. Some single mothers are not letting the fathers in – and then the single mothers get a lot of subsidies from the government and the fathers pay [maintenance]. Sometimes, the mother doesn’t want to let them [fathers] see the children. But most mothers want the children to have contact with the father.

On the other hand, as there are no maintenance payment under the 50-50 shared custody arrangement, one Swedish respondent out of those who were involved in 50-50 suggested that this has been taken advantage of by some fathers who are reluctant to pay:

S6: Sometimes the father says that he wants 50-50 so he doesn’t have to pay alimony but then doesn’t provide support [in raising the children]…. When both parents agree on 50-50 and both cooperate, then it’s great. But when it is imposed on you by the courts, then it can be terrible because the child gets caught in a war.

Forced, regular contact with the father of the child may therefore prove problematic for some single mothers, particularly if the separation or was difficult, or if there are great

137

discrepancies in the incomes of the father or mother. In some instances, respondents admitted their reluctance to share custody.

S1: I feel… a bit sad [about the 50-50]… It is sad for me that I don’t see her [her daughter] every week. That’s the biggest problem: not always having the child with you. But the older she gets, the better I feel about it.

S17: I have contact with the father of my child because we have 50-50 custody… I think it’s hard because my ex has a house and two cars, and I live in a little tiny apartment. So it’s 50-50 in that he pays his cost, and I pay mine. But my son doesn’t have a 50-50 living space….

Moreover, 67% of Swedish respondents involved in shared custody confided that they nevertheless continued to feel they are the primary care providers:

S11: I have 50-50 with the father, but mentally I have primary care. I have to keep track of everything, and when there is an outing, everyone comes to me first.

S18: Officially we share 50-50. But in Sweden, like in most countries, they always contact the mother first because she is really the primary caregiver… even though officially it is 50-50.

Nevertheless, while it has its drawbacks, the 50-50 scheme appears to provide Swedish single mothers with tremendous networking opportunities, largely by making an effort to minimize the disparity in resources (such as time, money or energy) expended by single mothers and fathers.

As outlined in figure 3.1 and 3.2, the interviews have shown that, while Canadian,

French and Swedish respondents are all highly active in informal networks with family and friends, they nevertheless have divergent networking profiles. In all sixty cases, an important amount of informal activities revolved around children – especially in Canada, where child- related social ties play an important role for a majority of single mothers interviewed.

Canadian respondents preferred to network with other parents on an informal basis, while

French single mothers turned to family. Swedish respondents tended to be involved in a mix

138

of networks, including family, and various types of friends. Thus, as the previous chapter has shown, women are more active in kin and familial ties which are associated with lower levels of civic engagement (Norris and Inglehart, 2006). However, the interviews indicate that familial and kin ties are only detrimental to civic engagement under certain conditions.

While Sweden and France both demonstrate high levels of kin and familial networks, the former exhibits high levels of civic engagement while the former does not. The findings suggest that the effect of familial ties is tempered in the Swedish case because respondents are active in a substantial number of other informal and formal networks, while the predominance of familial ties over other networks in the French case leads to lower levels of trust and network homophily. Moreover, in each country, informal networks often acted as a source of childcare, in Canada via networks with other parents, in France with family, and in

Sweden, relying on the 50-50 shared custody arrangement. Accordingly, respondents have highlighted the way in which informal networks are indispensable to them, acting not only as a form of association, but also enabling other types of networking.

THE CIVIC ENGAGEMENT OF SINGLE MOTHERS: TRUST

Since Canadian and Swedish respondents had the highest levels of overall networking, it is perhaps not surprising that Canada and Sweden have the highest score in terms of generalized trust, while the French, who suffered from low levels of networking, particularly in formal groups, exhibit extremely low levels of trust. Figure 3.3 indicates that

70% of Swedish and Canadian respondents say most people can be trusted, while just 10% say this in France.

139

Figure 3.3 Social Trust Levels Among Single Mothers in Canada, France and Sweden

SWEDEN CANADA 70% 70%

FRANCE 10%

Source: Sample of 60 personal interviews conducted in 2004-5.

These results largely echo those of the findings from the previous chapters. The fact that the most active are also the most trusting lends support to the central civic engagement premise that trust is an integral part of associational life. Canadian and Swedish single mothers were found to be highly active in a variety of diverse, bridging networks – many of which did prove to demand trust and reciprocity – such as the child-related networks in

Canada. Conversely, the French tended to withdraw from formal groups, opting for the comfort of bonding relationships like family ties, which do little to forge generalized trust.

The trust and reciprocity that is so central to the effectiveness and endurance of child- related networks between parents in Canada, for instance, is largely absent when a French single mother asks her parents to watch their grand-child. It is not that trust is missing from

140

the family tie, but it is interpersonal (of those who are close to you) rather than generalized

(of others in general) trust. Further, because interpersonal trust is likely to be pre-existing in a familial bond (blood is thicker than water), it remains separate from exchanges that take place within family networks (for instance, whether or not family members share reciprocal relationships), and is generally not required to start or maintain them. This suggests that family networks do not, on their own, count as a viable source of general social trust.

Moreover, there is reason to be skeptical of ‘trust’ measurements which rely on the common survey question: Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing with people? Like many questions aimed at measuring a norm, the responses elicited from this question tend to be easily skewed by the interviewee’s recent experiences. This is exemplified in the following responses when interviewees were asked about their levels of trust:

C16: After what happened to me [referring to two tenants who suddenly left without paying their rent]? Apparently not!’

F3: No – certainly not after having lost all my friends during my divorce!

Even aside from the issue of bias from recent life experiences affecting responses, this ‘trust’ question, often referred to as a reliable measure and included on the World Value Surveys, was tested during the interviews in all three countries, with poor results (Reeskens and

Hooghe, 2008; Glaeser, Laibson, Scheinkman and Soutter, 2000; Hardin, 2002; Uslaner,

2002a). 78% of the time, respondents did not understand the question as it was presented, had difficulty answering it, or probed the interviewer for further explanation, while others refused to answer altogether because they felt it was overly simplistic. This is illustrated in the following responses:

141

C4: That’s not a question that permits an answer! Trusted about what? Trusted with money? Trusted emotionally? I can’t answer that question.

F1: On what level are we talking about here? With friends? In general – like with friends or family, yes – I think they can be trusted, but politically, no. I think we’re manipulated a little. And the average person on the street – strangers – no. I don’t trust them. I could never leave my kids with strangers.

S15: It depends on the area. If it’s with my child, then “no,” I’m more likely not to trust a stranger.

Thus, a different method should be adopted to better measure trust – one that includes multiple indicators for generalized trust. Such an approach should include a series of questions on trust that provides more context to help respondents understand what is being asked, hopefully leading to a more accurate, reliable answer. The following questions are presented as examples of this approach:

- If you go to a museum or gallery, would you leave your coat in an unattended coat room? - If you were in a coffee shop, would you trust a stranger sitting nearby enough to ask them to watch your belongings while you went to buy a coffee? - Would you leave your cart full of groceries unattended outside the supermarket for as long as it took to go to the parking lot and get your car? - If you left your camera, in a camera bag, on a public bus or train, do you think someone would bring it to the lost and found? - If the ‘accessible’ fare-gate (for those who face mobility problems) was left unlocked, do you think most people try to slip through to ride for free rather than pay at the turnstiles?

Therefore, by placing respondents in a hypothetical situation, this approach might help respondents provide a more accurate and hence more reliable measure of generalized trust.

However, since these sample questions involve scenarios that centre on trusting others with one’s personal possessions, some may prefer to focus on a series of different, broader generalized trust questions, as found in ISSP Role of Government IV and the European

Social Survey, including:

- There are only a few people I can trust completely

142

- If you are not careful, other people will take advantage of you - Would you say that most of the time people try to be helpful or that they are mostly looking out for themselves?

The results indicate that, compared to the French, Canadian and Swedish respondents who are not only more active but also involved in a diversity of networks, are also more trusting. However, the interviews do not clarify the causal mechanisms of civic engagement

– that is, whether it is networking that builds trust, or whether it is already-trusting societies which produce citizens that are more likely to be civically active. This section also raised important questions about the validity and reliability of prevalent trust measures, and offered possible solutions to address these.

THE RELEVANCE OF CHILD-RELATED NETWORKS

The interviews underscore the importance of child-related networks, first and foremost, to Canadian single mothers, but also in Sweden and France. In each country, child- related networks are a constant rather than variable. They include both formal, and informal groups that stand alone valid, valuable social networks. Thus, play-dates, school-related activities like the school paper, the PTA, pick-ups and drop-offs at school or daycare, children’s activities such as sports teams, choir, aqua-tots or nursery rhymes at the library, parents groups of various kinds and, of course, sleepovers; all these social ties are found to build trust and encourage reciprocity between members of the network – particularly when the networks involve non-family members – where there is no pre-existing level trust in the network. The exchanges that take place within child-related networks, such as sleepovers for example, are built on reciprocity (I’ll take your kid this time, and you’ll take my kid next

143

time etc.). One example of the trust and reciprocity that binds members of child-related networks is the network as an elaborate child monitoring service:

S12: I want to know who my son’s friends are and who his friend’s parents are so that I know that no one will buy them alcohol or something like that. We are actually trying to start a little group so that if we see one of them smoking, we would tell the other parents and I would expect the same from the other parents. It sends a good signal to the child that there is always a parent around.

While it appears that the type of trust produced by child-related networks is particularized in nature (trust between the immediate members of the networks), unfortunately, the interviews do not clarify whether such associations also generate the generalized social trust required for civic engagement.

In addition, child-related networks, like other formal and informal networks, tend to interact with one another, with formal networks feeding into informal networks, and informal networks facilitating more formal, types of participation. This is partly because, irrespective of how generous or developed the daycare system is, such facilities tend to close around the time when networking activities begin: at the end of the workday. Consequently, single parents in Canada, France and Sweden must work to find alternative arrangements to meet their childcare needs. It may seem logical to assume that the burden of childrearing acts as a deterrent for the civic activity of single mothers. However, as the interviews demonstrate, childcare can act as a catalyst to networking. As one respondents illustrated, children often work as a draw:

C3: Who are my networks? Probably the strongest networks are people who have children of the same age that I’ve met through the schools, neighbours, parks, and we’ve connected with our children. I’ll give you an example when my child was little, we went to a library morning – and it was ending soon and I was desperate, frantic because this was the only socializing I was doing. So, we started a support [network] – we had 1 father and 8 mothers. We met every week at 1 person’s house and we did this for 5 years. The whole idea was that we could connect and have that support. Some are still very strongly involved, and others less, but we are all still more or less

144

connected – 10 years later. So it’s through the children… that’s your common bond – the kids. It’s a lot like having a dog. When you have a dog, you meet other people with dogs. Once we got our dogs, everyone stopped to talk to me. When you have a child – it’s the same thing. People stop and they comment about your baby, they’ll talk to you much more, other people with children will talk to you much more. It’s like you have this extra thing that lets you into all these other places and all these other connections.

Thus, children often serve as a source of child-centered social interaction.

Yet, while child-related networks are prevalent, they are consistently overlooked in civic engagement analyses, and are rarely included in surveys. Scholars have suggested that this lacuna stems from the tendency to focus on male-dominated activities in measuring associational life, devaluing care and child-related activities in the process. Thus, Lowndes asserts: “[d]espite its promise to link informal community-based activity with broader political phenomena, the social capital debate has often replicated the classic ‘public/private’ split, whereby it is assumed that women’s activity is ‘outside the political world of citizenship and largely irrelevant to it’” (Lowndes, 2006, p.215-16).

This notion that networks can be split between those that are ‘domestic’ and those that are ‘civic’ in nature is flawed. Many types of informal networks, which typically take place in the ‘private sphere,’ and which form an integral part of associational life, are often included in surveys, such as spending time with friends and family. Putnam is clear on the fact that networks need not be inherently political or formal to ‘count’ as civic engagement; what matters is that networks foster trust and reciprocity. Moreover, informal networks are often a platform for individuals to discuss political issues and concerns. Respondents in

Canada, France and Sweden confided that all types of networks, whether formal or informal, provided them with a forum to discuss political issues. When asked whether they discussed politics with their social networks (frequently, occasionally, or never), the majority of

145

respondents (60% in Canada, 80% in France and 45% in Sweden) replied ‘occasionally,’ followed by closely in Canada and Sweden by ‘frequently’ (30% in Canada, 5% in France and 40% in Sweden). As one respondent explained:

C3: when you’re a single parent, it’s just you and the kid, and you just crave adult conversation. That’s a really strong piece for a single parent because there isn’t that other person...

The power of child-related networks is, moreover, reflected in the effect these can have on the community at large. In the case of Canada, where parent-centered, child-related networks thrived, these webs of ties seemed to extend to the entire community. Thus, child-related networks play an important part in participating in, and developing a stronger community in a very real way. Several Canadian respondents spoke about the link between their children and their community in the following way:

C2: A lot of my networks do revolve around my kid… whether it’s daycare, school, or the neighbourhood parents. It’s a good community – the park is a great connection. We’ll go there and run into somebody. There’s a rink there and it’s enclosed so you can have dinners there.

C3: When you have a child, you get to know the community. My son’s best friend lives across the street, and the school is there – you’re less likely to move. And when you’re a single parent, you’re really tied to your community… We have a really strong community here – we have email listings in case you need a plumber or whatever… It gives you the sense of belonging, a sense of community.

C8: I’m really, really a part of my community… there are about six houses on the corner and my house is one of them and we all have each others’ house keys. We all socialize – last week we all felt sort of blah – so one family invited us all and made a Hawaiian theme – with pineapple, and Hawaiian pizza. And the kids go in their pyjamas from one house to another. I can truly rely on these people. I also have networks around my daughter’s school… One single mom just got a job so we all took turns picking up her kid from school – so it’s a lot like that. When I was in hospital, people came out of the woodwork and there were more people offering to take care of my daughter than I needed.

While not directly speaking about their community, several respondents in Sweden who resided in co-op housing spoke, in similar terms, about the sense of support, trust and

146

belonging fostered by their environment. This not only fed into their networking, with other co-op residents, but their networking, in turn, reinforced this community of sorts.

Therefore, the interviews illustrate the importance of child-related networks to the civic life of single mothers. While this was especially the case in Canada, child-related networks were also prevalent in Sweden and France. Not only do these associations build trust and reciprocity, and foster community, but they also often facilitate and are intricately linked to other forms of networking. Since these networks seem to throw single mothers and couples together in networks, it seems plausible that a larger number of parents – beyond this sample interviewed - rely on such child-related networks. This is all the more reason why such social ties must not be underestimated or devalued, but instead, should be included in standard civic engagement analyses.

The interviews confirm that care work, and the presence of young children, inevitably shapes the associational life of single mothers, most of whom are primary caregivers while also being active in the paid work force on a full time basis. The civic engagement pattern that emerges from the interviews suggests that Canadian and Swedish respondents have high levels of networking and trust, while those in France are considerably less active in formal networks, and demonstrating low levels of trust. Thus, at the very least, all single mothers interviewed were highly engaged in informal networks, although in different ways, with the

French preferring familial ties, the Canadians opting to network with other parents, and the

Swedes with a mix of family and friends.

The findings demonstrate that individual-level factors have an important impact on associational life, and lend support for three of the four explanations for the gender gap in

147

participation outlined in the preceding: the resource-, methodological- and agency-based accounts. The resource-based account, positing a positive correlation between having resources and being civically active, is vindicated here, as respondents’ struggle to balance work and family life was draining them of the resources which facilitate participation, such as time, money and energy. Indeed, respondents with lower incomes were found to form the majority of non-joiners. Nevertheless, the desire to participate espoused by Swedish,

Canadian and – to a lesser extent French - respondents motivated them to be active in a mix of formal and informal networks. The interviews also highlighted the merits of the methodological account of gendered participation which ultimately suggests that in order to gain a better understanding of the civic engagement of men and women, the concept of ‘civic activity’ must be broadened to include child-, care- and parenting-related groups. Indeed, the findings indicate that these often discounted forms of networking form an important part of respondent’s associational life, build trust, and act as a catalyst for further networking. As such, they should be included in the definition, measurement and analysis of civic engagement. Further, the lower levels of trust and networking reported by French respondents, coupled with their preference for tight familial bonds, provides evidence for the agency-based account of gendered participation which suggests that network homogeneity leads to narrow associational lives.

However, the interviews do not lend support for the orientation-based account of the gender gap in participation which suggests that women’s role as primary caregivers shifts their mental energies towards the family and away from politics and the community. The

Canadian case disproves this approach because while a high proportion of Canadian respondents say their social ties revolved around their children, they were among some of the

148

most civically active. Instead, single mothers in all three countries said that having children brought them much closer to their communities. Thus, the orientation-based account of the gender gap in participation falsely presumes women are locked a very real public/private divide where family and care-related networks are separate and somehow less valuable than other forms of sociability. Instead, the notion expressed by Herd and Meyer, as well as

Lowndes (Lowndes, 2006; Herd and Meyer, 2002) that children and care-related networks can act as a catalyst for participation resonates in the interviews. Further evidence against the orientation-based account can be seen in the findings about respondent interest in politics and voting behaviour. Contrary to the logic of the orientation-based account, 80% of

Canadian respondents, who were more likely to network with other parents and their children

- said they were interested in politics, while 60% said so in France and 75% in Sweden.

Additionally, 90% of Canadian respondents claimed they voted on a regular basis, while

60% said this in France and 85% in Sweden.

Finally, this chapter suggests that, in addition to individual-level factors, the state must also be considered for its role in shaping civic engagement. The findings highlight that the implementation of effective policies can help women reconcile the tension between work and family life which constrains their associational lives. This is reflected in the Swedish 50-

50 custody arrangements which an overwhelming majority of Swedish respondents asserted helped them balance their labour market activity, associational life and care work. Another clear example of the importance of state presence in the lives of single mother interviewees is the number of benefits awarded to respondents in Sweden and France which helps them compensate for their low stocks of resources, such as time and money. For instance, although more Canadian respondents belong to the higher income group, with 25% saying they earned

149

at least $70,000 after taxes compared to 5% of French respondents, and none of the Swedish respondents, no Canadian single mothers interviewed received any family benefits.1 In contrast, 80% of French and 90% of Swedish respondents received family benefits. In addition, 30% of both French and Swedish respondents received housing benefits, while no

Canadian respondent had such a provision. Additionally, 5% of French respondents and 45% of Swedish respondents received other types of benefits where Canadian respondents received nothing. This, moreover, does not account for state subsidized daycares and leisure centres, the Swedish ‘contact person’ system and such public programs and services geared to help relieve some of the family burden. In fact, in the case of Canadian respondents, it is the lack of state presence in terms of benefits and services which could explain respondents’ heavy reliance on networks with other parents and high levels of activities revolving around their children. The next three chapters explore the impact of the state in greater detail by looking at country-level factors which shape associational life.

1 Interviews were conducted prior to the implementation of the Canada Child Benefit which awards parents of children six years of age and under $100 taxable/month.

150

Appendix Interview Schedule

151

Questionnaire

A) DEMOGRAPHICS 1) Date of Birth:

2) Race/Ethnicity:

3) Where were you born (contintent)?

4) What is the highest education level you have attained?:

5) Number of Children and ages:

6) Other Dependents:

7) To which religious denomination do you belong to?: a) Apart from weddings, funerals or holidays/ceremonies, do you regularly attend church/temple/ synagogue? b) Do you consider yourself a religious person? c) Do you consider yourself a spiritual person?

8) What is your marital status:

9) I am interested in looking at a mix of people from different socio-economic backgrounds – with different incomes, etc. Therefore, I was wondering whether you could give me an estimation of your annual income before and/or after all deductions.

10) Are you currently on social assistance/welfare? a) If so, how do you feel about the welfare system?

11) Are you involved in any government programs or services other than welfare you can think of (pension, family benefits or childcare subsidy, employment insurance, housing subsidy, other)?

12) Are you working right now? a) If you are employed, what type of employment/occupation are you in? b) How many hours /week do you work? c) How flexible is your work?

13) Do you feel as though you have the choice to work in the paid labour market or to do unpaid work at home? a) If you had a choice, would you still choose to stay home or work?

152

14) How do you feel being a single mother and how do you think single mothers are perceived?

B) INFORMAL NETWORKS 1) People have social networks – that involve 1 or many people in their life that can be made up of family members, work colleagues, neighbours or friends, that they can lean on for support or whom they regularly interact with and can rely on. a) What types of people (family/colleagues/friends/neighbours etc.) b) How many social networks would you say you have?

2) How often do you visit with people who make up your social network?

3) What sort of activities do you do with people in your social network?

4) When getting together with friends or family, do you discuss politics (frequently, occasionally, never)?

5) Would you say that your social networks are mostly made up of men or women or both more or less equally, on average?

C) FORMAL NETWORKS 1) Do you belong to a group, organization or association? a) Which one/s would you say you are most active in or participate in the most – and what do you do? b) do you do unpaid work for them? c) Why did you join? d) How does belonging to this/these group/s make you feel?

2) What sort of benefits do you feel you are gaining from your membership in this group?

D) POLITICAL ENGAGEMENT 1) Would you say you are interested in politics?

2) Did you vote in the last election/do you regularly vote?

3) Have you, would you or would you never consider any of the following: a) signing a petition - b) joining boycotts - c) attending lawful demonstrations - d) joining unofficial strikes - e) occupying buildings or factories -

4) Do you think most people can be trusted?

153

5) I am going to name a number of organizations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all? a) Churches - b) armed forces - c) the legal system - d) the press - e) television - f) labour unions - g) the police - h) government in Ottawa - i) government in Ontario - j) political parties - k) parliament - l) the civil service - m) major companies - n) the environmental movement - o) the women’s movement -

E) CARE WORK 1) Would you say you are the primary caregiver of your child/children?

2) Are you in contact with the father of your child/children? a) If so does he visit? b) Does he provide financial support?

3) If you have a partner (not biological father) how involved is the partner with the childcare? a) Does he support to childcare financially?

4) Are your children in day care or afterschool programs? a) Is any of your childcare subsidized?

5) How often do you need childcare?

6) Why do you usually need childcare?

7) When you need childcare, who do you call or turn to first? Who is next?

8) Would you say that a lot of your social networks revolve around your children and/or their activities?

154

F) FOLLOW-UP Is it OK for me to contact you by phone if I need to follow-up/clarify or ask a few more questions?

G) SNOWBALL Do you know any other single mothers who you would be willing to approach to tell them about my research?

155

PART II: COUNTRY-LEVEL FACTORS

156

Chapter 4 The Impact of Welfare Worlds on Citizenship

Networking and the norm of trust are vital to democratic participation. The dominant civic engagement literature, devoted to outlining what associational life can do, has demonstrated that it can: build valuable civic skills and make important contacts; voice their concerns to decision-makers; develop political interest and knowledge; foster social cohesion by generating favourable norms like trust; and facilitate the solution of collective action problems. As Part I showed, social and economic inequalities tend to be reproduced in civic engagement levels. Inequality in participation is a threat to democracy because it can, for instance, result in a skew in the types of issues and concerns that are likely to be heard and addressed by decision-makers. A lack of networking and trust can also lead to a vicious circle in which communities get trapped in a cycle of distrust, apathy and social isolation which leads to social ills and is difficult to break out of. In light of its centrality to the democratic process, and all the benefits that can accrue from strong, active communities, it is important to understand how civic engagement works to establish who is in a better position to participate and why, and what can be done to generate civic norms and behaviour generally.

An important place to start is to consider the manner in which citizenship is conceptualized since, as the analyses in Part I demonstrates, this has important ramifications for the way associational life is understood. The preceding chapters have shown that the dominant civic engagement literature’s approach to citizenship, appeals to the civic republican and liberal traditions, both of which are rooted in formal equality. As such, it is designed to skip over how differences such as gender, race and class, as well as social and economic conditions, affect participation because it takes for granted that everyone is equal

157

in the sense that they are the same and are treated in the same way. However, the case of women demonstrated that, in practice, this is not the case. Despite having achieved formal equality in most respects, the findings in previous chapters suggested that women’s enduring social and economic disadvantage leads to lower levels of networking and trust. In light of this, a feminist approach to citizenship, founded on the notion of substantive equality, is called for because it can account for the impact such factors have on associational life.

Yet, in the effort to understand how civic engagement is shaped, focusing on individual-level factors such as class- and status-based inequalities alone is insufficient.

Such a limited approach fails to consider the effect that country-level factors have on citizenship and participation, or any potential interaction between levels of anlaysis. For instance, while an individual-level analysis highlights women’s disadvantages in citizenship and participation compared to men, a country-level analysis would reveal the difference in citizenship and participation between women in different countries as a result of divergent state policies; while some states implement policies that encourage traditional gender roles, others may provide a range of social provisions designed to help reconcile labour market activity and family responsibilities. This suggests the possibility of cross-level interaction between how state policies shape citizenship, and how the resulting experience of citizenship structures opportunities and resources at the individual-level is possible.

Therefore, in order to get a solid understanding on how civic engagement works, it is necessary to move beyond the uniquely individual-level account presented in preceding chapters to consider the impact of the state on civic life.

Accordingly, Part II shifts the focus from the impact of class- and status-based inequalities on participation to focus on the role of the state in shaping citizenship and

158

associational life, focusing on welfare states. While citizenship at the individual-level centres on the experience of inequality and how this structures everyday life including participation, at the country-level, it is approached as a political construct, created through the rights and entitlements of the welfare state. As the next chapters will show, states implement policies which influence citizenship in ways that promotes a vision of the polity that they deem preferable. For instance, by designing policies that either implicitly or explicitly reinforce or alleviate inequalities, states can construct very different types of citizenship which work to constrain or foster participation.

The purpose of this chapter is to explore how states structure women’s citizenship via the rights and entitlements of the welfare state. Drawing on Esping-Andersen’s welfare regime typology, it compares family, income security and housing policies in Canada,

France and Sweden to determine which is most favourable to women, focusing on equality outcomes. In each country, women’s citizenship status tends to differ based on how the state chooses to approach their roles as wives, mothers and workers, reflected in the rights and entitlements of the welfare state. The findings reveal that Sweden’s preoccupation with promoting equality has meant that it takes an active role in eliminating traditional gender roles, reflecting policies which offer greater flexibility and options to help women balance work and family life, while also encouraging men to take a more active role in caring for young children. France’s preoccupation with conserving tradition, social status hierarchies and the state’s power has led to policies which preserve traditional gender roles on the one hand, while justifying extensive intervention into the market and family on the other, by providing generous, comprehensive social welfare provisions. Canada’s preoccupation with minimal state intervention into the market and family has led to a system that does little to

159

address traditional gender roles. As such, it tends to leave women struggling to balance work and family duties. BEYOND THE INDIVIDUAL: THE STATE OF CITIZENSHIP It has already been shown that unequal citizenship status matters to associational life.

Not everyone in the community is the same or treated in the same way; differences such as gender, race and class exist and produce vastly different citizenship experiences. Social and economic context structures norms of trust and networking opportunities such that those who are worse off suffer from lower levels of civic engagement. Yet citizenship does not exist in a vacuum; the state has the power to decide who can claim rights and entitlements of the welfare state and, ultimately, who is included or excluded from the community.

Moreover, while disparity is not arbitrary, the state is in a position to either ignore, alleviate or reinforce the inequality which disproportionately plagues certain segments of the population through the rights and entitlements of the welfare state. This is something that has long been argued by the citizenship and welfare state literature, including the important work of T.H. Marshall, who draws a clear line between citizenship and the state. According to Marshall, the progression of rights and entitlements - starting with civil and political rights – would only be complete when social rights were also extended to citizens via the welfare state. The ‘welfare state’ generally consists of: a) the rights and entitlements which derive from social policies, but often overlap into other policy areas such as family, tax and labour policy; b) the manner in which resources are redistributed, for instance, eligibility criteria, and duration and amounts of benefits; and c) the outcome in terms of how it works to order society such as whether it targets some or treats all citizens equally. Strategies and discourses of the welfare state are shaped by an underlying logic that works to direct policy

160

implementation, design and outcomes. Welfare states which share similar practices and logics over time are referred to as welfare regimes (Esping-Andersen, 1990; Brush, 2002, p.163). Brush, for example, suggests that elements of regimes include signs, subjects, strategies, and sanctions - regimes as coherent systems through which people signify and contest meanings (Brush, 2002, p.163). Prior to twentieth century social citizenship,

Marshall argues that poverty relief via Poor Law was not an integral part of citizenship but an alternative to it. Claimants were stigmatized by being stripped of any citizenship rights because, to be eligible for assistance, paupers had to enter workhouses in which they were not only forced to relinquish civil rights of personal liberty, but also political rights, as they were disenfranchised (Marshall, 1992, p.15). However, the advent of the welfare state was accompanied by social and economic rights for all citizens to a minimum standard of living, culminating into an unprecedented, more complete citizenship status, following a long struggle for universal civil and political rights.

While elements of Marshall’s account of the struggle for citizenship from a few rights to a more complete citizenship that is linked to the rise of the welfare state may explain the general story of citizenship, this downplays the variation that exists across welfare states, including in their citizenship legacies. Marshall does concede that there is no universal citizenship ideal; that the rights and duties that make up citizenship are determined by the community, and thus vary over space and time (Marshall, 1992, p.18). However, his approach has come under attack for its Anglo-centrism, assuming the development of citizenship in Britain can be used as a template to explain the story of citizenship elsewhere.

For instance, Mann argues, “[t]here is one rather remarkable feature of [Marshall’s]

Citizenship and Social Class. It is entirely about Great Britain. There is not a single mention

161

of any other country…This certainly implies a general evolutionary approach, and indeed he does intermittently use the term ‘evolution’” (Mann, 1987, p.340). Others not only trace the divergent paths to full citizenship taken by different countries, but also show how the conceptualization of citizenship itself is greatly shaped by how the public and private arenas are defined (Turner, 1990). This is reflected, for example, in the difference between French and German citizenship. On the one hand, the French citizenship tradition stems from the active struggle from below which culminated in the French Revolution, and where an emphasis on the public arena led to an attack on the private arena, including the family and religious life. On the other hand, Germany’s strong, autocratic state tradition made for a more passive route to citizenship, granted from above, and where the private sphere was historically understood as a refuge from the great reach of the state (Turner, 1990). Scholars also show that the ‘evolution’ of rights from civil to social, as suggested by Marshall, was not always the case. For instance, social rights in Germany under Bismarck were awarded prior to the extension of many civil and political rights, introduced at this time as a ploy to maintain working class support for the regime and undermine support for Socialists (Mann,

1987). Moreover, many women were awarded social rights in their capacity as mothers and wives long before they were granted full civil or political rights (Sarvasy, 1997; Siim, 2000).

As such, scholars highlight the need to expand Marshall’s seminal citizenship story to include and better reflect the experience of other nations (Mann, 1987).

Finally, Marshall is also critiqued for his lack of attention to gender, and women’s complex relationship with regards to the market and the family (Gornick and Meyer, 2005, p.87). Since, for him, the main actors in society are representatives of social classes, and the citizen par excellence is male (Marshall, 1992, p.15), Marshall largely downplays other

162

forms of inequality, such as gender (O’Connor, 1998). He repeatedly links citizenship with paid work, but fails to acknowledge the ensuing gender implications (Sarvasy, 1997).

Moreover, while Marshall does recongize that women’s unequal citizenship status is

“peculiar,” he does not pursue the matter further in relation to women’s citizenship. For instance, although he declares that by the nineteenth century, civil rights were universal,

Marshall overlooks the way marriage contracts continued to limit women’s civil rights

(Sarvasy, 1997).

What is ultimately needed that Marshall’s approach is lacking is a model that goes beyond a simple relationship between state and citizen, to consider the way that the state, market, family and civil society interact with one another. Such a strategy, referred to by welfare state and citizenship scholars as the ‘welfare diamond,’ or the ‘responsibility mix,’ can illuminate the ways that variation in citizenship tradition translates into variation in the status of women across different states (Jenson, 1998, p.16-18). The relationships between the state, market, civil society and family are typically the result of long-standing legacies, and are reflected and reinforced in social citizenship (Janoski, 1998, p.16-18). According to

Janoski (Janoski, 1998, p.16-18), the size of each institutional sphere (state, market, community, family), as well as the degree of overlap between them, can provide a framework to compare divergent civil societies. For instance, the logic of relationships between the state/marke spheres that characterize different countries, and how these are mediated and reconciled, are reflected in the system of rights and obligations that each regime establishes, and the checks and balances that have developed – whether it be an understanding of the civil society as a check against the state, or the state as a check against the market.

163

Janoski demonstrates how institutions within each sphere interact to sustain divergent citizenship traditions across regime-types, mapping political theory and sociology onto different welfare state types (Janoski and Gran, 2002, p.18. Janoski, 1998, p.17-24,

108-112). Thus, he links Anglo-Saxon countries with liberal political theory (Locke, Smith,

Marshall, early Dahl, and Rawls) and pluralist political sociology which results in a form of citizenship where there is little consensus between self-interested citizens, whose freedom and privacy is protected from the state with an emphasis on individual rights (see table 4.1).

Under such a system, various interests are forced compete for power because they are kept at arm’s length from authority (Janoski, 1998). Western European countries are linked with civic republican and communitarian political theory (, Rousseau, Etzioni, Galston,

Oldfield, and Petit), and elitist political sociology which results in a form of citizenship where consensus between virtuous citizens is desirable, but where authority is elite-driven through business-dominated peak institutions (i.e. centralized labour and capital organizations) and experts, with little input from the grassroots, labour or other organizations (see table 4.1). Here, obligation to the ‘general will’ or the ‘good of the community’ ranks higher than rights – which tend to be universalistic (as opposed to individualistic) (Janoski, 1998). Finally, Nordic countries are tied to social or expansive democratic political theory (Habermas, Bohman, van Gunsteren, Barber, Warren), and neo- corporatist political sociology which results in a complex form of citizenship where consensus gives way to participation in the community, and where interest is aggregated into peak institutions that are integrated into deliberative institutions. Such a system strives to maintain a balance between the group and the individual, as well as rights and obligations

(see table 4.1) (Janoski, 1998).

164

Janoski’s approach brings to light the distinctive citizenship and civil society heritages that have developed, now firmly entrenched across different countries, and outlined in the work of welfare state scholars, most notably, Esping-Andersen. Esping-

Andersen, whose welfare regime typology dominates the comparative welfare state literature, elaborates on the way welfare regimes shape citizenship and address class and status-based inequalities, but goes beyond Marshall’s notion of social rights to consider divergent strategies for redistribution. Moreover, as the following section shows, he surpasses Marshall’s notion of social citizenship as a cure for inequality in society to examine just how social policy actually works as a system of stratification in itself (Esping-

Andersen, 1990, p.23).

THE ESPING-ANDERSEN TYPOLOGY Esping-Andersen’s approach suggests that we can understand why states cluster around regime-types (sometimes referred to as ‘welfare worlds’) by considering patterns of political mobilization and coalition structures, as well as the institutional historical trajectory of a regime (Esping-Andersen, 1990, p.29, 80). He categorizes welfare regimes, first, based on the degree to which they are ‘de-commodifying’ and ‘socially stratifying.’ De- commodification refers to the extent to which citizens are free from market dependency.

Thus, “de-commodifcation occurs when a service is rendered as a matter of right, and when a person can maintain a livelihood without reliance on the market... [it] entail[s] citizens can freely, and without potential loss of job, income, or general welfare, opt out of work when they themselves consider it necessary” (Esping-Andersen, 1990, p.21-23). According to this understanding, the mere presence of a social provision does not necessarily spell out emancipation of citizens from the market; the outcome of policies must be taken into account. For instance, a social policy can have a low de-commodification effect if its

165

benefits are too low or it stigmatizes those in need to such a degree that it forces recipients to turn to the labour market. Variation in de-commodification across regime-types is measured by combining the scores of a variety of variables for pension, sickness and unemployment cash benefit policies. Variables generally include eligibility requirements, built-in disincentives (i.e. wait periods for cash benefits and limits on duration of entitlements), as well as a benefit’s wage-replacement rate (Esping-Andersen, 1990, p.49-

54).

Social stratification, on the other hand, refers to the various ways in which states sort or rank citizens based on social policy traditions. For instance, some might strive to treat all citizens equally through universal programs, while others will opt for means-tested programs that emphasize social cleavages by classifying some citizens as ‘deserving’ and others as ‘undeserving.’ Considering the stratification of a regime should highlight the degree to which they amplify or reduce class or status differences. Social stratification includes a variety of measures designed to capture the specific dimension of stratification for each regime-type (i.e. the residual nature of liberal regimes, the occupationally-based selective nature of the corporate-conservative regimes, and the all-encompassing tendencies of social democratic regimes). Measures for each include: étatism as a percentage of GDP spent on pensions for government employees; as measured by the number of major occupationally-based public pension programs; the presence of means-tested benefits as the percentage of total social expenditures spent on means-tested poor relief; influence of the market as the percentage of private pensions out of total pensions; ‘universalism’ measured by the average of universal programs for sickness, unemployment and pensions;

166

and benefit equality measured by the gap between basic and maximum benefits for these schemes (Esping-Andersen, 1990, p.69-73).

Feminist scholars have challenged Esping-Andersen’s de-commodification and stratification concepts, arguing that they presume a male citizen and fail to account for women’s experiences. For example, the ‘de-commodification / commodification’ binary assumes that citizens spend most of their lives as paid workers in the labour market and that, when they are de-commodified (leave the labour market), their wages are substituted for welfare benefits. De-commodification in particular implies a certain freedom from market dependency. Yet, many women have a different citizen-profile, finding themselves outside of the paid labour market, or in part-time, seasonal, low-skilled work which is often not tied to the benefits that lead to de-commodification through welfare benefits in the first place.

Moreover, since much of women’s labour continues to be unpaid and takes place in the home, such as care work, it is not considered ‘commodification’ (Daly, 1994; Hobson,

1994). In addition, women’s comparatively inferior status in the paid labour market means they disproportionately make up the brunt of means-tested, meagre, stigmatizing welfare benefit recipients, as most of the more generous provisions are tied to occupational status

(Sainsbury, 1993; Orloff, 1993).

Some feminist analyses have examined how different welfare states work to reinforce or weaken the traditional male breadwinner model (Lewis, 1992; Fraser, 1994).

For instance, according to Lewis, Ireland and Britain represent a strong male-breadwinner model, where men and women are encourage to uphold traditional gender roles, with little or no support for women’s employment (Lewis, 1992). France, where the traditional family model remains, but women’s roles outside the home is somewhat more accepted than in

167

Britain or Ireland, exemplifies a modified breadwinner model. Here, policies and tax rules support women’s roles as mothers, while also offering provisions such as an extensive subsidized daycare system which facilitates female labour force participation. Finally,

Sweden typifies the weak breadwinner, or dual earner model, where the tax system and social benefits encourage men and women to be carers and workers (Lewis, 1992).

Other scholars have also shown that welfare states establish ranks between women as well, reflected for instance in the ways different types of families are treated, or the differences between married and single mothers, as well as widows (Millar, 1996; Orloff,

1993; Hobson, 1994). Highlighting how welfare states reinforce different gender roles for married and single mothers, Millar points to Scandianvian countries and, to a lesser extent,

France and Belgium, as examples of countries that encourage employment for all mothers

(Millar, 1996). Alternatively, Ireland, the Netherlands, Luxembourg, the UK and Australia represent countries that discourage employment for all mothers (Millar, 1996). Finally,

Southern European countries, along with Germany, Austria and the US, tend to encourage employment for single mothers, but discourage participation in paid work for married mothers (Millar, 1996).

In light of such gendered critiques, Esping-Andersen acknowledged that his original approach to studying welfare regime typologies did fail to address gender issues. He conceded that the concept of de-commodification “…may adequately describe the relationship between welfare states and the standard full-career male worker… [and] is not easily applicable to women considering that their economic role is often non-commodified or at least only partially commodified...[and that] the functional equivalent of market dependency for many women is family dependency” (Esping-Andersen, 1999, p.44).

168

Esping-Andersen also came to consider the trade-offs many women are forced to make in areas such as balancing paid and unpaid work; he points to a clear negative correlation between female employment rates and unpaid hours of domestic work, demonstrating the constraints women in particular face with regards to engaging in paid employment. For instance, in countries with a high level of female labour market participation, as in Denmark,

Finland, the US and Canada, unpaid weekly hours are relatively low - between 25 to 35 hours per week. Where women have lower employment rates, in countries such as the

Netherlands, Italy and Spain, they tend to work long domestic hours - 39 to 40 hours a week.

(Esping-Andersen, 1999, p.57-8).

In response to feminist critiques of his typology, Esping-Andersen (1999) turns to the concepts of familialization/de-familialization. The former refers to a system in which households are assigned the brunt of famialial burdens, while the latter refers to a system in which households are encouraged to turn to the welfare state or the market for relief of their familial obligations. He measures familialization by examining the percentage of the aged living with their children, unemployed youth living with parents, and women's weekly unpaid hours, while indicators for de-familialization include public spending on family services, the percentage of children under three years of age in childcare, and the percentage of elderly receiving public home help. (Esping-Andersen, 1999, p.51, 61-2; Bambra, 2004, p.203).

For feminist scholars, de-familialization is the answer to gender-blind de- commodification since, “…just as social provision can give workers leverage against the forces that compel them into the labour market, [de-familialization] potentially provides women leverage against the forces that relegate care work, childbearing, household

169

formation, and sexuality to normative femininity in the context of marriage” (Brush, 2002, p.164). However, feminist scholars have highlighted that Esping-Andersen’s particular take on de-familialization should be made to distinguish between de-familialization by public means via the state, and de-familialization by private means via the market - since these have very different implications for families and, hence, women. A market-based de- familialization system, where claimants must purchase privately-run services directly from the market (having to pay for childcare, for instance), is sure to aggravate class dynamics since only those with higher incomes will be able to afford the service – particularly compared to a system where the state subsidizes the service (such as a public daycare system) (Leitner, 2003, p.361; Bambra, 2004). Moreover, de-familializing countries that favour state-based strategies generally involve some intervention into the family that require direct action to relieve the care burden that tends to otherwise disproportionately fall on women’s shoulders. Alternatively, while they may not explicitly promote traditional gender roles to the extent that familializing countries do by creating policies that discourage women’s entrance into the paid labour market or encourage their roles as mothers, countries that de-familialize through the market nevertheless implicitly promote traditional gender roles – by refusing to set in motion the changes required to shift away from modes of social order that promote inequality (Leitner, 2003). Thus, it is important to consider the gendered effects of social and family policy when looking at de-familialization – both implicit and explicit.

Esping-Andersen evaluates the combination of de-commodification, de- familialization and social stratification along three regime dimensions, each rooted in distinctive historical legacies, reflected in their policies and politics: liberal, corporate-

170

conservative, and social democratic (see table 4.1). It is important to note that Esping-

Andersen’s regime-types are not meant to represent strict categories, but rather a group of countries that score most highly on, and thus tend to cluster around, a particular regime-type dimension. For instance:

The social democratic world comprises the five nations whose social insurance programs are most universalistic in coverage and homogenous in benefit level. The liberal world includes the five countries most marked by means testing and private (as opposed to public) health and retirement insurance. The conservative world is constituted by the five nations with the highest scores on a dimension tapping the degree to which social insurance programmes are differentiated by occupational and public-private status group distinctions (Hicks and Kenworthy, 2003, p.29).

Esping-Andersen’s regime typology is an ideal-type of approach; he does not claim to have found a system with categories which perfectly fit each state, routinely warning that there are no “pure cases” (Esping-Andersen, 1990, p.28-9, 49). Nevertheless, subsequent chapters demonstrate that countries are inclined to cluster around particular regime-types. A COMPREHENSIVE OUTLINE OF WELFARE REGIMES Based on Esping-Andersen’s approach, together with the work of other welfare state and citizenship scholars, a more comprehensive understanding of welfare states emerges; one which identifies broad patterns in polities at the level of policy choices and practices, and general institutional logic, corresponding to the way countries cluster around regime- types: liberal, corporate-conservative, and social democratic (see table 4.1). Since, in liberal welfare-regimes, which typically includes Anglo-Saxon countries like Canada, the United

States and the UK, the logic of the market prevails, the state is kept at bay by individualizing risk, and promoting market solutions. Thus, entitlements are targeted to those in abject need, and tend to be means-tested. This regime-type is connected to liberal political theory, emphasizing individual rights and championing the notions of ‘less-eligibility’ and ‘self-

171

help’ (Janoski, 1998, p.18, 20; Esping-Andersen, 1999, p.74-5). Citizens are conceptualized as autonomous agents that are passively distrustful or actively managing the state, and civic virtue is equated with personal accountability and industriousness (or reliance on the market over the state) (Janoski, 1998, p.226). As a result, social and class/status group rights tend to play a secondary role or are avoided (Janoski, 1998, p.18, 20; Jepperson, 2000, p.10-16).

Authority in the liberal polity is such that government is seen as an instrument of a society that is organized in a pluralist (as opposed to corporatist) manner, with actors lobbying the state, but not directly involved in the policy-making process. Consequently, the state, community and market spheres only overlap minimally. For instance, while there may be some overlap between institutions in the community sphere such as voluntary and welfare state organizations, such interactions are rarely found between the market and state spheres

(Janoski, 1998, p. 112-114). In the liberal regime-type, women are expected to participate in the labour market, reflected in their high rates of full-time employment and earning levels

(Moller, Misra and Budig, 2006, p.20. Orloff, 2002, p.14). Yet they continue to be seen as care providers, and since caregiving is viewed as a private matter in liberal regimes, there is often a lack of support for women’s informal carework, and few public childcare options.

Thus, emphasis is on private provision of care, both via unpaid carework, and care purchased in the market. Since de-familialization takes place through the market, class cleavages appear amongst women, with poverty hitting single mothers hardest (Moller,

Misra and Budig, 2007, p.7-8, 24, 27; Christopher, 2002, p.79). As such, feminist scholars have tended to categorize liberal regimes as supporting the traditional male breadwinner model, where women are seen as caregivers first and secondary earners (Lewis, 1992).

Moreover, since the little assistance available usually comes in the form of means-tested,

172

meagre poverty relief, this has the effect of forcing single mothers into market-dependency

(Millar, 1996).

Alternatively, corporate-conservative regimes, usually Western European countries like France, Belgium, Germany and Italy, tend to be moderately de-commodifying.

However, the influence of a strong Catholic legacy, reflected in a preference for the principle of subsidiarity - which suggests that neither state nor community should undermine the autonomy and responsibility of the family - as well as a great attachment to corporatist modes of social organization, manifests itself today in a system that continues to maintain social cleavages via occupational, insurance-based social security. ‘Corporatism’ here refers to the type of social stratification involved in the distribution and organization of welfare. It first emerged in modern Western Europe in reaction to the rise of , when the old feudal social order, including the petty bourgeois class, Catholic and other religious groups, as well as traditional intellectual circles, felt threatened by what was deemed the radical individualism and social disorganisation that accompanied capitalism. As the old social order collapsed and the state increasingly took over the social security functions that had been handled by intermediary groups such as mutual aid societies, guilds and corporations, the welfare state in essence restored this old social order the via status-differentiated, occupationally-based schemes and services (Esping-Andersen, 1990, p.24-5, 27, 38-41).

Rooted in communitarian theory, citizens in corporate-conservative regimes are conceptualized as bonded brothers and sisters, and civic virtue is equated with fraternal responsibility (Janoski, 1998, p.18-20, p.226). The high degree of étatism and elitism inherent in these polities, led by peak institutions with influential experts who are directly involved in the policy process, is often held responsible for their smaller public spheres (e.g.

173

voluntary organizations) (Esping-Andersen, 1990, p. 109; Schofer and Fourcade-

Gourinchas, 2001, p.819, 823; Salamon and Wojciech Sokolwski, 2001, p. 17, Appendix

Table 2, p.25; Curtis, Grabb and Baer, 1992; Curtis, Baer and Grabb, 2001, p.802). Despite exhibiting lower levels of civic engagement, the state has maintained some control over the nonprofit sector and over key social elites there, while satisfying demands for social welfare.

This is exemplified in a highly subsidized third sector that has grown over time alongside government spending (Salamon and Anheier, 1998, p.229, 239, 242-43). Worms, for instance, points to the growth of associational life in France alongside welfare state development (Worms, 2002, p.144). Society, organized along corporatist lines, emphasizes

‘the nation’ rather than ‘the individual.’ Either the state or business will have greatest influence on society – elites tend to play dominant role and mass participation is secondary

(Janoski, 1998, p.112-14). The combination of corporatism and elitism results in a civil society where the state routinely merges with the market and community spheres.

Familialism is strongly attached to Catholic social teachings, specifically – the principle of subsidiarity, which holds that the state and community should not undermine the autonomy and responsibility of the family; in other words, “the state should not do what the family can do...” (Handy, 1992, p.64). This principle goes hand in hand with the corporatism of Western European countries, where the positioning of guilds and fraternal organizations so close to the family unit made them better equipped to serve their needs compared to the remote central state. (Esping-Andersen, 1990, p.61) Thus, this strong belief in the principle of susidiarity and influence of corporatism, help explain why familialism prevails in corporate-conservative countries. Familialism is reflected in the historical presence of

174

pronatalist strategies (Esping-Andersen, 1999, p.85-6).1 Moreover, women continue to be seen as caregivers and are rewarded for their care work, for instance, through care allowances and extensive childrearing leaves which may weaken their attachment to the paid labour market (Moller, Misra and Budig, 2007; Ostner, 1996, p.6-7; Orloff, 2006, p.13-14).

This explains the greater gender gaps in employment and earnings found in many corporate- conservative regime countries, and the disparity in poverty rates between single and non- single mothers (Moller, Misra and Budig, 2007; Kilkey and Bradshaw, 1999, p.177; Moller and Misra, 2005, p.26. Christopher, 2002, p.78-80).

However, some corporate-conservative countries, such as France and Belgium have also established strategies that encourage women to enter the labour market – by providing public or subsidized childcare and parental leaves. In these cases, such policies have alleviated gender gaps in employment and earnings, and reduced poverty rates among both single and married mothers alike (Kilkey and Bradshaw, 1999, p.174-75; Moller and Misra,

2005, p.21-24). Thus feminist scholars tend to highlight a divide between corporate- conservative countries, such as Germany that, on the one hand, support a strong breadwinner family model, providing little support for women to enter the labour market and assistance to lone mothers; and France and Belgium, on the other hand, which provide a mixed or modified breadwinner model, with a strong support for women’s caring roles and labour market activity (Lewis, 1992; Millar, 1996).

In contrast, the social democratic regimes, which typically consist of Nordic countries, such as Sweden, Denmark and Finland, are inclined to be highly de-

1 According to Heitlinger, pronatalism “…implies encouragement of all births as conducive to individual, family and social well-being… [it] can be seen as operating on several levels…The cultural, and especially the ideological, meanings of pronatalism have been historically associated with right-wing coercive , laws and practices that overemphasize natalist goals, reinforce the traditional family model of father as breadwinner and mother as homemaker, and severely limit reproductive choice” (Heitlinger, 1991, p.344-5).

175

commodifying, characterized by universal benefits. Such regimes correspond to an egalitarian-driven social democratic theory that strives to address discrimination against class/status groups through the expansion of rights (Janoski, 1998, p.20). Accordingly, citizens are conceptualized as neighbours and co-workers that are cooperative and involved, and civic virtue is equated with reciprocal group responsibility (Janoski, 1998, p.226).

Despite the state’s prominent role, authority in social democratic polities remains vested in society that is structured along a corporatist decision-making structure. This variant of

‘corporatism,’ sometimes referred to as ‘neo-corporatism,’ emerged in the post-war period but blossomed in the 1960s and 1970s and is often associated with Scandinavian countries.

It refers to a system of interest organization - a sort of ‘institutionalized collaboration’ in which employer organizations, labour unions and the state are involved in a tripartite decision–making relationship (Birnbaum, Adler and Adler, 1982; Schmitter, 1974). Thus, interest groups have a direct or informal hand in shaping policy through a consensus- building process (Janoski, 1998, p.109). Social democratic regime-types are characterized by great overlapping spheres, especially between the state, community and market spheres, and a vibrant third (or nonprofit) sector (Thomas Janoski, 1998, p.112-14; Jepperson, 2000, p.10-16; Schofer and Fourcade-Gourinchas, 2001, p.819, 823; Salamon and Sokolwski,

2001, p. 17, Appendix table 2 p.25; Curtis, Baer and Grabb, 2001, p.802). This is consistent with a history where working-class elements were likely able to exercise some political power in alliance with other social classes (Salamon and Anheirer, 1998, p.228-9, 241-2).

Since men and women are viewed as both earners and carers, policies support women’s participation in the labour market (flexible work weeks, public , generous leave), as well as men’s participation in caring (e.g. paternity leaves) (Moller, Misra and Budig,

176

2007). Work-family reconciliation have helped reduce poverty rates for both married and single mothers (Moller, Misra and Budig, 2007; Kilkey and Bradshaw, 1999, p.173-74).

Feminist scholars concur that social democratic countries tend to be the most supportive of women’s citizenship, weakening the traditional breadwinner family model in favour of a dual earner family model, all the while providing support for lone mothers as well (Lewis,

1992; Millar, 1996).

Thus, Esping-Andersen’s welfare regime model highlights how, informed by their respective historical legacies, states differ with respect to their relationship with the market, family and civil society, resulting in varied approaches to citizenship. The next section demonstrates how this has led to divergent approaches to women’s status, gender roles and family models. COMPARING FAMILY POLICY IN CANADA, FRANCE AND SWEDEN

Social policies in each regime are compared in this section to determine the degree to which they were designed to curb inequalities – particularly in relation to women’s status. It is particularly vital to consider both de-commodification and commodification in the case of women since the degree to which states support women’s participation in the labour market has been particularly important to the goal of gender equality. Historically, the obstacle to women’s equality was less about their degree of freedom from the market, as Esping-

Andersen’s ‘de-commodification’ concept suggests, and more about seeking relief from the devalued and invisible work taking place in the home via the market (commodification). In other words, there is a link between women’s autonomy from the family and unpaid caring work on one side, and their ability to participate in paid labour on the other. This view is

177

supported, for instance, by studies that report a negative correlation between the presence of young children in the home, and women’s decision to enter the paid labour market (Gornick and Meyer, 2005, p.47).

This review pays particular attention to social policies which shape women’s ability to balance paid and unpaid work, including: family policies2 which support women’s participation in the paid labour market (such the provision of public or subsidized childcare, and parental leaves), and women’s care work (such as family or children’s allowances, and childrearing leaves); housing policy; as well as income security programs (social assistance or allowance), that have long contributed to the family economy (Jenson and Thompson,

1999, p.1; Palley, and Bowman, 2002, p.345-6).

1) Family Policies Liberal welfare states such as Canada typically shy away from heavy government intervention in the family, preferring instead to emphasize the importance of self-reliance, individualism and privacy (O'Hara, 1998, p.7). This is reflected in Canada’s historically weak family policy, characterized by a very small collection of meagre, generally means- tested benefits and a preference for market solutions (Sylvie Morel, 2002, p.24-7; Strong-

Boag, 1995, p.126). This approach means that families are forced to fend for themselves, inevitably aggravating social stratification in civil society between those who can afford to purchase benefits, and those who cannot.

Alternatively, France is often pointed to as an example of a state with a generous and well-developed family policy. Yet a review of its social security history reveals that the lingering imprint of French étatism and gender-biased familialism created instead a more

2 Palley and Bowman provide a good definition of family policy: "social policies contributing to the well-being of caregivers and children... defined broadly to include all government benefits that contribute to the health and well-being of families with children through direct cash transfers… tax expenditures … in-kind benefits… and direct services...” (Palley and Bowman, 2002, p.345-6).

178

contradictory system. Women’s seminal roles in the family as wives and mothers have long been a target of French state intervention. For instance, the 1804 and 1939 Family Codes

(Code de la Famille) reinforced gender roles and the traditional family under the Napoleonic civil code, such as the dominion of the father over the family (patria potestas) highlighting women’s lack of civil rights (Nord, 1994, p.827-9; Siim, 2000, p.47-8; Crompton, 2000). In addition to the family, institutions of the ancient régime such as the Catholic Church have also been a target of the state. The state weakened the Church through a form of systematic secularization (laïcité), culminating in the disestablishment of the Church in 1905, formally separating Church from state, and making religion a strictly private concern (Mabille and

Valasik, 2004, p.237). As a result of such étatism, French family policy is intertwined with conservative familialist and pronatalist3 concerns, as well as the historical tug-of-war between the state and Church in the area of social welfare. To weaken the Church’s power, an ardent ‘state familialism’ emerged, “... a form of defense of family interests that …[can be distinguished] from 'Church familailism,' in the sense that it is based on science rather than religious values, and puts forward the protection of families by means of different types of social provisions through a 'bureaucratic management' of families…” (Revillard, 2006, p.5-6). Many familialist policies have been pronatalist in nature, with the state using scientific studies to justify encouraging large families, such as the finding that France was undergoing a population crisis.4 While today’s French family policy offers ‘choice’ for parents and supports women’s employment, exemplified in highly subsidized childcare

3 For more on pronatalism, see Heitlinger, 1991, p.344-5. One example of French policies that may be considered pronatalist and familialist in nature is the 1939 allocation de Mere au Foyer (Homemaker’s allowance) which became the Allocation de Salaire Unique (single-pay allowance) of 1941-1978, paid to the husbands whose wives were not waged workers. It provided benefits for homemakers to care full-time for children and, along with family allowances, represented 90% of a workwoman’s income for a family with two children in 1947. As a result of this policy, the participation of mothers with two children in the labour force diminished from 23% to 17% from 1946-1954 (Revillard, 2006). 4 For examples of ‘state familialism’ see supra note #3.

179

services, policies that encourage the traditional family nevertheless endure. For instance through the ‘family quotient,’ the French income tax system promotes horizontal equality between families (from those without children to those with children), favouring large traditional families (Fagnani, 2005, p.41; Dang and Monnier, 2004, p.696; Avdeyeva, 2006, p.45; Girard, 1994, p.582). It is designed in such a way that “…households that have the same number of mouths to feed are to be taxed equally, regardless of the distribution of income within the household.” (Saint-Jacques 2009). In addition to this, France uses joint taxation of couples (rather than individual taxation) in the calculation of income tax, a system that encourages women to stay home because tax benefits are greater when the gap in earnings between the first and second earner are wide.

Conversely, with an understanding that the state should play a key role in supporting families, an explicit Swedish family policy was forged, endorsed by the Myrdals, that suited the Social ’s goal of full-employment.5 A generous Swedish family policy developed, centred on promoting gender equality not only by endorsing female labour force participation through subsidized childcare and flexible leaves, but also with measures designed to encourage male caregiving. For instance, Swedish parental leave reserves specific days for fathers, and has resulted in comparatively high take-up rates (O’Hara,

1998, p.15-7). Unlike France and Canada, part of the relative effectiveness of introducing the Swedish approach to family policy might be explained by the degree of consensus about

5 During the early to mid-twentieth century, a swell of ideas from intellectuals about the best way to modernize Sweden was unravelling, reflected in the work and influence of Alva and Gunnar Myrdal. The Myrdals highlighted Sweden’s low fertility rate, but rather than recommending a population policy that reinforced a traditional family model, they proposed a ‘voluntary parenthood’ approach, where the state, through financial incentives, would encourage families to reproduce without hampering women’s rights. The Social Democratic Party or Socialdemokratiska Arbetarpartiet (SAP) began following this course of action, enacting laws, starting in the late 1930s and into the 1950s, that provided assistance to single-mothers, created housing and childcare benefits and allowances, and anti-discrimination policies in employment (Ginsburg, 2002, p.213; Haas, 1996, p.48-9; Holmwood, 2000, p.35-42).

180

the goal of gender equality and the role of the state’s duty to relieve some of the burden held by families through generous benefits.

A) FAMILY AND CHILDREARING LEAVES Family leaves, including those reserved for mothers (maternity), fathers (paternity), to be shared or split between mother and father (parental), as well as ‘sick child’ leaves, have been shown to encourage women’s labour market attachment by giving parents the time and earnings replacement to care for their newborn or sick child, while ensuring job security (Janet Gornick and Meyer, 2005, p.48). One exception to this is what scholars refer to as ‘childrearing leaves’; these can be distinguished from other leaves by how extensive they are (typically lasting more than one year per child); their lower, flat-rate benefit level; and in the fact that they are sometimes means-tested and may not guarantee job security

(Morgan and Zippel, 2003, p.52-3). Unlike other leaves, childrearing leaves are said to reduce women’s incentives to return to work by weakening their skills, limiting career paths and opportunities, and reducing earning levels (Jaumotte, 2004, p.65).

In Canada, provincial provide unpaid maternity and parental leave as well as caregiving or sick leave while the federal government provides paid leaves. In both cases, these benefits are part of broader employment insurance or labour legislation. As table

4b.1 shows, until recently, family leave in Canada excluded certain categories of employees such as the self-employed. However, the Fairness for the Self-Employed Act (December

2009) will make such benefits available for the self-employed starting in 2011. Canadian employment insurance also makes eligibility dependent upon a certain minimum of hours worked which tends to exclude part-time workers (see table 4a.4 for provincial leave policies). Moreover, duration and amounts of leave remain limited. Thus, while maternity and parental leaves have been extended and improved over time, paternity and ‘sick child’

181

leave – when it is available - remains limited (typically around four days) and unpaid.

Moreover, only five provinces offer some form of ‘sick child’ leave and with the exception of Quebec, all are unpaid (see table 4b.1).

Alternatively, France’s family leaves are available to all employed persons and amounts tend to be quite generous, for instance with 100% replacement rate for maternity and paternity leave. Parental leave, encompassed in the Complément de libre choix d’activité

(CLCA) which is part of the Prestation d’acceuil du Jeune Enfant (PAJE) (replacing the

Allocation Parentale d’Education (APE) in 2004), are available for two years for a single child, four years for a parent with two children, and up to five years for three of more children. The CLCA is a flat-rate benefit requiring a minimum number of years worked to be eligible. Since these leaves last longer than one year, they qualify as childrearing leaves, echoing France’s pronatalist roots by reinforcing traditional gender roles, and doing little to encourage male caregiving. Recipients of these extended French leaves are disproportionately low to middle-income women, working largely in blue-collar and clerical

(as opposed to managers or professionals), part-time positions. Furthermore, since the benefit level is meagre, beneficiaries of the French childrearing leave tend to be attached to a spouse that has a stable, well-paying job. In 2002, 98% of beneficiaries were women and about 80% chose to take the full benefit (Périvier, 2003, p.73-75, 267; Burri, 2003).

Finally, Swedish family leaves are available to all citizens, irrespective of employment history or status. Flat-rates are available for those who are not employed, while benefits are set at a generous 80% wage-replacement rate for those who are employed (see table 4b.1). In addition, Swedish parental leave is extremely flexible. It can, for instance, be taken in blocks of time until the child is eight years old. Parents are also entitled to 60 ‘sick

182

child’ days per child per year at an 80% wage replacement rate. In addition, parental leave allots specific, non-transferable ‘mommy’ and ‘daddy’ days to encourage a sharing of childcare duties. Therefore, comparing overall eligibility criteria, relative benefit amounts, flexibility of the benefit and duration times, Canada’s leave policies appear to be the least beneficial to women compared to those of France and the highly egalitarian policies of

Sweden.

B) CHILDCARE6 Since women continue to shoulder a disproportionate share of unpaid care work, even as female labour force participation rates have soared, the availability of affordable, quality childcare is of great importance to women’s status. On the one hand, costly childcare might discourage women from entering the work force as it “…may be viewed as a tax levied on mother’s hourly wages” (Gornick and Meyer, 2005, p.47-8). On the other hand, liberal welfare regimes have high rates of female labour force participation, despite low levels of public or subsidized childcare, highlighting that state-run day care and early learning systems are not a prerequisite for women’s labour force participation (Sarvasy,

2001, p.8). Nevertheless, while a market-driven child-care system may not directly impede women’s labour force participation, it exacerbates social stratification and inequality between higher and lower income mothers. Thus, in order to minimize disparities, the

6 There are many different forms of childcare. They include: parental care in the home, when a parent, typically the mother, takes care of child in the home; informal care services, when friends or family provide unregulated care at little or no cost; formal care services may not only involve daycare centres or family daycare in private homes providing regulated care, but also the unregulated care provided by a nanny; and finally, pre-primary school. Public care refers to childcare that is administered by the state and can thus be distinguished from publicly subsidized childcare, where the state funds but does not provide the services. Private childcare services are those not run by the state, but managed independently, and can include commercial (for-profit) and non-profit care. Childcare subsidies are usually targeted at formal day care and pre-primary school (Cool, 2004).

183

availability (hours of operation) and accessibility (cost) of quality childcare must be considered.

As tables 4a.6, and 4b.2 show, in the liberal regime of Canada, where the state has historically deferred to the market, private childcare services prevail. Thus, quality childcare will only typically be a viable option for higher-income households since parents are expected to pay about 50% of the cost of childcare. Outside of tax deductions and a few provincial tax credit measures, childcare expense subsidies are typically targeted to lower- income families. Québec is the only province that has initiated a lasting, affordable ($7/day) childcare scheme (for details, see tables 4a.6, and 4b.2). In 2001, British Columbia initiated a four-year childcare plan that established, among other things, $7 a day for before and after school care, and $14 a day for full day care. However, this program was quickly terminated

(Kershaw, 2004, p.931, 933; Coalition of Child Care Advocates of BC, 2004). As a result, until children enter full-day public kindergarden at age five (or four in Alberta, Manitoba,

Saskatchewan and Quebec), the majority of Canadian families (about 76%) with children under four or five are forced to send their children to less expensive, unregulated facilities

(OECD, 2003, p.66). Moreover, there is no comprehensive leisure centre system in place to care for children when care facilities or schools are closed.

Early childcare and education in France fall under the auspices of the state including the Social Services, Employment and Solidarity Ministry (Ministère des Affaires sociales, du Travail et de la Solidarité) for infants, and the Education Ministry (Ministère de la

Jeuness, de l’Éducation nationale et de la Recherche) for children three years and up, resulting in a more accessible, affordable and, hence, egalitarian system (OECD, 2006, p.326). Parents pay about 27% of the costs of childcare for infants 0-3 (crèches), while

184

children above three attend preschool (école maternelle) for free. Moreover, there is a well- developed system of leisure centers available to care for children when school is closed which is especially important in France because schools close in the middle of the week

(Wednesdays). Thus, while France does demonstrate a degree of familialism and pronatalism, adhering to a corporate-conservative logic, it has nevertheless developed a comprehensive public and subsidized childcare and early education system. This contradiction can be explained by looking to French étatism - childcare as the byproduct of the state’s continuing project to replace the Church in its role as provider of welfare and education by developing full-day preschools in the late nineteenth century. By the 1960s and

1970s, the state had implemented of a comprehensive system of daycares and preschools, which persists today (Morgan, 2003, p.271. OECD, 2006, p.328). Despite this highly organized public early childhood education system however, the majority of French parents

(about 64%) care for their infants in the home until age three, reflecting France’s conservative, familialist legacy.

As in France, childcare and early learning services in Sweden are administered by the state. Facilities for children 1-6 are under the auspices of the Ministry of Education and

Science (OECD, 2006, p.409; Sarvasy, 1997, p.8-9). However, they are more heavily subsidized, with parents paying only about 9% of the cost of preschool (forskola) for children aged 1-4 (or free for the low-income), and free of charge starting as early as age 4.

Thus, most Swedish parents rely on centre-based facilities, which typically offer full-day care. In addition, about 50% of children aged 3-12 are enrolled in leisure centers which care for children when school is closed.

185

In light of this review of childcare policies, it seems that Canada’s lack of a comprehensive state run or subsidized daycare or leisure centre system exacerbates social cleavages between those who can afford to buy social provisions from the private sector, and those who cannot. France has developed a rather sophisticated daycare and early learning system. However, familialist traditions linger and, ultimately, Sweden’s services are most affordable and accessible, making them the most egalitarian for women.

C) FAMILY/CHILD BENEFITS AND SERVICES Allowances geared towards families with children have the capacity to help increase household incomes and reduce poverty. In this sense, they may serve to lessen social stratification by promoting equality. As table 4b.3 shows, Canadian family and child allowances have been meagre, and targeted to those most in need (see table 4a.3 and 4a.5 for provincial data) (Jenson and Thompson, 1999, p.3; O’Hara, 1998, p.25-6, and 29; Graham and Al-Krenawi, 2001, p.414). Since the state has historically distanced itself from involvement in family matters, federal and provincial family benefits are often delivered though the tax system (deductions or credits) rather than through direct cash transfers, exemplified in the Canada Child Tax Benefit which replaced what was once a modest but universal family allowance.7 Although the federal government initiated a universal allowance in 2006, the Universal Child Care Benefit, for all parents with children under age six, the payment just $100/month/child and is taxable. In addition, unlike its European counterparts, Canada has no guaranteed maintenance program to compensate custodial

7 Canada’s 1944 family allowance payments amounted to approximately $5.95 per month, or 5% of an average monthly income of a family with two children at that time. Jane Jenson, et.al, 1999, p.3. It was allowed to erode over the years and was increasingly aimed at low-income families, before being terminated and replaced in 1993 by the income-tested, tax-based Canada Child Tax Benefit (CCTB) – part of the National Child Benefit (NCB) (O’Hara, 1998, p.25-6, 29; Graham and Al-Krenawi, 2001, p.414). The NCB also includes the National Child Benefit Supplement.

186

parents when the non-custodial parent fails to pay maintenance. Canadian provinces only provide child maintenance enforcement systems. However, the Government of Quebec may, under limited circumstances, provide an advance not exceeding $ 1500, for a three month period. Advances may only be considered in cases where the debtor or non-custodial parent is late with maintenance payments but where it is certain they will pay (R.S.Q. P-2.2, 2002).

In contrast, France offers a myriad of family benefits tailored to meet the needs of a variety of family types. The basic family benefit (Allocation Familiales),8 which is not non- means tested, amounts to approximately $180 CAD per month (the benefit increases with the number of children). However, remnants of France’s pronatalist policy remain; for instance, the basic family benefit is only available with the birth of the second child.

Furthermore, the state provides a birth grant, as well as means-tested supplemental benefits to large families, identified by scholars such as Heitlinger as pronatalist (Heitlinger, 1991, p.352). Unlike Canada, France has established a guaranteed maintenance cash benefit

(Allocation de Soutien Familial) to compensate custodial parents (about $160 CAD/month) when non-custodial parents default on their maintenance payments.

All Swedish families with dependents under age 16 and, in certain circumstances, even over 16, are eligible for a children’s allowance. The benefit is about $139 CAD per month per child. It is set at a flat rate, with supplements available for larger families. The

Swedish state also provides innovative services to struggling families such as the ‘contact person’ program, teaming parents with carefully selected ordinary citizens who are paid a stipend to provide support, including childcare. Moreover, like France, Sweden has

8 The Allocation Familiales, one of France’s earliest family benefits, started in 1913 to benefit large families, was expanded in 1932 to include the families of commercial and industrial workers with two children, and was made available to the general public starting in 1938.

187

established a guaranteed maintenance program (Underhållsstöd), compensating custodial parents when non-custodial parents fail to pay support (about $180 CAD/month).

In view of the above, while Canada tends to offer few benefits, those it has developed tend to be meagre, targeted and largely tax-based, avoiding direct state intervention in the family. On the other hand, France has a variety of family benefits that are quite generous, designed to accommodate different family types. However, to this day, family benefits reflect France’s historically, pronatalist agenda. Thus, compared to Canada and France, Sweden’s family benefits and services appear to be the most egalitarian for women, offering all-encompassing generous provisions, as well as innovative services designed to ensure all citizens meet their needs as parents.

2) Housing Policy Safe, adequate, affordable housing is a basic component of an adequate standard of living and personal autonomy. This is acknowledged in article 11 of the International

Covenant on Economic, Social and Cultural Rights, article 25 of the Universal Declaration of Human Rights, as well as article 31 of the Revised European Social Charter (Reitsma-

Street, Schofield, Lund and Kasting, 2001). As a result of family breakdown through divorce, separation or domestic violence, as well as disadvantages and a precarious relationship to the labour market stemming from their dual roles as primary care providers and paid workers, many women face barriers to safe, adequate or affordable housing (Edgar and Doherty, 2001, p.3-4, 15).

As table 4b.4 suggests, Canadian housing policy has tended to promote home ownership, and is marked by a strong desire to protect individual property rights. Programs thus treat housing as a commodity rather than as a social need or right (Reitsma-Street, Lund and Kasting, 2001). Since housing in Canada is generally seen as distinct from the wider

188

social security system, it is treated as a labour policy tool to stimulate construction employment in the private sector. For instance, the first housing scheme was introduced in

1919, following WWI, to boost employment. This was repeated in the aftermath of the Great

Depression, with the introduction of the 1935 Dominion Housing Act, and the 1938 National

Housing Act, and at the end of WWII, with the 1944 National Housing Act (Prince, 1995, p.726; Wekerle, 1997, p.173). The limited federal housing initiatives of the 1960s and

1970s, such as the Cooperative Housing Program, were eventually downloaded to provincial and municipal governments by the 1980s and 1990s and funding for new social housing was stopped altogether for a seven year period, from 1993 to 2000.

Canada offers no right to housing, and aside from Québec’s minimal (up to $80 a month) means-tested housing benefit targeted to low-income families, offers no stand-alone housing allowance scheme. Instead, the state has addressed housing problems via meagre, means-tested supplements that typically refer to the housing component of social assistance schemes. These welfare ‘shelter allowances’ are notorious for failing to cover the cost of rent, forcing many recipients to tap into other means to make ends meet, such as drawing from that nominal portion of welfare that is meant to cover other basic needs, such as food

(Reitsma-Street, Lund and Kasting, 2001, p.62; Wekerle, 1997, p.175; Prince, 1998, p.828,

841). As a result, social housing in Canada is relatively small, constituting about 6% of the total housing sector, and it is generally understood that there is a housing crisis. For instance, Reitsma-Street et.al state: “[t]he number of people with no secure housing or with chronic housing problems has reached record levels comparable to the Depression era.’

Since 1993, the crisis has worsened, with mayors of large cities calling homelessness a national disaster” (Reitsma-Street, Lund and Kasting, 2001, p.6).

189

French housing policy, alternatively, reflects its corporate-conservative legacy, with a tendency towards statist government intervention in the name of social solidarity – even despite decentralization laws first passed in 1983. No matter what phase of French housing policy one examines, whether it be the post-war phase of active re-building and development; the 1960s and 70s with its emphasis on urban renewal of inner-cities and the rehabilitation of social housing high-rise estates in the suburbs; the post-1973 economic crisis period which saw the introduction of means-tested housing allowances to help a population suffering through high unemployment levels; or the more recent interest in devolution of power and increased efficiency - the French state has made its presence known

(Aubert and Fribourg, 2005, p.9; Blanc, 2004, p.286-7). Today, social housing makes up about 17% of total French housing, and a high demand continues for low-rent spaces

(Fitzpatrick and Stephens, 2007, p.14, 25-6). Poorly built social housing, as well as the contradictory effects of immigration and social policies, have resulted in a growing and enduring ‘social exclusion’ problem in the suburbs, la banlieu, where low-rent, high-rise estates are located (Maurice Blanc, 2004, p.290-3).9 For instance, on the contradictory effects of the French housing allowance, the Aide Personnalisée au Logement (APL), Blanc notes:

In the social housing estates, households with incomes slightly above the official poverty level have no right to APL benefit, so that their rents increase greatly (often from 40 to 50 per cent) and they attempt to move out. In contradiction with its aims, APL produces a new type of segregation, concentrating the poor in modernized (but

9 The concept of social exclusion, which is French in origin, was first used in the 1970s to describe the large number of people excluded from social insurance entitlement. Prior to postindustrialization, the view promoted in France was that social risks were randomly distributed – but with the high levels of long-term unemployment starting in the 1970s, it became evident that risks were not so randomly distributed – that certain groups of people were much more at risk than others – namely, immigrants, the poorly educated, youth, and women. The policy response to social exclusion is ‘social insertion’ which aims to bring the disadvantaged back into society – in line with the notion of social solidarity (Béland and Hansen, 2000).

190

still socially devalued and stigmatized) social housing estates (Maurice Blanc, 2004, p.290-3).

Similarly, with regards to the paradoxical effects of ‘urban renewal’ strategies, “[i]n some circumstances, modernization strategies centered on the community have produced the unexpected result of reinforcing isolation and stigmatization of social housing estates”

(Maurice Blanc, 2004, p.290-3). Finally, Blanc asserts that changes to French immigration laws had the unexpected effect of turning what were otherwise foreign, seasonal, workers - into immigrants looking for long-term housing; foreign workers could no longer risk leaving

France for fear of being refused re-entry later. This, combined with considerable, systemic, discrimination, led to the segregated, low-rent estates in the outskirts of large urban centres

(Maurice Blanc, 2004, p.290-3). Although France has a variety of housing allowances, tailored to meet the needs of various types of recipients, from those with low incomes to large families, these are means-tested. Moreover, France has instituted the right to housing for ‘high priority’ homeless persons who have waited a prolonged period of time for housing.

True to its social democratic roots, housing policy in Sweden developed as a part of broader welfare state initiatives, and works above all to equalize the housing market to ensure that all citizens have access to adequate housing. Thus, housing policy has been approached as a social right and, as such, has obligated the state to work to secure this right for individuals.10 This approach has necessitated a historically strong state presence and, since the 1930s, when housing policy was first initiated, there has been considerable state intervention on many levels, including in financing, through long-term government loans

10 “The Swedish political document in which the policy theory of state correctives to the housing market has been spelled out most clearly is the Cabinet Proposal of 1974… Swedish housing policy is said to be based on the idea that good and commodious housing in a good environment is an ‘indispensable social right,’ and it is the responsibility of ‘society’ to guarantee security in the housing field” (Bengtsson, 2001, p.260).

191

with interest rate guarantees started in 1941; various forms of rent-control, first with rent- splitting, or hyressplittring, from 1942 to 1968, followed by the establishment of the ‘utility- value’ system, or bruksvärdessystemet, from 1968 until the present; housing allowances, first started in 1948 and endure to this day; as well as massive housing building and renewal projects. For instance, under the ‘million program’ in mid-1960s,the state created a plan to build a million housing units over a ten-year period. Moreover, during the 1980s, the

Swedish state engaged in a large-scale housing modernization strategy (The Swedish

Institute, 1996; Ditch, Lewis and Wilcox, 2001, p.20; Kemeny, 1981, p.95-7). ‘Social housing,’ available to all citizens (no means or needs tests), makes up nearly 20% of the total housing market, and is so common in Sweden that the term is rarely used.11 Housing allowances in Sweden are comprehensive, benefiting about 30% of households with children.12 Although Sweden has tended to approach housing as a social right, however, it has not made this a fully enforceable right, with the exception of emergency shelters for the homeless. Furthermore, since the 1990s, the state has introduced increasingly liberal reforms that reduced housing allowances, made them more targeted, weakened of rent-controls in favour of market prices, and generally deregulated the sector (Turner and Whitehead, 2002).

Nevertheless, Canada’s lack of both a comprehensive housing policy and the recognition that this is integral to social security, make it the least beneficial to women.

France’s housing legacy, with its relatively accessible housing allowances, makes it much more egalitarian. Yet, the problems plaguing social housing in large urban suburbs – a

11 Despite this, “[t]here is a small tendency for households with lower incomes to live in non-profit housing. This is attributed to the lower space-standards of municipal housing and the willingness of those who can afford it to rent or buy larger accommodation available in the private sector. It is not a consequence of any official selection procedure” (Ditch, Lewis and Wilcox, 2001; Fitzpatrick and Stephens, 2007, p.25-6; Priemus and Dieleman, 2002. p.193). 12 Swedish housing allowances are not available for single person households, aged 29-65, without children. (Gustafsson, 2004, p.44; Ditch, Lewis and Wilcox, 2001, p.45-6).

192

problem that has been in the making since the 1970s, stem from contradictory policy outcomes and suggest that French housing policy is limited in its effectiveness and in reducing social cleavages. Thus, despite the recent deregulating trend, Sweden’s accessible housing allowances, large public housing market and underlying policy logic that views housing as integral to social security make it the policy which best promotes women’s status.

3) Income Security Policies13 Researchers have documented the economic burden of parenthood, showing that since women continue to be primary caregivers, they are especially vulnerable to economic hardship. Moreover, without the added support of a partner, single mothers are often left to bare all the care and earnings responsibilities on their own, leaving them even more susceptible to slipping into poverty. For instance, in a recent ten-country study, “single mothers in every nation except Sweden and Belgium [were] more likely to live in poverty”

(Moller, Misra and Budig, 2006, p.26; Christopher, 2001, p.3). In many countries, mothers of young children are more likely to be unemployed, work on a part-time or seasonal basis, have lower average earnings and experience greater poverty than childless women (Moller,

Misra and Budig, 2006, p.19-26). Income security is critical to understanding the impact of welfare regimes on women’s citizenship and to alleviating inequalities since it is inextricably linked to the financial well-being of families.

Table 4b.5 compares the cases of Canadian, French and Swedish income security programs, and tables 4a.2 and 4a.7 detail the history and current systems in place in

Canadian provinces. Rooted in British poor law tradition in English Canada, and the French

13 For the purposes of the research at hand, ‘income security’ policies includes any benefit that serves to supplement the income of recipients when they have no other means to do so. In other words, this analysis is limited to what might be deemed ‘assistance of last resort,’ including, but not limited to welfare, and social or public assistance. Unemployment or employment insurance benefits are excluded from this analysis.

193

principle of subsidiarity in Québec, Canadian social assistance is a provincial responsibility that has historically been modest, stigmatizing, and a system of last resort. Many of

Canada’s early poor relief systems were based on the British model, rooted in the principle of ‘less eligibility’ which ensured that benefit levels were below that of the lowest employment wage. An extension of this principle introduced in Canada was the stigmatizing workhouse test, “…[which] had the dual aim of detecting fraudulent claims for assistance as well as implementing less eligibility as a principle” (Guest, 1999, p.38). While some improvements were made in the 1960s, these were slowly but steadily eroded by the 1980s and 1990s.14 Social assistance payments in Canada tend to be meagre, ranging from $266 a month for a single employable person in New Brunswick in 2005, to $560 in Québec (table

4b.5); these amounts are particularly low considering they are meant to also cover the cost of basic needs such as shelter and food.

Aside from a myriad of social security schemes tailored to meet very specific groups or needs through social insurance schemes, social assistance in France was historically understood to reside in the realm of the family, private charities or the discretion of local authorities (l’aide sociale). However, during the 1970s and 1980s, changes in the family and soaring unemployment rates forced the state to implement two important French income security schemes including: 1) the Allocation Parent Isolé (API), a means-tested income security scheme worth about $1100 CAD per month, targeted at lone parents, and available until the youngest child is three years old, and; 2) the much more modest and stigmatizing

14 Changes in the 1960s under the Canada Assistance Plan (CAP), a social welfare 50-50 shared-cost program between the federal and provincial governments, improved social assistance somewhat. It prohibited work as a condition to receive benefits, improved benefits and broadened its eligibility requirement so that anyone who was poor could apply for relief. However, the tide eventually reversed and, by 1996, CAP was dismantled, funding for social assistance was cut, and workfare programs began to multiply across the country (Graham and Al-Krenawi, 2001, p.411; Brodie and Isabella Bakker, 2007, p.15).

194

Revenu Minimum d’Insertion (RMI), a means-tested provision making benefits conditional upon participation in work activity projects worth just over $660 CAD per month,1 and available to anyone over 25 years of age as a system of last resort. Some studies painted a bleak picture of the program as a biased, stigmatizing system that fails to connect claimants to meaningful work and whose ineffectiveness risks turning its make-work projects into a

‘revolving door’ for the socially excluded (Murard, 2002; Enjolras, Laville, Fraisse and

Trickey, 2001, p.43-4; Noëlle Brugi, 2006; Béland and Randall, 2000, p.57; Morel, 2002, p.7-14; Alvestegui and Charbonnel, 1999, p.88; Demailly, Bouchoux and Outin, 2002;

Périvier, 2003, p.2; Cordazzo, 2005, p.420, 421).

In 2009, a new benefit, the Active Solidarity Income, Revenu de Solidarity Active

(RSA), was initiated, replacing the RMI, the API and other unemployment benefits. Thus, the RSA combines an unemployment scheme with a benefit of last resort. Claimants must be

25 years and older – although those who are under 25 with one or more children, or who have worked two out of the last three years may also qualify. It is a means-tested allowance, targeted at low-income families and individuals, which can be conditional upon participation in make-work projects. Monthly payments for a single person are approximately $610 CAD, and are calculated based on how many children are in the household, the sum amount of other monthly benefits received, and the total household revenue. Despite the fact that the

RSA has been touted by the French government as an improvement on earlier social minima, reducing the number of confusing and complex social provisions into one global benefit, and placing even greater emphasis on helping recipients secure employment, studies

1 The RMI is only one of many social minima policies in France. Between 1975 and 2002 alone, a total of seven social minima were established to try and alleviate, and integrate, the socially excluded by providing security for those falling outside the purview of insurance schemes, and to address the consequences of changes in the traditional family (Dang and Monnier, 2004, p.696; Letard, 2005; Conseil Superieur de L'Emploi Des Reveus et des Couts, 1997; Enjolras, Laville, Fraisse and Trickey, 2001, p.49).

195

suggest that this benefit does not mark any significant improvement over the earlier programs for those who are most vulnerable (Fabre and Sautory, 2009). For instance, since the RSA is conditional on the total household revenue, benefits are much less fixed, fluctuating over time depending on the receipt of a new allowance, or any other rise in income, and making it difficult for recipients to find security and stability. Moreover, studies suggest that the RSA does not solve the problem endemic to many provisions dependent upon participation in work activity projects, forcing claimants into low-skill, short-term jobs rather than meaningful work or professional advancement (Fabre and Sautory, 2009).

Swedish social assistance has its roots in the Swedish poor laws which made poor relief a local responsibility. Changes to social assistance legislation from the 1956 poor laws to the 1918 ‘social aid law’ (Socioalhjälpslagen) did little to reduce the stigma attached to the benefit or its central features (Halleröd, 2009). The more impressive changes came with the passing of the Social Service Act of 1982, marking a shift towards welfare state expansion, and the 1998 Social Service Act which among other things restricted eligibility to those of 25 years and older. These changes established a national social assistance norm, and attempted (albeit with little success) to prompt a shift in thinking away from the negative connotations of ‘the social assistance’ (socialbidrag) label, to ‘economic support’

(ekonomiskt bistånd) (Halleröd, 2009).

Swedish social assistance is available to anyone who is 25 years of age or older and:

1) poor enough to pass the means and needs test and who has exhausted all other avenues of assistance including selling assets, or; 2) has exhausted unemployment benefits. The average monthly payment for a single person is about $500 CAD per a month for the base benefit which amounts to approximately 50-60% of an average Swedish household income

196

(Gustafsson, 2004, p.36, 40). This amount does not include the additional benefit, which can be quite substantial, covering other costs such as housing (see table 4b.5). While unemployment insurance and social assistance in Sweden have long been tied to active labour market policy, the emphasis on activity measures increased during the 1980s and

1990s. Those who refuse to participate in work activities may have their benefits reduced or terminated (Gustafsson, 2004, p.37). As in Canada and France, social assistance in Sweden has long been stigmatizing. For instance, before 1945, long-term social assistance recipients in Sweden were denied the right to vote, and scholars have revealed that a substantial number of people who would otherwise qualify for social assistance chose not to apply, linking their non-use to the shame and stigma involved in receiving the benefit (Gustafsson,

2004, p.41; Korpi, 1975, 121).

Thus, in Canada, France and Sweden, income security programs tend to be nominal, means-tested, stigmatizing and often conditional upon participation in work/training programs or only available for a limited time. Canada, however, has the lowest rates – particularly in light of the fact that the state offers fewer social provisions than France or

Sweden. France has one of the most generous benefits, particularly for families with children, but payments are not fixed as they are sensitive to any changes in family income levels, and thus fail to offer much-need security and stability. Social assistance rates in

Sweden are also very high when the base and additional benefits are combined, but are more stable than the French system, making it the most egalitarian for women.

SOME WELFARE REGIME OUTCOMES As table 4b.6 details, an analysis that goes beyond social policy – to consider equality outcomes - reveals a similar pattern in terms of the effects of regime-types on

197

women’s status. For instance, the rate of women’s labour force participation serves to illustrate the degree to which regimes encourage or deter women’s paid work. While all cases studied here have relatively high rates of female labour market participation, differences are apparent. Compared to the Canadian liberal regime, Sweden and, to a lesser extent, France, have more ambitious welfare states. Both European countries opted for an interventionist approach to the family through policies which, whether intentionally or not, had the effect of supporting women’s entrance into the paid labour market, for instance, by providing daycare as well as a variety of family leaves. This is reflected in the outcome for women with young children. Thus, Sweden leads in the percentage of mothers with children under three years of age who are engaged in the paid labour market (73%), followed by

France (66%) with Canada having the lowest rate (59%) (see table 4b.6).

While a majority of women claim to be working part-time across all three cases, lingering familialism may explain the lower rates of women’s full-time employment in

France; Sweden leads (44%), followed by Canada (41%), with France having the lowest percentage of women saying they worked full-time (39%) (see table 4b.6). Even though part-time employment may help women balance unpaid care work and their attachment to the paid labour market, particularly if they are mothers with young children, part-time work typically found in lower-paid, lower-ranking positions that offer minimal benefits and limited career opportunities.

Whether through meagre, means-tested benefits, an insurance system based on employment status and contribution levels, or generous and universal benefits, welfare states were largely developed to respond to the threat of social risks such as poverty. Accordingly, consulting the poverty rates of women in Canada, France and Sweden can shed light on how

198

effective regime-types are at tackling this dilemma. As might be expected, given its higher de-commodification levels and low social stratification, the social democratic regime of

Sweden has the lowest poverty rate; it is followed by France and Canada (see table 4b.6).

Moreover, women tend to experience poverty more often than men in both France and

Canada, but not in Sweden, where researchers suggest that men are slightly more likely to be poor than women (see table 4b.6). Across all cases, however, lone mothers demonstrate higher poverty rates than other mothers. Single mothers face greater poverty (38% have incomes below the poverty line) than non-single mothers (12% have incomes below the poverty line), resulting in a sizeable 26% difference in poverty levels between two different types of mothers. In France, both single and non-single mothers fare better, and the gap between the two shrinks. Thus, 13% of French single mothers have incomes below the poverty line compared to 7% for non-single mothers, a 6% difference. Women in Sweden have the lowest levels of poverty, with the narrowest difference between single and non- single mothers. 4% of Swedish single mothers have incomes that fall below the poverty line compare to 2% for non-single mothers – a difference of 2% (see table 4b.6).

When considering the impact of benefits provided through taxes and cash transfers, the same pattern holds; as such, welfare benefits have the greatest impact on reducing poverty in Sweden, followed by France and Canada (see table 4b.6). Thus, despite certain enduring gender equality problems such as the gender-based occupational segregation that stems from the large number of women working in the ambitious Scandinavian welfare states, countries like Sweden generally tend to be more egalitarian when compared to other regime-types (Esping-Andersen, 1990, p.201-2; Hadas and Shalev, 2009; Hadas and

Semyonov, 2003; OECD, 2002, p.69, p.89-90). The differences revealed here between

199

regime-types and equality outcomes, led by Sweden and France - and with Canada lagging, suggests the depth and breadth of policies in Sweden’s social democratic and, to a lesser extent, France’s corporate-conservative regime-types have an important impact on reducing poverty rates. Yet, in all three regimes, tax and transfer-based benefits appear to have the greatest effect on reducing poverty among non-single mothers over single mothers; in light of women’s enduring social and economic disadvantage, this finding suggests that the best chance for women with children to escape poverty continues to be through reliance on a man

(Moller and Misra, 2005). This is the case in Canada where findings indicate that after tax benefits and social transfers are calculated, poverty rates are reduced by 31% for single mothers and 39% for non-single mothers – an 8% difference. In other words, social benefits have a slightly greater impact on the reduction of poverty for non-single mothers than lone mothers. Compared to Canada, French tax benefits and social transfer are more effective at reducing poverty for both single and non-single mothers alike. After tax benefits and social transfers are considered, the poverty rate of French single mothers is reduced by 64% - twice that of Canadian single mothers. French non-single mothers, moreover, see their poverty rate reduced by 69%. Thus, French benefits lead to a narrow 5% difference between non-single and lone mothers. Swedish tax benefits and social transfers have the greatest impact, reducing poverty for single mothers by 89%, and 90% for non-single mothers – just a 1% difference). Moreover, looking at table 4b.6, one can deduce that in light of the hours of unpaid and paid work women complete, they are more likely to not only be poor economically, but also constrained in terms of other resources that are needed to engage in associational life, such as time and energy.

200

The goal of this chapter was to move beyond how class and status-based individual- level factors shape civic engagement, using Esping-Andersen’s welfare regime typology to illustrate the power the state in shaping women’s citizenship in divergent ways. Although inequality and gender roles persist across countries, some states have actively sought to alleviate these by encouraging dual parenting and dual income families, facilitating women’s entrance into the paid labour market through various policy schemes which make women less vulnerable to falling into poverty. This review of family, income security and housing policies in liberal Canada, corporate-conservative France, and social democratic

Sweden demonstrates how regime-types differ in their approach to women’s inequality, and its implications for their citizenship. These diverse approaches reflect different understandings of traditional gender roles and their effects on society, as well as the way regime-types vary in their relationship between the state, market, family and civil society; these factors appear inextricably linked. The lack of support for state intervention in the family and market that is typical of liberal regimes has resulted in limited social welfare provisions for women looking to balance labour market activity with the duties they continue to disproportionately bear as primary caregivers. Canada had fewer and more meagre family, housing and income security benefits, leaving women largely fending for themselves, and resulting in the highest poverty rates for women. France was found to offer a much wider array of social provisions, with moderate-to-high benefit levels that help women reconcile work and family life. The outcome of these more generous policies is lower poverty rates for women compared to the Canadian case. Sweden was found to have the most comprehensive policies with the highest benefits levels, striving to eliminate traditional gender roles and encouraging women’s labour market activity. In addition to the

201

most ambitious welfare state, Sweden also displays the lowest levels of poverty among and between women. Yet, state intervention in the family and market alone does not necessarily spell out greater ‘gender equality’ either, as experiments with pronatalism and familialilism over time in countries like France and to a lesser extent Sweden have shown. The focus of this chapter has been on the way the state structures women’s citizenship through the rights and entitlements of the welfare state. The following chapter examines how the state shapes participation.

202

Appendix Comparative Social Politics and Outcomes

203

TABLE 4.1 A GENERAL OUTLINE OF THREE WELFARE REGIME­TYPES WELFARE REGIME-TYPES LIBERAL CORPORATE-CONSERVATIVE SOCIAL DEMOCRATIC May include US, UK, Canada May include France, Germany and Italy May include Sweden, Norway, Denmark Liberal – emphasizing the individual, and individual rights over Communitarian – emphasizing the community (society or nation) and Social/Expansive Democracy – emphasizing rights to obligations (paying taxes, military service), particularly obligations over rights. Communitarian theory preoccupied with empowerment and participation. Striving for the expansion of liberties (i.e. negative rights or freedoms from state or social effective and just functioning society. The ‘good society’ built through individual and organizational rights, and addressing interference) and is aimed at allowing individual extensive mutual obligations and where individual rights play smaller role. Aims discrimination of class/status groups. Emphasis on balance freedom of action. Class/status group rights not easily to foster a strong community based on shared identity, mutuality and between group and individual rights and obligations leads to Political Theory integrated into liberal theory, nor are social rights because they integration. Generalized exchange of rights and obligations - citizens complex self-identity linking individual interests through (Janoski, 1998) require greater obligations. Relationship between rights and should fulfill obligations without expecting immediate returns. participation in community activities (eg. work, neighbourhood, obligations is contractual or one of immediate reciprocity or or welfare-related needs) while ensuring individual civil rights restricted exchange – for each right there is generally an equal remain intact. obligation.

Pluralist – interest groups are kept at arm’s length for fear of Elitist – peak institutions and their experts highly influential in Neo-Corporatist – aggregates interest groups into peak corruption and/or bias. Influence policy by pushing interests governance, and little direct democratic participation is present. Form of federations, which are formally or informally involved in policy Political and campaigning – slow, cumbersome process with little government may vary from being elite-controlled to business-dominated process - either direct consultation with interest groups or self- Sociology consensus. Founded on liberal interpretations of rights and where peak corporations control outcomes with little input from labour, administration of government by peak federations at local, obligations focusing on individual. Welfare state benefits are grassroots or other organizations. Emphasis is on principle of regional and national levels. Participation less cumbersome (Janoski, 1998) modest and means-tested, preference for market solutions to subsidarity by elite organizations from the church to the trade unions than under pluralism because interests are aggregated and social problems, and reliance on the state seen as solution of and base benefits on social insurance principles rather than general consensus-building process embedded in institutions. Adheres last resort. revenues or means-tests. The state encourages regulative solutions, and to expansive or social democratic theory, where rights and market can be extensively organized into somewhat noncompeting obligations are in balance. firms. emphasizing obligation as opposed to rights.

The liberal polity: This polity, exemplified in Anglo-Saxon The state-nation/corporatist polity: This polity, exemplified in France The social corporatist polity: This polity, exemplified in Nordic Polity countries, tends to follow a ‘societal’ model of authority. (state-nation) or Germany (state-corporatist), tends to follow a ‘statist’ countries, tends to follow a ‘societal’ model of authority. 1 Moreover, society is organized along ‘associational’ lines. This model of authority. A high level of statism is negatively associated with Society is likely to be organized in a highly ‘corporate’ way. Tendencies low level of statism typical of liberal polities has been civic engagement. Moreover, depending on the regime, in terms of the The low levels of statism coupled with high levels of associated with fostering civic engagement. way society is organized, can either be ‘corporate’ (Germany), or a corporateness typical of social corporatist polities are credited version of ‘associational’ (France). with fostering high levels of voluntary membership.

Effects of The selective welfare state, with its divisive system, is inclined The conservative welfare state, with its étatist system, is inclined to The universal welfare state with its encompassing system institutions on to produce lower levels of trust in those that are singled out. produce blocks of trust that vary between privileged/unprivileged fosters trust. trust groups, hindering generalized trust. (Rothstein and Stolle, 2003)

1 This section draws from the work of Jepperson, as well as Schofer and Fourcade-Gourinchas. Jepperson studies polities along two fundamental dimensions of structuration: the composition and coordination of the society axis (associational vs. corporate), and collective agency axis (societal vs. statist), and tie the outcomes to levels of voluntary activity (Jepperson, 2000, p.8-16; Schofer and Fourcade-Gourinchas, 2001, p.819, 823).

204

Political Power Power moves horizontally – citizen self-organizing. Historical Power moves vertically – citizen to authority. Historical strength of Power moves in a matrix of vertical and horizontal (Janoski, 1998; weakness of leftist parties. conservative, Catholic party, tradition of absolutism/authoritarian negotiations. Historical strength of leftist parties have shaped 2 Esping-Andersen, statehood. this regime-type. 1990)

Conception of Autonomous individuals that are passively distrustful or Bonded brothers and sisters that are deferential or submissive but will Neighbours and co-workers that are cooperative and involved. Citizen actively managing the state. act when instructed. (Janoski, 1998; Esping-Andersen, 1990)

Role of Discretionary civic virtue based on personal accountability and Socially coercive civic virtue supported by fraternal responsibility. Participatory civic virtue with some social pressure based on Citizenship markets. reciprocal group responsibility. (Thomas Janoski, 1998; Esping- Andersen, 1990)

State, market, Role of the state and family remain marginal; role of market is Role of market is marginal, role of the state is subsidiary and role of the Role of the family and market remain marginal; role of the state family nexus central. family tends to be central. is central. (Esping-Andersen, 1999) Third Sector Liberal model of state-nonprofit relationship - large nonprofit Corporatist model of state-nonprofit relationship - state induced to make Social democratic model of state-nonprofit relationship. - high Tendencies sector associated with low government social welfare spending. common cause with nonprofit organizations for varying historical government social welfare spending associated and lower levels (nonprofit Reliance on private philanthropy for funding may explain high reasons. The third sector functions as one of several ‘pre-modern’ of public subsidies for and employment in the third sector. organizations level of employment in the third sector (required for fund mechanisms sustained by state to maintain control over key social elites Third sector, funded through dues/fees, very active, but less as raising). Model is consistent with a history where opposition while satisfying demands for social welfare. Third sector highly a service-provider, and more as vehicle for expressive groups. relying on from landed elites or working-class movements were subsidized, but not been consistent over time – the scale of the nonprofit Model consistent with a history where working-class elements paid/unpaid staff) weak/absent, and traditional preference for individual self- sector tends to grow with government spending. were likely able to exercise some political power in alliance (Salamon and sufficiency over government intervention in social welfare. with other social classes. Anheier, 1998; Salamon and Sokolwski, 2001)

Voluntary High voluntary membership and activity levels. Volunteering Moderate voluntary membership and activity levels, but lower than High voluntary membership levels - greater membership and Membership and dominated by service organizations. High levels of civic liberal and social democratic models. Volunteering dominated by lower activity levels. Volunteering dominated by expressive Activity engagement linked to the nature of the liberal polity (low expressive and service organizations. Lower levels of civic engagement organizations, with smaller levels of service organizations. statism); Protestant roots; and weak family ties - in the absence linked to the nature of the state-nation/corporatist polity (high statism); High levels of civic engagement linked to the nature of the (Schofer and of an extensive family network, people turn to associations. Catholic roots; strong family ties – extensive family network to rely on social corporatist polity (low statism, high corporatism); Fourcade- may lessen need to join associations. Protestant roots; weak family ties - in the absence of an Gourinchas, 2001; extensive family network, people turn to associations. Curtis, Baer and

Grabb, 2001; Reher,

1998; Salamon and Sokolwski, 2001; Jepperson, 2000)

2 In the case of France, Jacobinism worked to enfeeble the Church and promote secularization through laïcité. This ultimately hindered the viability of an effective, enduring Catholic party. Thus, the French welfare state developed under a Conservative, but secular, party.

205

Conceptualization Grounded in pluralist principles - nearly avoids overlap Mixed type of corporatist and elitist elements - the state sphere strongly Extensive overlapping spheres - especially the state, public and of the Civil between state, public and market spheres. For instance, its overlaps with the market and public spheres; the market and public market spheres. For instance, overlap between state and public Society voluntary and welfare state organizations in the public sphere, sphere, however, overlap much less. For instance, share similar spheres since third sector is often supported by state subsidies; and much less overlapping areas with the market or state characteristics of overlap as social democratic regime, except less labour and capital organizations overlap with state government (State, Market, spheres. Like social democratic regimes, liberal regimes tend to overlap between unions, employers, religion or voluntary organizations. administration of welfare services and in peak political Public, Private have larger public sphere (eg. voluntary participation) than Either the state or business has greatest influence on society – elites bargaining. Like liberal regimes, social democratic regimes spheres) corporate-conservative regimes. tend to play dominant role and mass participation is secondary. Smaller have larger public sphere (eg. voluntary participation) than (Thomas Janoski) public spheres (eg. voluntary participation) than liberal or social corporate-conservative regimes. democratic regimes.

Highly commodifying because market seen as emancipatory. Moderately de-commodifying; maintains social cleavages, for instance Highly de-commodifying; universal benefits driven by The dominant mode of solidarity tends to be the individual, and via occupational, insurance-based system. The dominant mode of egalitarian concerns. The dominant mode of solidarity tends to Welfare locus of solidarity market. Stratification – selective welfare solidarity tends to be one of either kinship, corporatism or etatism, and be universalism, and locus of solidarity is the state. Capitalism state, exemplified in means-tested programs; system is divisive locus of solidarity is the family. Stratification – the conservative welfare Stratification – since this welfare state favours universal (Esping-Andersen, because targeted provisions produce a deserving/undeserving state is etatist in that it tends to distinguish between the privileged from programs, it tends to alleviate social division since it includes 1990. Esping- citizen dichotomy; there is little impartiality since the system unprivileged; consequently, system fails to produce impartiality since all citizens; in striving to treat all citizens equally, this model Andersen, 1999; promotes stigma; rules of the system are complicated and fraud citizens receive different treatment; rules of the system are complicated works towards impartiality; rules of the system are easy. De- Rothstein, Stolle, is likely. Defamilializing/market oriented – where family and and fraud is possible. Familializing system – with considerable role for familializing/state-centered – where the state plays a 2003) state play less of a role in welfare, deferring instead to the the family in welfare. considerable role in welfare, and the state and market play a market. secondary role.

Tends to follow the liberal primary earner/secondary carer History of conservative familialism and pronatalism. While some states Tends to follow the earner-carer model – aims to help women Motherhood and regime. In this model, men and women are largely treated as (i.e. France, Belgium) have since developed policies to support women reconcile paid work and family (flexible work weeks, public 3 workers, expected to engage in the labour market. This is enter the paid labour market, they maintain signs of their old familialist child care, generous leave), and encourages men in caring role Employment reflected in women’s full-time employment and earnings levels roots, resulting in somewhat ambiguous family policies. Motherhood (eg. paternity leaves). Mothers are better able to maintain full- which tend to be higher than in other models. Care viewed as a tends to decrease the likelihood of full-time employment, but in states time employment under this earner-carer model, and there are private matter, reflected in low level of support for formal or with work reconciliation policies, such as public or subsidized few differences in earnings between single women and informal childcare. Rather, emphasis is on private provision of childcare, there tends to be fewer differences in earnings between single mothers. This model represents the best strategy for women, care, both via unpaid carework, or care services purchased in and non-single mothers. including single mothers, with the least amount of earnings or the market. The lack of family policies in primary earner model employment penalties for single and married mothers alike, and means that poverty tends to hit single mothers hardest. with relatively low levels of poverty – even among single mothers.

3 Moller, Misra and Budig actually distinguish between two Conservative models: the Conservative Primary Care/Secondary Earner model (eg. Germany) or the Conservative Choice model (eg. France) (Moller, Misra and Budig, 2007 Gauthier, 2002, p.79).

TABLE 4a.1 PROVINCIAL DEMOGRAPHIC PROFILES (2001)1 CANADIAN PROVINCES 206 Program BC AB SK MB ON QC NB NS PE NL opulation Total: 3,907,738; Total: 2,974,807; Total: 978,933; Total: Total: Total: 7,237,479; Total: 729,498; Total: 908,007; Total: Total: 512,930; Women: Women: Women: 497,180; 1,119,583; 11,410,046; Women: 3,704,635; Women: 372,790; Women: 468,920; 135,294; Women: 261,965; 1,988,640; men: 1,488,220; men: men: 481,750 Women: Women: men: 3,532,840 men: 356,705 men: 439,085 Women: men: 250,965 1,919,100 1,486,585 569,980; men: 5,832,990; 69,540; men: 549,600 men: 5,577,055 65,750

Educational No high school: No high school: No high school: No high school: No high school: No high school: 24% No high school: No high school: No high No high school: attainment 19% 22% 28% 28% 21% high school: 25% 29% 26% school: 29% 35% (25-64 yr olds) high school: 24% high school: 23% high school: 22% high school: high school: Trades: 13% high school: 24% high school: 18% high school: high school: 16% Trades: 14% Trades: 15% Trades: 16% 23% 24% College: 16% Trades: 14% Trades: 17% 20% Trades: 22% College: 19% College: 19% College: 16% Trades: 13% Trades: 11% University: 22% College: 17% College: 18% Trades: 16% College: 14% University: 24% University: 21% University: 18% College: 16% College: 19% University: 16% University: 20% College: University: 14% University: University: 18% 20% 25% University: % of visible 17% minorities in provincial 22% 11% 3% 8% 19% 7% 1% 4% 1% 1% population Roman Top 3 religious No religion: 35% Roman Catholic: Roman Catholic: Roman Roman Roman Catholic: Roman Catholic: Roman Catholic: Catholic: Roman Catholic: denominations Roman Catholic: 26% 30% Catholic: 27% Catholic: 34% 83% 54% 37% 47% 37% 17% No religion: 23% United Church: No religion: No religion: No religion: 6% Baptist: 11% United Church: United Anglican: 26% United Church: United Church: 20% 18% 16% Muslim: 2% United Church: 16% church: 20% United Church: 9% 14% No religion: 15% United Church: United Church: 10% Anglican: 13% No religion: 17% 16% 12% 7%

Employment Women: 55%; Women: 63%; Women: 58%; Women: 58%; Women: 58%; Women: 53%; men: Women: 51%; Women: 50%; Women: Women: 42%; rate (15 yrs +) men: 64% men: 76% men: 69% men: 69% men: 69% 65% men: 59% men: 60% 56%; men: men: 48% 65%

Unemploymen Women: 8%; men: Women: 5%; men: Women: 6%; men: Women: 6%; Women: 7%; Women: 8%; men: Women: 10%; Women: 10%; Women: Women: 19%; t rate (15 yrs 9% 5% 7% men: 6% men: 6% 9% men: 14% men: 11% 13%; men: men: 24% +) 14%

Lone mothers: Lone mothers: Lone mothers: Lone mothers: Lone mothers: Lone mothers: Lone mothers: Lone mothers: Lone Lone mothers: $33,829; $35,375; $28,473; $30,415; $38,110; $33,231; $27,234; $26,625; mothers: $25,149; Average lone fathers: lone fathers: lone fathers: lone fathers: lone fathers: lone fathers: $45,996; lone fathers: lone fathers: $30,116; lone fathers: family income $47,480; couples: $51,206; couples: $38,374; couples: $9,448; $53,326;couple couples: $63,975 $36,464; couples: $36,843; couples: lone fathers: $33,923; couples: $70,033 $76,903 $62,023 couples: s: $79,794 $57,254 $60,121 $38,330; $53,709 $64,223 couples: $58,366

% of families’ income from 6% 5% 7% 6% 4% 7% 10% 9% 11% 14% government transfers (couples/lone lone spouses 18-64 mothers: yrs) lone mothers: lone mothers: 5,245; lone lone mothers: lone mothers: lone mothers: 40,100; lone 401,245; lone lone mothers: lone mothers: lone mothers: fathers: lone mothers: 136,455;lone 92,945; lone 34,160; lone fathers: 9,060; fathers: 84,860; 267,570; lone fathers: 28,075; lone 36,695; lone 1,055; 18,935; lone Family types fathers: 31,960; fathers: 23,580; fathers: 7,910; couples: couples: 68,025; couples: fathers: 6,535; fathers: 7,440; couples: fathers: 4,120; couples: 917,615 couples: 694,760 couples: 223,555 253,690 2,704,880 1,683,960 couples: 180,490 couples: 218,770 32,120 couples: 131,330 1 Percentages were rounded. All data from Statistics Canada, 2001 Census of Canada.

TABLE 4a.2 OVERVIEW OF PROVINCIAL SOCIAL ASSISTANCE HISTORIES2 CANADIAN PROVINCES 207

BC AB SK MB ON QC NB NS PE NL Provincial Conservative Conservative Residual Conservative Conservative Residual Market/Family Market/Family Residual Redistributive Regime-Type Enforcement Enforcement

Differentiated Differentiated Avoided Differentiated Differentiated Avoided Uniformly treated Uniformly treated Avoided Basic assistance to between between responsibility for between deserving / responsibility for care all categories of all categories of responsibilit all in need. Less deserving/ deserving/ care of indigent. In deserving / undeserving of indigent. In recipients; recipients; y for care of stratifying (not undeserving poor; undeserving poor; providing no undeserving poor; accepted providing no viable endorsed endorsed principle indigent. In categorical) and accepted accepted viable alternative, poor; accepted responsibility alternative, principle of less of less eligibility. providing no stigmatizing (no Description responsibility for responsibility for reinforced other responsibility for ‘deserving’ reinforced other eligibility. Stigmatizing indoor viable work tests). Open of Assistance deserving poor. deserving poor. mechanisms for for ‘deserving’ poor. mechanisms for Stigmatizing relief (poor houses) alternative, to state Regime Responsibility to Responsibility to distribution of poor. Responsibility distribution of indoor relief for all recipients. reinforced intervention, from undeserving only undeserving only welfare – Responsibility to undeserving welfare – including (poor houses) for Mirrored Poor other partially due to inception- grudgingly. grudgingly. including the to undeserving only the market and all recipients. Laws. Insistence on mechanisms seasonal economy, Actively Avoided market and family. only grudgingly. family. Historical Mirrored Poor primacy of market for reflected in: little 1950 stigmatized responsibility for grudgingly. Actively role of church in Laws. Insistence and family. distribution private charity, undeserving poor. undeserving Avoided stigmatized providing relief. on primacy of of welfare – and significant Work altogether. responsibility undeserving market and including the relief to wide requirements for for undeserving poor. Work family. market and range of relief. altogether. requirements family. recipients. for relief. Adopted Adopted Timing of Adopted Adopted While some Adopted Adopted Adopted categorical categorical home categorical home Adopted While assistance categorical categorical categorical categorical categorical allowances relief relief categorical was comparatively Allowance allowances for allowances for allowances existed allowances for allowances for comparatively late – comparatively comparatively later, allowances widespread, it Adoption deserving poor deserving poor early – such as deserving poor deserving poor such as mothers’ later, and more and more restrictive comparativel was not until comparatively comparatively 1917 mother’s comparatively comparatively allowance in 1940. restrictive eligibility – y late – such 1949, after joining early, such as early, such as allowance, they early, such as early, such as eligibility – mothers’ allowance as mothers’ the Dominion of 1920 mother’s 1916 mother’s were extremely 1919 mother’s 1920 mother’s mothers’ not until 1930. allowance in Canada that a allowance. allowance. meagre. allowance. allowance. allowance not 1950. Mother’s until 1943. allowance was adopted. Market Provincial Market Market/Family Market/Family Conservative Conservative Mixed model performance Conservative Redistribu- Conservative Regime-Type performance Enforcement Enforcement tive 1960s - federal cost- Shift since 1980s With 1930s rise of Shift since 1980s Categorical Maintained its Continues to be Continues to be sharing and Quiet with increased categorical benefits Gradual shift Joining with new positive, approach non-categorical categorical, categorical, Revolution resulted categorization of and poor law from relief Confederation in Description incentive-driven remained, even approach to relief favour work favour work in generous, non- recipients, and rescinded in 1958, by discretion 1949 gave access of Assistance approach to under CAP, until –during and after requirements, requirements, categorical benefits. positive, relief became to relief as a to federal Regime encouraging 1970s when CAP. But new intrusive, intrusive, But 1980s brought incentive-driven stratifying with right. Less unemployment labour market province became focus - discretionary, discretionary, new ‘mixed’ model: approach to deserving/undeserv focus on insurance. from 1950- participation. For fully responsibility stratification stratifying and stratifying and renewed categorical labour market ing distinction. work Provincial relief 1990 instance, offering for assistance. Led elements appeared patriarchal patriarchal approach, reduced participation. For Patriarchal and require- became stratifying wage subsidies for to shift to non- 1980s, and elements – elements – benefits, emphasized instance, more moralistic elements ments, (deserving/undeser employers hiring categorical relief; stigmatization to before and after before and after labour market benefits to the before and after stratifying or ving distinction), welfare recipients. stringent work all as disincentive CAP. CAP. participation. Mix of highly CAP. stigmatizing stigmatizing and requirements to leaving market market-performance employable. recipients – marginal - before forced reliance on or family in favour and market/family before and and after CAP. market or family. of assistance. enforcement models. after CAP.

2 Adapted from Boychuck, 1998, p.24-41, and 58-96.

3 TABLE 4a.3 PROVINCIAL FAMILY-RELATED SUPPLEMENTS, SUBSIDIES AND ALLOWANCES 208

CANADIAN PROVINCES

Program BC AB SK MB ON QC NB NS PE NL Description Earned Income Alberta Family Saskatchewan Child Related ON Child Care Aide aux parents Working Income Family Assistance Benefit Employment Tax Employment Income Support Supplement for pour leurs revenues Supplement Program Credit Supplement Program Working de travail / Parental Families Wage Assistance Eligibility: Eligibility: Eligibility: Eligibility: Programme (PWA) Families with Families (children Families with Family income Eligibility: Eligibility: Eligibility: working income of > 18) with at least $3000-25,000/yr below $14,188 Families Eligibility: Families families (children families (children Program more than $3750; $6500 in family working income, get max. Assets (children > 7) (children >18) > 18) with income >19) with income Supplements reduction if working income child support, or may not exceed with 1 parent earning no more than < $3750 but > $16,500, not on for Families income is more but less than self-employment $200,000. employed, $22,000 ($15,000 for >$25,921. Max. to social assistance with Earned than $20,921. $50,000 in family studying or single parents), and families with for 3+ mnth in Income net income. training. Max assets >$45,000 for income > $20,921. past yr. No program No program benefit when renters and $90,000 income below for homeowners. Benefits: Max. Benefits: Max. Benefits: $2200- Benefits: $20,000. Benefits: Benefits: $250/yr. $50.41/ mth for $500 credit 3996 depending Monthly Benefits: Max. Additional benefit family with 1 /child/yr or $1000 on number of supplement up Benefits: Max. $3782/yr ($315/mth) of up to child, $81.16 for 2 for families with children, to max. to $30/child. annual benefit for 1 child families $20.83/mth. children, $111.66 2+ children. Credit of $12,000 (up to $1000 with $12,000/yr for 3 children and calculated at rate $185/mth for 1 ($92/mth) per income. Access to $27.50/mth for of 8% family child, $222 for 2 child under 7. childcare at $2/day. each additional working income and $259 for 3 Max. amount child. above $6,500+, children). for single reduced by 4% of Reduction when parent is $1310 incomes over family income ($109/mth) per $25,000. above $15,000. child under 7.

BC Family Bonus Saskatchewan Family Allowance NB Child Tax NS Child Benefit NF and LB Child Eligibility: Family Child Benefit Eligibility: Families Benefit Eligibility: Benefit income > $20,500 Eligibility: Family with children >18 Eligibility: Families with Eligibility: in the previous tax income > $15,921, and income > Families with income of > Families with Child Tax yr; families < with children >18 $15,000 for single income > $20,000 $15,999 get full income > $16,744 Benefits $20,500 eligible No program get max. Families No program No program parent/$21,000 for 2- get annual tax-free benefits; families No program get full benefit. for basic Family with $15,921- parent families; $250/child under between $16,000- Families earning Bonus depending 30,000 income incomes > $60,000 18. Income < 20,921 receive $16,744-21,744 on income and eligible for partial get some benefits $20,000 get some partial benefits. can get partial number of benefits. depending on number benefits depending benefit. children. of children. on number of children. Benefits: Tax-free Benefits: Tax-free, Benefits: Tax-free Benefits: When Benefits: Tax-free mthly max. of mthly max. $20.83 mthly max. from Benefits: Tax-free combined with mthly max of up $10.25 for 1st for 1st child, $160.41-316.67 for max. of National Child to $17/mth for 1st child, $26 for 2nd $37.83 for 2nd and single-parents, and $20.83/mth/child. Benefit child, $26 for 2nd, and $31.91 for 3rd $43.83 additional $52.08-208.33 for 2- Supplement, totals and $30 for each child. children. parent families. $1700/children/yr. additional child.

3 Adapted from Beauvais and Jenson, 2001, p.60-3.

209 TABLE 4a.4 PROVINCIAL LEAVE POLICIES4 CANADIAN PROVINCES

Program BC AB SK MB ON QC NB NS PE NL Description Eligibility: Eligibility: Eligibility: Eligibility: Eligibility: Eligibility: Eligibility: Eligibility: Eligibility: Eligibility: pregnant women. pregnant women, pregnant women pregnant women pregnant women pregnant women. pregnant pregnant pregnant women pregnant women employed for at employed for 20 employed in same working at least women women in same in same job for 20 in same job for at Unpaid least 12 months. wks of 52 wks job for 7 months. 13 wks before job at least 1 yr. continuous wks. least 20 wks. Maternity before leave. due date. Leave Benefits: 17 wks; starts no Benefits: 18 wks, Benefits: 18 wks, Benefits: 17 wks, Benefits: 17 Benefits: 18 wks, Benefits: 17 Benefits: 17 Benefits: 17 wks, Benefits: 17 wks, later than 11 wks at least 6 wks must starting any time starting no earlier wks, starting at starting 16 wks wks, starting up wks, starting no starting up to 11 starting no earlier before due date, be taken after during 12 wks than 17 wks before time of live birth before due date; to 11 wks earlier than 16 wks before due than 17 wks before end no later than 6 delivery; after due date; due date; – even if more Plus, flat-rate before due date. wks before due date. due date. wks after delivery; additional 3 wks additional 6 wks extension based on than 17 wks one-time payment date. additional 6 wks available w available w employer before due date. of $360 to new available with doctor’s doctor’s discretion. mothers to cover doctor certification. certification. 2 wks not covered certification. by federal benefit.

Eligibility: new Eligibility: only Eligibility: birth Eligibility: Eligibility: birth Eligibility: birth Eligibility: birth Eligibility: birth Eligibility: birth Eligibility: birth parents (birth and for adoptive parents, primary birth/adoptive and adoptive and adoptive or adoptive and adoptive and adoptive and adoptive adopting). parents. caregiver, parents in same parents, or parents. parents. parents in same parents in same parents in same adoptive parent. job for 7 months. guardian. job for 1 yr. job for 20 wks. job for 20 wks.

Benefits: 35 wks Benefits: adoptive Benefits: 12 wks Benefits: 37 wks; Benefits: 35 wks Benefits: 52 wks Benefits: 37 Benefits: 52 Benefits: 35 wks Benefits: 35 wks, Unpaid with mat. leave, 37 parents with at for birth parents, starting at end of with mat. leave, continuous wks wks for total of wks starting at for birth parents, starting at end of Parental Leave without starting least 12 months of 18 for adoptive mat. leave for 37 without; which can be 52 wks end of mat. 52 for adoptive mat. leave for end of mat. leave seniority are parents. birth mother; starting at end of taken any time combined with leave for birth parents; starting at birth mothers; unless other entitled to 8 wks taken within 52 mat. leave for within 70 wks of mat leave. mothers; taken end of mat. leave must begin within arrangement with of leave. wks after child’s birth mother; child’s arrival; starting at end within 52 wks for birth mothers; 35 wks of child’s employer; must arrival for within 35 wks returning to same of mat. leave after child’s taken within 52 arrival for start within 52 wks fathers/adoptive after child’s job for leave less for birth arrival for wks after child’s fathers/adoptive of adoptive child’s parents & arrival others than 12 wks long mothers; taken fathers/adoptive arrival for parents; returning arrival. returning to same who qualify; or equivalent jobs within 52 wks parents; fathers/adoptive to same/equivalent or equivalent job. returning to for longer leaves; after child’s returning to parents; returning job; seniority same/equivalent seniority arrival for same/equivalent to same/equivalent accumulates. job; seniority accumulates. fathers/adoptive job; seniority job; seniority accumulates parents; accumulates. accumulates. seniority accumulates. Employees are Employees are Employees Employees are Employees are entitled to 5 days entitled to 12 days working for entitled to 5 days entitled to 3 Other of upaid leave per of unpaid leave to company with unpaid leave / days unpaid Leave Related employment yr to care for sick over 50 employment yr to care leave / yr to Childcare meet care immediate family employees is care for family. 2 for family. responsibilities of (up to 12 wks for entitled to 10 days are paid if at immediate family. serious illness). days unpaid job for 2 mths. family leave / yr. 4 Adapted from Beauvais and Jenson, 2001, p.42-44, p.72 - Box 1.

210 TABLE 4a.5 PROVINCIAL CHILD MAINTENANCE ENFORCEMENT PROGRAMS5 CANADIAN PROVINCES

BC AB SK MB ON QC NB NS PE NL Family Maintenance Maintenance Maintenance Family Régime de perception Family Support Maintenance Family Support Program Maintenance Enforcement Enforcement Enforcement Responsibility des pensions Orders Service Enforcement Support Enforcement Enforcement Program Program Program Office alimentaires Program Orders Program Program Program

Any parent with Any parent with Any parent with Any parent with Persons with All separated parents. Any parent with Any parent with Any parent Any parent with an existing court ordered court order, a family custody or family support court order or with support support order or maintenance order maintenance. maintenance or support order or control of a order or registered order in agreement that has or registered written agreement. agreement. child, with agreement. agreement. receipt of been filed with Eligibility agreement. child support social court. order, or assistance domestic (either contract filed financial with court and assistance or Family day care Responsibility subsidy) Office

Provides Provides Provides Provides Provides Provides enforcement Provides Ability to issue Maintenance Receives and enforcement for enforcement, enforcement to automatic enforcement to and can also advance enforcement to garnishment of Enforcement disburses court and monitoring of including ensure enforcement encourage payment. ensure compliance income sources, to Office can ordered funds for Benefits agreements/orders. cancellation of compliance. and compliance. and also assists issue a lien on real garnish support and Family divers’ licenses, computerized parent on social property, to seize wages or any maintenance, Maintenance reporting of monitoring of assistance to bank accounts, etc. income traces delinquent Incentive debtors to Credit payments. obtain child Special source, payers, transmits encourages parents Bureau, and use of support order. enforcement is attaché liens support orders to on welfare to new series of provided for single to property, other provinces, provide family media advisories parents on social impound garnishes wages. payment on time. to locate assistance by motor individuals. Family vehicle, Family and Social Maintenance suspend Services helps Income Support drivers’ clients on social Program. licenses, and assistance obtain take other orders. remedites in accordance with Maintenance Enforcement Act.

5 Adapted from Beauvais and Jenson, 2001, p.68-9.

TABLE 4a.6 INFORMATION ON EARLY CHILDHOOD EDUCATION AND CARE BY PROVINCE (2001)6 211

CANADIAN PROVINCES

Program BC AB SK MB ON QC NB NS PE NL Description Childcare for No public No public No public No public No public $5/day childcare; No public No public No public No public preschool aged childcare; income- childcare; childcare; childcare; income- childcare; income- subsidies only if childcare; needs- childcare; income- childcare; childcare; children tested subsidies income-tested income-tested tested subsidies tested subsidies public care is tested subsidies tested subsidies income- income-tested for low-income subsidies for subsidies for for low-income for low-income unavailable for for low-income for low-income tested subsidies for families low-income low-income families for families licensed care families for families for subsidies for low-income families for families for licensed care licensed care licensed care low-income families for licensed care licensed care families for licensed care licensed care Allocation to regulated childcare $273 $110 $97 $338 $232 $980 $105 $91 $187 $101 for each child 0-12 yrs in province

$ children 0-12 for whom there is 12% 9% 4% 12% 9% 21% 10% 8% 14% 6% regulated child care space

% of regulated childcare centres 58% 44% 99% 92% 83% 86% n/a 57% 54% 36% that are non-profit

% children in regulated childcare 25% 22% 51% 48% 44% Quebec provides 23% 23% 34% 24% receiving subsidies publicly-funded services rather than subsidies Median monthly 0-17 mths: parent fees for full- 0-17 mths: $650 0-17 mths: 0-17 mths: n/a 0-17 mths: $573 0-17 mths: $783 0-17 mths: $477 0-17 mths: $380 0-17 mths: $470 0-17 mths: $n/a time centre based 18 mth-3yrs: $547 $525 18 mth-3yrs: 18 mth-3yrs: $383 18 mth-3yrs: $603 18 mth-3yrs: $455 18 mth-3yrs: $360 18 mth-3yrs: $412 $440 18 mth-3yrs: childcare 3-5 yrs: $460 18 mth-3yrs: $405 3-5 yrs: $368 3-5 yrs: $541 3-5 yrs: $440 3-5 yrs: $360 3-5 yrs: $412 18 mth-3yrs: $380 $450 3-5 yrs: $380 $380 3-5 yrs: $360 3-5 yrs: $425 3-5 yrs: $360

% National Child Benefit reinvestment funds 9% 20% 0 26% 0 Quebec does not 31% 6% 50% 17% spent on regulated participate in the childcare National Child Benefit

6 Adapted from Beauvais and Jenson, 2001, p.60-3. Friendly, Beach and Turiano, 2002.

TABLE 4a.7 PROVINCIAL SOCIAL ASSISTANCE, RANGE AND ADEQUACY OF BENEFITS, AND EMPLOYABILITY PROGRAMS (2000-01)7 212 CANADIAN PROVINCES

BC AB SK MB ON QC NB NS PE NL Work Those deemed Those deemed Those deemed Those deemed Those deemed 18-24 deemed Participation in Those deemed Those deemed Participation in require- employable employable employable employable employable work measures employable employable work measures required to engage employable required to engage required to engage required to engage required to engage voluntary but required to required to engage voluntary but ment in work measures; in work measures; in work measures; in work measures; in work measures. encouraged. No engage in work in work measures; encouraged; (all welfare single parent required to engage single parents single parents parents Voluntary for criteria for measures; Single parents single parents programs are employable when employable when employable when employable when others. Parents employability of single parents employable when employable needs test) child 7+. in work measures; child is 2+. child 6 yrs+. child is school employable when parents. employable child is 1+. when child is aged. child 2+. when child is 2+ yrs. single parents 1+.

Child tax benefits, employable when Child tax benefits, Child tax benefits, Child tax benefits, Child tax benefits, Child tax benefits, Child tax Child tax benefits, Child tax Range of health beenfit, health benefits, health benefits. health benefits, health benefits, health benefits, benefits, health health benefits, benefits, health Benefits additional benefits child is 6mths+. additional benefits additional benefits additional benefits additional benefits benefits. additional benefits benefits. (aside from such as Christmas such as utilities such as winter such as school- such as shelter and such as the back- basic and school start-up and school-related clothing and related allowance school-related to-school

assistance) allowances for subsidies for 2- school-related for 2-parent allowance for allowance for 2- families with parent families allowances for families with 2 families with parent families children. with 2 children. families with children. children. with 2 children. children.

$6457 for single Child tax benefits, $5978 for single $5558 for single $6829 for single $6415 for single $3374 for single $4817 single $5846 for single $3276 single employable is additional benefits employable is employable is employable is employable is employable is employable is employable is employable is 34% of poverty such as health 37% of poverty 29% of poverty 36% of poverty 34% of poverty 21% of poverty 30% of poverty 36% of poverty 20% of poverty Adequacy line; $14,069 for benefits and line; $12,367 for line; $11,403 for line; $13,828 for line; $13,318 for line; $12,888 for line; $12,250 line; $12,530 for line; $14,670 of single parent 1 school-related top- single parent 1 single parent 1 single parent 1 single parent 1 single parent 1 for single single parent 1 for single Welfare child is 60% of ups for families child is 61% of child is 48% of child is 59% of child is 57% of child is 64% of parent 1 child is child is 62% of parent 1 child is Benefits poverty line; with children. poverty line; poverty line; poverty line; poverty line; poverty line; 61% of poverty poverty line; 73% of poverty ($/yr) $18,412 for couple $18,210 for couple $17,585 for couple $18,330 for couple $16,919 for couple $16,206 for couple line; $18,275 $19,399 for couple line; $17,474 with 2 children is $5030 for single with 2 children is with 2 children is with 2 children is with 2 children is with 2 children is for couple with with 2 children is for couple with 52% of poverty employable is 60% of poverty 50% of poverty 52% of poverty 48% of poverty 53% of poverty 2 children is 64% of poverty 2 children is line. 27% of poverty line; line; line; line; line; 60% of poverty line; 57% of poverty line; $11,619 for line; line; single parent 1 child is 49% of Including training, poverty line; Including training, Including training, Includes Including Including training, Including Including training Including volunteering or $18,395 for couple counselling and skills and volunteering, job volunteering, education, training, in private or 3rd job training and Make- placement with 2 children is assistance with placement searches, training, training and volunteering, job volunteering, sector, counselling, work program. 52% of poverty searches. Includes programs, and substance placement placement; job-finding volunteering, education,. Childcare and line; Includes job childcare and counselling abuse programs. programs. includes childcare clubs. Childcare searches. Childcare activities other assistance training or health benefits. services. Childcare Childcare Participation subsidy. Appeals assistance Childcare exemptions available to placement, Appeal process. assistance assistance includes 23 hrs of process. available. subsidies available. participants. No volunteering, and available. Appeals available. Appeals free childcare. Appeals available. Appeals Appeals appeals process. education. process. process. Appeals process. process. process. process. Childcare assistance available. Appeals process.

7 Welfare incomes and calculations for the adequacy of benefits are annual, for 2001. Adapted from: National Council of Welfare, 2002, p.26-42; Gorlick and Brethour, 1998; Beauvais and Jenson, 2001; Jenson and Thompson, 1999. Not all benefits or employment activities are included here, for instance, benefits to disabled persons were excluded. For full details see aforementioned reports.

TABLE 4a.8 THE STATUS OF WOMEN BY PROVINCE: POVERTY, WORK AND POLITICAL REPRESENTATION8 213 CANADIAN PROVINCES Program BC AB SK MB ON QC NB NS PE NL Description Provincial 1917 1916 1916 1916 1917 1940 1919 1918 1922 1925 suffrage

% women in provincial 1984: 8% 1984: 8% 1984: 12% 1984: 5% 1984: 7% 1984: 7% 1984: 6% 1984: 6% 1984: 6% legislatures 1984: 11%; 1993: 19% 1993: 18% 1993: 19% 1993: 22% 1993: 18% 1993: 16% 1993: 10% 1993: 28% 1993: 6% 1993: 25% Employment Employ- and pay No employment Employment Employment No employment Employment equity Employment Employment ment equity No employment equity Employment or legislation. equity policy, but equity policy since equity policy; policy, no legislation; equity policy but equity policy since policy, no policy or policies and equity policy Equal pay no legislation; 1983, no equal pay equal pay legislation no legislation; 1975, no legislation; legislation; equal legislation since 1991; legislation equal legislation legislation; equal legislation 1951; 1964. Loi d’Équité equal pay legislation; equal equal pay pay legislation legislation since 1957. 1952. pay legislation proactive Pay Salariale in 1997 for legislation 1961; pay legislation legislation 1971. 1994; equal pay 1956; proactive Equity Act in 1987 public and private Pay Equity Act 1957; Pay Equity 1959; proactive legislation Pay Equity Act in for public and sectors. 1989, for public Act passed 1988 Pay Equity Act 1953; pay 1985 for public private sectors. sector and unions. for provincial 1988 for public equity sector and unions. Effectively public service sector. 1995 legislation since repealed in 1997. Act effectively 1988. repeals this.

% mothers in Infant: 63%; Infant: 60%; Infant: 65%; Infant: 66%; Infant: 67%; Infant: 67%; Infant: 67%; Infant: 71%; Infant: 77%; Infant: 64%; paid labour toddler: 69%; toddler: 71%; toddler: 74%; toddler: 75%; toddler: 74%; toddler: 75%; toddler: 74%; toddler: 72%; toddler: 79%; toddler: 66%; market child 6-15 yrs: child 6-15 yrs: Youngest child 6- Youngest child 6- Youngest child 6- Youngest child 6-15 Youngest child 6- Youngest child 6- child 6-15 yrs: child 6-15 yrs: (2001) 77% 84% 15 yrs: 82% 15 yrs: 86% 15 yrs: 82% yrs: 79% 15 yrs: 79% 15 yrs: 78% 85% 74%

Poverty rate Single mothers: % by family Single mothers: Single mothers: Single mothers: Single mothers: Single mothers: Single mothers: 42%; Single mothers: Single mothers: n/a; 2-parent Single mothers: type by 47%; 2-parent 39%; 2-parent 43%; 2-parent 55%; 2-parent 38%; 2-parent 2-parent families: 53%; 2-parent 49%; 2-parent families: n/a; 50%; 2-parent province families: 11%; families: 10%; families: 9%; families: 11%; families: 8%; 11%; childless families: 9%; families: 11%; childless families: 14%; (2001) childless childless childless couples childless couples: childless couples: couples: 10%. childless couples: childless couples: couples: n/a. childless couples: 8%. couples: 8%. hildren: 7%. n/a. 6%. 11%. 12%. couples: 14%. % low Women: 57%; earners Women: 54%; Women: 47%; Women: 52%; Women: 54%; Women: 54%; men Women: 60%; Women: 56%; men 43%. Women: 63%; working full Women: 51%; men 46%. men: 53%. men: 48%. men:46%. 46%. men 40%. men 44%. men 37%. year full time men: 49%. by sex (2000) Women: Women: 39,205; Women: 36,035; Women: 292,965; Women: 137,075; Women: 22,830; Women: 32,315; Women: 3,695; Women: 27,390; 60 hrs+/wk Women: 93,424; men: men: 14,005 men: 13,380 men: 73,850 men: 38,800 men: 7,025 men: 8,465 men: 1,095 men: 95,30 unpaid 109,300; men: 25,700 housework 30,190 (15 yrs+) Women: Women: 50,265; Women: 54,155; Women: 444,600; Women: 187,945; Women: 31,630; Women: 44,215; Women: 6,110; Women: 27,975; 60hrs+/wk Women: 148,095; men: men: 17,035 men: 19,455 men: 129,680 men: 59,055 men: 10,980 men: 13,405 men: 2,160 men: 11,175 unpaid 168,840; men: 42,030 childcare 50,950 (15 yr+)

8 Percentages are rounded. Adapted from Friendly, Beach and Turiano, 2002; Government of Canada; Bakan and Kobayashi, 2000, p.8-12 and Appendix II; Aucoin, 2005; National Council of Welfare, 2004, p.40; De Léséleuc and Kong, 2004; Canadian Centre for Justice Statistics, 2003; Statistics Canada, 2003, p.27; Iyer, 2002; Studlar and Matland,1996, p.273.

214 TABLE 4a.9 THE STATUS OF WOMEN BY PROVINCE: VIOLENCE AGAINST WOMEN, SHELTERS, AND VICTIM SERVICES9 CANADIAN PROVINCES

Program BC AB SK MB ON QC NB NS PE NL Description

Civil domestic Victims of Protection Against Victims of Domestic Domestic Correctional Services Domestic Violence Victims of violence Crime Act Family Violence Domestic Violence Violence and Violence of Quebec Act Intervention Act, Family legislation as Act (1999) Act (1995) Stalking Protection Act contains provisions 2001 – await Violence Act of 2003 (rage Prevention, (2000) related to spousal assent (1996) of remedies Protection and violence. aside from Compensation Criminal Code) Act (1999) Women: 7%; % of reported Women 9%; Women: 10%; Women: 9%; men Women: 8%; men 6%. Women: 6%; men: Women: 6%; men: Women: 8%; men Women: 6%; Women: 6%; men: Spousal men 6%. men 7%. 8%. men: 7%. 5%. 7%. 7%. men: n/a. 4%. violence (1999- 2004) Women: 11%; % of stalking Women: 11%; Women: 13%; Women: 9%; men: Women: 11%; men: 7%. Women: 10%; men: Women: 10%; Women: 12%; Women: Women: 9%; men: (1999-2004) men: 7%. men: 7%. 6%. men: 8%. 7%. men: 7%. men: 9%. 10%; men: 4%. n/a. 99 agencies Victim services 104 agencies 63 agencies 24 agencies 15 agencies reported 18 agencies reported 16 agencies 11 agencies 2 agencies 11 agencies (Police, court reported reported assisting reported assisting reported assisting 1218 assisting 391 people: reported assisting reported assisting reported reported assisting or community- assisting 1000 423 people: 64% 418 people: 71% assisting 240 people: 84% 87% Women, 13% 179 people: 69% 224 people: 69% assisting 35 156 people: 75% based) (2003) people: 75% women, 36% men. women, 29% men. people: 89% women, 16% men. Women, 31% Women, 31% people: 60% Women, 25% women, 25% women, 11 men. men. men. Women, men. men. men. 40% men.

Of those Snapshot of Of those Of those reporting, Of those reporting, Of those reporting, 2941 Of those reporting, 1 Of those Of those reporting, Of those Of those reporting, women’s reporting, 736 618 housed: 52% 225 housed: 46% reporting, 233 housed: 50% 1084 housed: 59% reporting, 119 171 housed: 51% reporting, 25 62 housed: 58% shelters housed: 56% women, 48% women, 54% housed: 44% women, 50% women, 41% housed: 54% women, 49% housed: 48% women, 42% (April 15th women, 44% children. children. women, 56% children. children. women, 46% children. women, 52% children. 2002) children. children. children. children.

99 shelters; 8 9 transition 8 transition 27 shelters and 13 shelters (10 10 shelters; 8 aboriginal About 90 15 shelters, 1 houses; 1 1 transition houses; 8 Shelters 86 transition rural family funded by SK women’s shelters; 100 shelters/transition dedicated Mi’Kmaq Family house; 4 women’s centres; (2003) houses; 16 safe violence Social Services) resource counselling houses. Aboriginal shelter. Treatment Centre; outreach 1 sexual assault home programs; prevention centres and 2 second-stage centres; 4 programs for 5 second-stage staff serving centre (province- 9 second-stage and 2 second-stage housing facilities second-stage abused women; housing facilities; 5 regions; 3 wide service); 1 housing housing projects. offering services housing about 119 6 women’s second-stage provincial programs; 37 to women and facilities; urban transitional centres; 2 abused apartment association against women’s centres children. support support workers women units. family violence (funding ceased programs for provice-wide. counselling (umbrella 2004). various programs; 2 sexual organization for communities. assault centres. shelters)

9 Percentages are rounded. Adapted from Friendly, Beach and Turiano, 2002; Bakan and Kobayashi, 2000, p.8-12 and Appendix II; Aucoin, 2005; National Council of Welfare, 2004, p.40; De Léséleuc and Kong, 2004. Canadian Centre for Justice Statistics, 2003; Government of Nova Scotia, 2003.

TABLE 4b.1 FAMILY AND CHILDREARING LEAVE POLICIES IN CANADA, FRANCE AND SWEDEN10 215

WELFARE REGIMES

LEAVES & TAKE­UP CANADA FRANCE SWEDEN MATERNITY Maternity Leave Congé de Maternité Havandeskapspenning Eligibility -Excludes self-employed until 2011, except in QC; must -All employees and self-employed workers with 10 months of -Only for women whose work jeopardizes their pregnancy; all have worked for employer for several months insurance contributions others use parental leave.

Length of Leave -15-18 wks, depending on province -16 wks -60 days

Rate -55% of earnings up to $413CAD/wk; -100% earnings up to €2,589 (just under 3800CAD)/month -80% of earnings (2006);

PATERNITY Paternity Leave Congé de Paternité Pappadagar Eligibility -1 yr continuous employment – self-employed excluded -All employees and self-employed workers -All fathers until 2011

Length of Leave -4 days; QC has 3-5 wks available for fathers -2 wks -10 days (+60 days of father’s quota for parental leave)

Rate -None -100% earnings up to €2,589 (just under 3800CAD)/month (2006) -Employed at least 240 days before birth to get 80% of earnings; otherwise, get 60SEK (just under $10CAD)/day.

PARENTAL Parental Leave Allocation Parentale d’Education (APE), now the Prestation Föräldrapenning d’acceuil du Jeune Enfant (PAJE) as of 2004 which includes the Eligibility -Must have worked at least 600 hrs over previous year. Complément de libre choix d’activité (CLCA) -All parents; for full benefit, must work 240 days for birth. No/low -All employees must have worked 2 years over the last 5 years income parents receive minimum benefit. before the birth of the child. Parents may work 16-32 hours/wk, - 1 year including maternity leave: 37 wks of parental but benefit is reduced. - 390 days of leave (13 months), including maternity leave; each Length of Leave leave in most cases for 1 parent or share between parent is allowed 195 days of leave each. Parents can transfer some parents but not exceeding 35 wks; parents can take leave -nearly 3 ½ years, including maternity leave: 3 years parental of their leave to other parent, but 60 days are reserved for each at the same time; in QC, parents can take 40-55 wks leave starting with the birth of the second child. Since 2004, parent (mommy/daddy days). Single parents get full leave. An (since 2006). CLCA recipients may get benefits starting with the first child for a additional 90 days is allowed for either parent – making total leave maximum of 6 months. at 480 days. But additional 90 days is paid at reduced flat-rate. Paid leave can be broken up and taken until child turns 8 yrs old.

-Up to 35 wks/family at 55% of average insured -Both APE and its successor the PAJE’s CLCA benefit are flat- -390 days at 80% of earnings up to 295,500 SEK (about $46,000 Rate earnings up to $413 CAD; QC – 55 wks at 70% of wage rate payments of €513 (or about $750CAD) /month (2005) to CAD); remaining 90 days at 60 SEK (just under $10CAD)/day. for 25 wks and 55% for 30 wks; or 75% of wage for 40 families below certain income; families with only 1 child receive Non-eligible parents receive 180 SEK (of just under $30 CAD)/day wks. less; supplement for larger families available. for 390 days, then 60SEK(just under $10CAD)/day for 90 days.

OTHER 5 provinces have 3-5 days leave to care for family. BC, Congé Pour Enfant Malade - all employees eligible for 3 days Tillfällig Föräldrapenning – 60 days/child under 12 at 80% LEAVE SK, ON, and NB have unpaid leave, and QC has paid unpaid leave to care for sick child under 16 yrs old. earnings, and for child 12-15 with doctor’s note; 1 day/yr can be leave for those who qualify (working 2 months.) taken to visit school of child 6-11 yrs old (Beauvais and Jenson, 2001) LEAVE About 66% of mothers received some leave; about 25% Most mothers take maternity leave, although length varies; About Almost all families use parental leave; 80% of fathers took paternity TAKE-UP were not eligible for benefit; about11% of men took 2/3rds of eligible fathers took leave in 2003. leave in 2004; 90% who became fathers after 1998 took parental parental leave. leave.

10 Drawn from Department of Trade and Industry, 2006.

TABLE 4b.2 RECENT EARLY LEARNING AND CHILDCARE POLICIES IN CANADA, FRANCE AND SWEDEN 216 WELFARE REGIMES CANADA FRANCE SWEDEN For children under 5, parents largely turn to private-run, For children under 3, the publicly administered, For children 1-6, pre-school, forskola, is unified and run centre-based and family day care. Starting at age 4 some decentralized, center-based crèche care system is by Ministry of Education, available full-time; family day TYPE OF CARE provinces offer part-day junior kindergarten for targeted available; individual, private childcare options include care, familiedaghem are found particularly in rural areas; BY AGE GROUP groups, except in Ontario, where it is widely available; 6 family day cares, assistants maternelles, or crèche there is also private day care arrangements provided by (OECD, 2007b) yrs old children may attend public, fully subsidized familiale, as well as in-home caregivers, garde à parents, cooperatives, churches, corporations and other kindergarten, most of which is part-day; legal domicile; for children 3-6, a single, universal preschool providers which covers about 13% of children; from age entitlement to free service starting at 5-6 yrs; system, école maternelle, exists, fully funded and 6-7, children attend public, pre-school class, compulsory public education begins at age 6.11 organized by the state as part of national education forskoleklass; legal entitlement to free service generally system; legal entitlement to free service generally starting at 4; compulsory public education begins at age starting at 3 yrs; compulsory public education begins at 7 (OECD, 2006c;. Gunnarsson, Korpi and Nordenstam, 1999). age 6.12 About 24% of children 0-6 have access to regulated About 64% of children 0-3 are cared for by parents, Most children under 18 months are cared for by parents; % OF CHILDREN child care, most of which are centre-based, non-profit, followed by licensed family childcare (18%), crèche 45% of 1-2 yr olds and 86% of 2-3 yr olds, and over IN EARLY services provided by community organizations. There is (8%), grand-parents (4%) or through other arrangements 91% of 3-6 yrs olds are in early learning and childcare LEARNING AND thus a large number of children in unregulated care. In (6%); for children 3-6, almost 100% enrolment in école services; 91% of 6-7 yrs olds attend forskoleklass. CHILDCARE Quebec, 34% of 0-3 yr olds and 48% of 3-4 yr olds have maternelle (OECD,2006b, p.328). access to regulated child care; about 95% of children attend kindergarten (OECD, 2006a, 2006, p.299).

Pre-school services are typically open at around 7:30am Pre-school services opened from 8:30am-4:30 pm; Pre-school open from 6:30am-6:30 pm; Monday-Friday. CARE SERVICE – 5:30 pm. There are few services outside of the regular Opened Saturdays, closed Wednesdays. Leisure centres Leisure centres before and after school or during HOURS OF work schedule or for parents working extended hours (Centres de loisirs sans hébergement - CLSH) are holidays – opened 7:30am-5:30pm (CMK foundation of OPERATION (Doherty, Friendly and Beach, 2003, p.38). Most opened Wednesdays (schools are closed), before and childcare and the Ministry of Social Affairs and Employment in kindergartens are part-day. after school, and holidays – hours vary (CMK foundation the Netherlands, 2004). About 50% of children 3-12 years of childcare and the Ministry of Social Affairs and Employment old are enrolled in leisure centres (OECD, 2006c, p.410). in the Netherlands, 2004). Child Care Expense Deduction (CCED)– federal tax For employed parents with children under 6 yrs of age, MEASURES TO deduction for child care expenses for children under 16 some or all coverage of child care at parent’s home, COMPENSATE yrs old; income must be below certain limit; in 2009, Allocation de garde d'enfant à domicile (AGED), or PARENTS FOR income had to be $10,320 or less (Canada Revenue caregiver in the home, Aide à la famille pour l’emploi CHILD CARE 14 EXPENSES13 Agency, 2009). Some provinces, such as Quebec and d’une assistante maternelle agrée (AFEAMA). Nova Scotia, provide tax credits for childcare expenses. Ceilings are based on number and age of children. There is no tax deduction for using collective care costs such as crèches, and a non-refundable tax credit for working parents is equivalent to about 25% of childcare expenditures.

11 Other provinces that run junior kindergarten programs include Nova Scotia, Quebec, Manitoba, Saskatchewan, and Alberta (Doherty, Friendly and Beach, 2003). 12 Other forms of centre-based care include: home-based care, accredited family day care, various crèche services, and haltes garderies (OECD, 2006b; OECD, 2004). 13 This section draws from Immervoll and Barber, 2005, p.59-62. 14 The Complément de libre choix du mode de garde is replacing the AGED. Centre des Liaisons Européennes et Internationales de Sécurité Sociale, Le régime françcais de la sécurité sociale, http://www.cleiss.fr/docs/regimes/regime_france3.html (last visited October 2010). “AFEAMA covers 100% of the social contributions linked to recruitment, part of the wage, and there is also a tax reduction. AGED contributes 50% for the hiring of employees. They also benefit from tax reductions up to 5,000 euros/yr” (CMK foundation of childcare and the Ministry of Social Affairs and Employment in the Netherlands, 2004).

217

CHILD CARE For 3-6 yr olds, about 0.2% of GDP; for all child care, Approximately 1% of GDP (OECD, 2006b). About 1.9% of GDP invested in early learning and child EXPENSES AS % 0-12, funding amounts to 0.4% of GDP care services in 2004. Funding per child in center-based OF GDP services are second highest, next to Denmark in all OECD (OECD, 2006c).

Parents using community services pay about 50% of For children 0-3, parents pay about 27% of costs; for Parents pay about 9% of costs to pre-school, about 2% costs (OECD, 2006a); There are limited fee subsidies for children above three, cost is free (OECD, 2006b). of the average income; parent with 1 child pays low-income parents (see provincial table); For 1998, the maximum of USD 135 monthly, 2 children USD 107

COST TO PARENTS median fee for full-time centre care for a child aged 0-17 monthly, and 3 children USD 54 monthly. Parental fees months was $531/month; $477/month for a child aged waived for low-income families. Pre-school free for all 18 months-3 yrs; $455 for a child 3-5 (Doherty, Friendly 4-5 yr olds (OECD, 2006c). and Beach, 2003, p.39); In Quebec, parents pay a flat rate of $7/day (OECD, 2006a).

TABLE 4b.3 RECENT FAMILY OR CHILD BENEFITS AND SERVICES IN CANADA, FRANCE AND SWEDEN 218 WELFARE REGIMES CANADA FRANCE SWEDEN Universal Child Care Benefit (UCCB) -starting in Allocation Familiales – available starting with birth of second Barnbidrag – benefit for all children under 16 2006, all parents of children under 6 yrs old may child. The benefit is set annually and increases in relation to the regardless of family income (Palley and Bowman, 2002, receive $100/month per child. This benefit is number of children. For 2007, for a family with two children, the p.364; Swedish Ministry of Health and Social Affairs, 2005); offered to the lower-income parent and is amount would be about €120 (about $180CAD)/month, and €153 can be extended to children over 16 and over attending taxable.15 (just over $220 CAD)/extra child. This benefit is not means-tested secondary school; SEK 950 (about $139 and is taxable.17 CAD)/month/child; Additional child allowance for National Child Benefit – an intergovernmental families with three of more children – supplement of FAMILY OR initiative aimed at low-income families. It includes Allocation forfaitaire – benefit granted to families with 3+ children SEK 254 (about $40 CAD) /month is paid for 3rd child, CHILD provincial income-tested child benefits for low- to child who has reach 20 yrs old – the regular family allowance SEK 760 BENEFITS AND income families (about $20,000 or under) with cut-off; in 2007, the benefit was about €75 (about $110 (about $120 CAD)/month for the 4th child; SEK 950 SERVICES benefits ranging from the low $20s to mid $30s CAD)/month until the child is 21. This benefit is not means-tested, (about $139 CAD)/month payable for 10 months/yr for depending on the province, income and number of but is taxable.18 children in secondary school (Swedish Ministry of Health children. The federal component is the base benefit and Social Affairs, 2005). of the Canada Child Tax Benefit (CCTB), a Complément familial – means-tested, taxable benefit for families refundable tax credit for low-income families. For with 3+ children over 3 yrs but under 21. In 2007, the benefit was Contact person or Family – Anyone in need of 2006-07, a family with income below $20,435 about €156 (just over $225 CAD)/month. additional social welfare may request help from a with 1 child received $1255 CAD/yr, and $1255/yr volunteer. Need is assessed by a social worker and is for 2 children. The federal government also Prestation d’acceuil du jeune enfant – (PAJE) – means-tested, meant as a complement to social work. This program is provides the National Child Benefit Supplement taxable benefit targeted to low-middle income families includes a not meant as a means to regulate or control of ‘clients.’ (NCBS) – is a top-up for low-income families. For package of benefits for infants, available until child is 3 yrs old; Contact persons are also not meant to be used in cases 2006-07, a family with income below $20,435 this includes (rates for 2007) a one-time birth grant worth about that are highly difficult, or where needs may be met by with 1 child received $1945, and $1720 for 2 €855 (just over $1250 CAD); a basic benefit of €171 (nearly $230 other means, such as relatives. Contact persons are children.16 CAD)/month; and benefits to help cover the costs of child care or ordinary citizens (no social work or psychology training stay home to care for the child.19 requirement) volunteers, with stable lives, no serious personal problems, and who are found to be able to forge bonds and provide support to others. While contact persons are paid, amounts are very low and vary based on their duties (Andersen, 1993, p.55-6). No guaranteed maintenance program. Allocation de Soutien Familial (ASF) -Cash benefit provided to Underhållsstöd - Cash benefit provided to custodial GUARANTEED Quebec does provide ‘advances,’ and all provinces custodial parent when non-custodial parent fails pay maintenance. parent when non-custodial parent fails pay MAINTENANCE provide some form of enforcement and/or This benefit is not means-tested, but is taxable (CRDS). Maximum maintenance. 1173SEK (just over $180 CAD)/month. monitoring of payments. (See provincial tables) benefit in 2007 about €112 (just over $160 CAD)/month.20 Parent responsible for maintenance must repay the state (Swedish Ministry of Health and Social Affairs, 2005). 15 The UCCB is sold as an attempt by the government to support parental child care choice; but there is no mechanism designed to ensure parents spend the money on child care. It is, thus, essentially a general family or child allowance. The UCCB does not affect the receipt of any other benefit – with the exception of the Canada Child Tax Benefit Supplement, which was targeted to low-income families and was rolled into the UCCB. Government of Canada, Canada’s Universal Child Care Plan, http://www.universalchildcare.ca/en/home.shtml (last visited October 2010). 16 Government of Canada, National Child Benefit Website, visit: http://www.nationalchildbenefit.ca/eng/home.shtml (last visited October 2010). 17 For France, this section draws from: Centre des Liaisons Européennes et Internationales de Sécurité Sociale, Le régime françcais de la sécurité sociale, http://www.cleiss.fr/docs/regimes/regime_france3.html (last visited October 2010). 18 For France, this section draws from Centre des Liaisons Européennes et Internationales de Sécurité Sociale, Le régime françcais de la sécurité sociale, http://www.cleiss.fr/docs/regimes/regime_france3.html (last visited October 2010). 19 As of 2004, PAJE is to incrementally replace all infant-related benefits, including the APE, AFEAMA and AGED. Caisses d’Allocation Familiale, Prestation d’accueil du jeune enfant, http://www.caf.fr/wps/portal/particuliers/catalogue/metropole/paje (last visited October 2010). Programs designed to support parents in their choice to reduce or stop the paid labour force participation include to care for their child at home include: the Complément de libre choix d'activité (CLCA) et Complément optionnel de libre choix d’activité (COLCA). 20 Centre des Liaisons Européennes et Internationales de Sécurité Sociale, Le régime françcais de la sécurité sociale, http://www.cleiss.fr/docs/regimes/regime_france3.html (last visited October 2010); The Clearinghouse on International Developments in Child, Youth and Family Policies at Columbia University, http://www.childpolicyintl.org/ (last visited October 2010).

219

TABLE 4b.4 HOUSING AND HOUSING POLICIES IN CANADA, FRANCE AND SWEDEN WELFARE REGIMES CANADA FRANCE SWEDEN RIGHT TO No. Yes – starting December 2008, enforceable right for high priority groups Yes – enforceable right to emergency shelter for HOUSING (homeless) who have experienced long delays in being allocated social roofless people. Also, enshrined in article 2(2) of (Fitzpatrick and housing. Also enshrined in preamble of Constitution, and ratified right to 1975 Constitution, and ratified right to housing, Stephens, 2007; housing, article 31 of Revised European Social Charter. article 31 of Revised European Social Charter. Government of Finland, 2006, p.17-9, and table in Appendix) No stand-alone housing allowance in Aide Personnalisée au Logement (APL) - Means-tested benefit targeting low Bostadsbidrag - Means-tested benefit. Amounts Canada – only limited ‘shelter allowances’ income urban-centre renters, with contract between the landlord and state. based on family income, household type and cost of that are part of provincial social assistance Cannot be combined with AL. Amounts vary according to income, rent, housing. 1 child: SEK 600 (just under $100 benefits, as well as the means-tested Quebec location, household type, and size of housing. For a family with 1 child in CAD)/month, 2 children: SEK 900 (about HOUSING shelter allowance for low-income families 2007, the benefit amounted to about €101 (just over $145 CAD)/month. In $140/monthCAD)/month. Nearly 30% of Swedish ALLOWANCES21 with dependents worth a maximum of 2006, the APL constituted 44% of all housing benefits funding. households with children receive housing allowance. $80/month in 2007. Allocation Logement (AL): Allocation Logement Familial (ALF) -Means-tested benefit targeting families and young couples who rent. Cannot be combined with APL. Applicants must be married less than 5 yrs, or caregiver to dependent 65+ or 21 years or under, or be in receipt of family allowances. Amounts based on income, household type, and rent. In 2006, the ALS constituted 26% of all housing benefits funding.

Allocation Logement Sociale (ALS) -Means-tested benefit targeting low- income renters. Cannot be combined with APL. Amounts vary based on income, household type, and rent. In 2006, the ALS constituted 30% of all housing benefits funding.

21 For the purposes of this research, ‘housing allowances’ do not include loans issued to homeowners or landlords for construction or renovation. For Canada, see country section - http://www.childpolicyintl.org/; for France, see Centre des Liaisons Européennes et Internationales de Sécurité Sociale, Le régime françcais de la sécurité sociale, http://www.cleiss.fr/docs/regimes/regime_france3.html (last visited October 2010); République Française, Aides Financière et Prêts: Logement, Portail de l’administration publique, http://vosdroits.service- public.fr/particuliers/N305.xhtml?&n=Logement&l=N15 (last visited October 2010); Caisse nationale des allocations familialies, 2008, p.1; for Sweden, Palley and Bowman, 2002, p.364; Swedish Ministry of Health and Social Affairs, 2005.

220

Social Housing (Fitzpatrick and Stephens, 2007) Social housing makes up 6% of housing sector and Social housing makes up 17% of housing sector and it is stable and Social housing makes up 18% of housing it is stable. growing thanks to new housing initiatives as part of Social Cohesion laws. sector and it is in decline due to demolitions in Size and trend: low-demand areas. Primarily provincial public housing authority and, 92% are Habitation à loyer modéré –HLM, some of which are public to lesser extent, co-ops and private landlords who (50%), and some of which are private under local control (42%). 95% are municipally-run housing companies. Landlords: get rent supplements.

Provinces and municipalities - typically via means- The state - based on means-testing and maximum income cut-off, but Municipally-run housing companies – based testing, and maximum income cut-off. ceiling set high enough to include not only low but also middle-income on legal framework that stipulates non- Eligibility: tenants. discrimination. No means-test, and no maximum ceiling.

High demand – 150,000 households on waiting lists Strong demand due to insufficient construction level throughout last Low demand – in some remote areas, there are for public housing. It is generally understood that decade. social housing vacancies. Demand: there has long been a housing crisis (Reitsma-Street, Schofield, Lund and Kasting, 2001, p.6). Generally yes.

Yes; eviction procedures are lengthy and expensive. Security of Tenure: No. May be asked to leave if income rises above threshold. Generally, the poor and non-whites in housing projects in outskirts of Some middle-class and working class Income and Racial major urban centres that tend to carry stigma; some newer social housing neighbourhoods, but ethnic and socio-

Mixity: Some income mixing in market rental units, but built in more attractive neighourhoods attract greater mix. economic segregation has worsened in 1990s. still much segregation; measures like co-ops tend to result in more mixing.

Transition Between 2003-04, there was an estimated 543 In 2006, there was an estimated 33 shelters available for women. In 2006, there was an estimated 160 shelters Housing22 shelters available for women. available for women.

22 For Canada, Statistics Canada, Violence Against Women… By the Numbers, available at: http://www42.statcan.ca/smr08/smr08_012_e.htm (last visited October 2010); for France and Sweden, Hagemann- White, Katenbrink and Rabe, 2006, p.52.

TABLE 4b.5 INCOME SECURITY POLICIES IN CANADA, FRANCE AND SWEDEN 221 WELFARE REGIMES CANADA FRANCE SWEDEN Social Assistance Revenu de Solidarity Active (RSA) Socialbidrag / Ekonomiskt bistånd Eligibility With exception of First Nations communities on Means-tested program of unemployment and last resort for Means-tested program of last resort, for citizens with no reserves that run social assistance programs funded by anyone 25 yrs or older, or younger if they have children or time or age limit. Expenditure item are adjusted annually the federal government, social assistance is run and have worked for 2 out of the last 3 years. Benefits can be by the central government, with local discretion on therefore varies by province (see provincial table). All made conditional upon participation in work activity projects. remaining items. are means-tested programs of last resort, typically Benefits calculated based on the value of other benefits available to those 18 years of age and over, intended to received, income and the number of children in the help meet the basic costs of living with varying degrees household. of requirement for participation in work activities. Adjustments of payment rates is at the discretion of the government (Adema, 2006, p.22).

Length Unlimited Unlimited Unlimited

There are two parts to the benefit; the basic cash benefit is Rate The basic social assistance rate provides for needs such Family benefits and housing allowances are included in the intended to cover food, clothes, leisure activities, phone, as food, clothing, shelter, and utilities. For a single household income calculation. The average rate for a single newspaper and TV. For a single person, the average rate person, the average basic rate in 2005 ranged between person in 2010 was approximately €460 or just over $610 for this in 2006 was approximately 3420 SEK/month or $280//month in New Brunswick to $560/month in CAD. Benefit rates increase with the number of dependents. just under $500CAD/month. The second part of the benefit Quebec. Recipients with children or disabilities, may Thus, for a two-parent family with two children, payments is intended to cover costs such as housing, electricity, also qualify for additional services or cash benefits such would be about €966 or $1288 CAD. travel, home insurance, union membership, fees for as allowances for transportation, childcare or drug and unemployment insurance, and in some cases, furniture, non-insured medical benefits (National Council of Welfare, dental and medical care. This part of the benefit can be 2006, p.1, 7, 10-19). quite high, quickly adding up to 4000SEK or about Earnings Benefits are reduced as income rises – there are no earnings $600CAD. Benefit rates increase with the number of exemptions and Many provinces have restrictions on the amount of exemptions. dependents. Thus the benefit is relatively high compared to deductions income a recipient can earn without losing or seeing a other welfare regimes. reduction in benefits. For instance, recipients will see their earnings deducted in full in British Columbia; this No earnings exemptions. Benefits are reduced crown-for- is also the case in Nova Scotia and New Brunswick crown against work incomes (Bergh, 2006, p.214). during the first month on welfare, and in Ontario and Saskatchewan during the first three months. Aside from this, recipients are granted earnings exemptions – for a flat-rate amount (anywhere from a maximum of $75- 200/month for a single person) or percentage of earnings (anywhere from 10-30% of wages) from employment for which their welfare benefit levels will not suffer (National Council of Welfare, 2006, p.20-7). Social Assistance For 2003, public social expenditure was 16.8% of GDP, For 2003, public social expenditure was 26.3% of GDP, 2.1% For 2003, public social expenditure was 31.3% of GDP, Spending 11.2% of which was spent on social assistance. of which was spent on social assistance. 2.2% of which was spent on social assistance. (Adema, 2006, p.13)

222

TABLE 4b.6 COMPARING WELFARE REGIMES: THE STATUS OF WOMEN IN CANADA, FRANCE AND SWEDEN WELFARE REGIMES POLICIES & CANADA FRANCE SWEDEN ARRANGEMENTS

SUFFRAGE 1918 ; aboriginal women living on reservations – 1960 1944 1919

POLITICAL REPRESENTATION Women’s share of parliamentary seats, 1998: 21% Women’s share of parliamentary seats, 1998: 11% Women’s share of parliamentary seats, 1998: 40% (Kenworthy and Malami, 1999, p.243)

GENDER POVERTY Male poverty rate: 12%; Female poverty rate: 13% Male poverty rate: 5%;Female poverty rate: 6% Male poverty rate: 3%; Female poverty rate: 2% 23 GAP (Gender poverty ratio: 1.13%) (Gender poverty ratio: 1.16%) (Gender poverty ratio: 0.6%)

SINGLE/NON- Single mothers: 38%; Single mothers: 13%; Single mothers: 4%; SINGLE MOTHER Non-single mothers: 12% Non-single mothers: 7% Non-single mothers: 2% POVERTY RATE (Christopher, 2002, p.70)

% REDUCTION OF Single mothers: 31%; Single mothers: 64%; Single mothers: 89%; POVERTY RATES Non-single mothers: 39% Non-single mothers: 69% Non-single mothers: 90% DUE TO TAX/TRANSFER SYSTEM (Christopher, 2002, p.72)

PARTNERSHIP Married: 71%; Cohabitating: 14%; Divorce ratio per 100 Married: 51%; Cohabitating: 11%; Divorce ratio per 100 Married: 45%; Cohabitating: 20%; Divorce ratio per 100 STATUS24 : 49 marriages: 39 marriages: 59

Housework: men 9.8 hrs/wk; women 16.8 hrs/wk Housework: men 14:08 hrs/wk; women 29:36 hrs/wk Housework: men 16:18 hrs/wk; women 23:41 hrs/wk 25 UNPAID WORK Other housework: men 7.7 hrs/wk; women 13.3 hrs/wk Other housework: men 2:16 hrs/wk; women 5:09 hrs/wk Other housework: men 3:30 hrs/wk; women 5:14 hrs/wk

23 Data from the Luxembourg Income Study for the mid-1990s. The ‘gender poverty ratio’ is the female poverty rate divided by the male poverty rate. A gender poverty ratio over 1 means that women’s poverty rates are higher than men’s, while a ratio below 1 means women’s poverty rates are lower than men’s. Christopher, England, Smeeding and Phillips, 2002, p.227-8. 24 Cohabitating and Married percentages for Canada from: Statistics Canada, 2001b. Data for France and Sweden from 2002, see OECD, 2006d, p.49; Data for Divorce ratios per 100 marriages from 2001; OECD, 2005a, p.33. 25 For Canada, data is from 2005, for those aged 25-54, Statistics Canada, 2006d. For France and Sweden, data from: Anxo, 2002.

Most work 40-44 hrs/wk. Women lead among those Most work 35-9 hrs/wk. Women lead among those working Most work 40-44hrs/wk. Women lead among those 223 USUAL [PAID] working fewer than full-time (40 hrs/wk) hrs. For fewer than full-time (35 hrs/wk) hrs. For instance, 2-3% of working fewer than full-time (40 hrs/wk) hrs. For WORKING HRS/WK instance, 3-4% of men work 0-34 hrs/wk, while 9-13% men work 0-34 hrs/wk, while 9-14% of women do so. instance, 3-4% of men work 0-34 hrs/wk, while 5-20% BY GENDER (OECD, 2005b) of women do so. 30% of women work 35-39 hrs/wk, However, men lead among those working more than full-time of women do so. 20% of women work 35-39 hrs/wk, while 17% of men do so. However, men lead among hrs (35 hrs+/wk). Thus, 46% of women work 35-39 hrs/wk, while 17% of men do so. However, of those who work those working full-time hrs or more (40hrs+/wk). while 51% of men do so. Of those who work 40-44 hrs/wk, 40-44 hrs/wk, 61% are men, while 38% are women. Few Accordingly, of those who work 40-44 hrs/wk, 53% are 15% are men, while 10% are women. Moreover, 8-10% of tend to work over 44hrs/wk, and while men tend to lead men, while 32% are women. Moreover, 5-8% of men for men for 45+hrs/wk, while only 3-4% of women do so. among them, the gender gap is not gaping. Thus, 3-5% 45+hrs/wk, while only 2% of women do so. of men for 45+hrs/wk, while 1-4% of women do so.

MOTHER’S EMPLOYMENT RATE BY AGE OF Under 3: 59%; Under 3: 66%; Under 3: 73%; YOUNGEST CHILD Ages 3-5: 68%; Ages 3-5: 63%; Ages 3-5: 83%; (Clearinghouse on Ages 6 to 14: 76% Ages 6 to 14: 68% Ages 6 to 14: 77% International Developments in Child, 2005)

WAGE GAP* Women earn 23% less than men in similar full-time job. Women earn 12% less than men in similar full-time job. Women earn 15% less than men in similar full-time job.

WOMEN’S PAID Full-time work: 41%; Full-time: 39%; Full-time: 44%; LABOUR FORCE Part-time work 70% Part-time work: 79% Part-time work: 73% PARTICIPATION (OECD, 2002b, p.69)

Pay equity legislation first started in the province of Pay equity between men and women in the 1946 Constitution Equal pay in some civil service positions in 1947. Pay Ontario in 1951, followed by Saskatchewan in 1952. and Articles L.140.2 and thereafter of the Labour Code equity between men and women in the 1980 Act on PAY AND Between 1956-73, all other jurisdictions followed suit. (European Foundation for the Improvement of Living and Working Equality between Men and Women/Equal Opportunities EXMPLOYMENT Conditions, 2002). EQUITY In 1953, Canada passed the Fair Employment Practices Act, as amended since (Iyer, 2002, p.41-44). LEGISLASTION OR Act covering the civil service, and passed the Female POLICY Employees Pay Equity Act in 1956. In 1977, the Canadian Human Rights Act prohibited wage discrimination between men and women, and in 1985, s.15 of the Canadian Charter of Rights and Freedoms prohibits discrimination on various grounds, including sex (Iyer, 2002, p.41-44).

Abortion first partially legalized in 1969, and broadened first decriminalized in 1975 – available on demand Abortion first partially legalized in 1938, and broadened REPRODUCTIVE in 1989 and 1990 to allow for free on request; for 1st trimester; after 10th week, 2 medical professionals must in 1946 and, finally in 1975 women’s right to abortions RIGHTS most conducted in first trimester. Contraception vouch that woman’s health or fetus is endangered to obtain on demand. In 1938, contraceptives were legalized and legalized in 1969. abortion. Contraception legalized in 1967. in 1964, contraceptive pill approved.

*‘Gender wage gap’ is measured here as the difference between male and female median full-time earnings expressed as a percentage of male median full-time earnings. For Canada and Sweden, data from 2004, for France, data from 2003. OECD, 2006d, p.72-3.

224

Chapter 5 A Hinderance or a Catalyst? Theorizing the Role of the State in Civic Engagement

The dominant civic engagement literature has shown that associational life is integral to

democracy. It helps citizens develop civic skills, builds social cohesion and civic norms such as

trust, and provides a means for citizens to voice their concerns to decision-makers. In order to

further our understanding of civic life, however, it is important to examine the factors which

shape networking and trust in the first place. Part I has elaborated on the effect of inequality and

the need to adopt an approach to citizenship which takes difference and social and economic

conditions into account to gaining a more complete understanding of civic engagement.

However, it seems impossible to fully comprehend democratic participation without also

considering another key player in the project of democracy: the state. To that end, this chapter

identifies and evaluates prevalent theories about the effect of the state on civic engagement.

Dominant county-level theories seeking to explain cross-national variation in civic involvement

focus on the influence of religion, diversity and family ties, but fail to account for the role the

state. Yet, as the last chapter highlighted, the state can shape citizenship by designing policies

which ignore, address or reinforce inequalities - all of which have an impact on participation.

The findings reveal that only a few comparative studies about civic engagement exist,

and tend to either view the state as a hindrance or a catalyst to associational life. According to

the former view, the provisions offered by large, generous welfare states can make private

provisions of welfare via networks redundant, crowding-out civic engagement. However, there are many faults with this approach, including a lack of evidence supporting the thesis, and the fact it rests on traditional gender roles, since private welfare is typically unpaid work performed by women. According the approach which views the state as a catalyst to associational life, states

225 are not a threat to networking, but work and develop in tandem with civic organizations. Political systems and generous welfare states are seen as fostering trust and networking. However these theories are also shown to be flawed for several reasons, including their inconsistent and contradictory findings, and their inability to explain the experiences of Western European countries, like France or Germany. In light of this, it is clear that no adequate theory of state currently exists which can explain comparative levels of civic engagement. EXPLORING PREVAILING COUNTRY-LEVEL CIVIC ENGAEGMENT THEORIES The few comparative studies conducted on associational life up until now have uncovered great variation across countries, suggesting that this is an area that warrants further investigation.

To that end, over the next few chapters, several analyses are conducted mostly drawing on data from the 1990 and 2000 waves of the World Values Survey, across twelve advanced Western democracies selected because they participated in questions relating to associational life for both waves. These countries include: Canada, Great Britain, the U.S., France, Belgium, Italy,

Germany, Austria, Finland, Denmark, the Netherlands, and Sweden. The analyses measure associational life by appealing to network- and norm-based indicators commonly found in the civic engagement literature, new and old. From these, a civic engagement index was built which incorporates four different indicators, including:1

- a voluntary membership index that includes fifteen organizational categories (social welfare for the elderly; religious organizations; education, arts, music and cultural activities; unions; political parties; local political action; human rights; conservation, the environment, ecology and animal rights; professional associations; youth work; sports and recreation; women’s groups; the peace movement; health organizations; and other) and tells us whether respondents belong to any group. Depending on the analyses, either a scale or dummy version of this variable is used;

1 The civic engagement index is built from World Values Survey indicators. This survey does not include measures for the vote.

226

- an unpaid work index which includes fourteen categories (social welfare for the elderly; religious organizations; education, arts, music and cultural activities; unions; political parties; local political action; human rights; conservation, the environment, ecology and animal rights; professional associations; youth work; sports and recreation; women’s groups; the peace movement; health organizations) and serves to distinguish between active members who volunteer, and simple ‘card-carriers.’2 Depending on the analyses, either a scale or dummy version of this variable is used

- a political action index which includes five categories (sign a petition; join in boycotts; attend lawful demonstrations; join unofficial strikes; occupy buildings or factories) and offers a glimpse into the various alternative forms of participation respondents may be involved in.3 Depending on the analyses, either a scale or dummy version of this variable is used.

- a social trust dummy variable based on the commonly used question, ‘do you think most people can be trust?’4

Using the civic engagement index enables us to gain a preliminary sense of civic life across

divergent Western advanced democracies. The findings detailed in figure 5.1 indicate that the

US, as well as Nordic countries, display comparatively high levels of civic involvement.

Thus, 85% of Swedish respondents say they are civically engaged, while 78% of Americans

do, 77% of the Dutch, 73% of Danes, and 67% of the Finnish. Canada is not far behind, with

62% saying they lead active associational lives, with Western European countries showing decreasingly lower scores. Accordingly, 52% of Belgian respondents said they participate,

followed by 45% of both German and Austrian respondents, while 42% of the British and

40% of Italians say they are civically engaged. France trails behind all other countries with

36% of respondents saying they are civically involved. The relationship between welfare regime-types and the civic engagement index is quite strong, positive, and statistically

2 The category ‘other’ is not included in unpaid work because it is not included in the survey for all countries in the analysis. 3 For the 15 category membership index, the alpha score is .64. For the unpaid work scale, it is .65, and for the political action index it is .75. 4 The alpha score for 4 category overall civic engagement index is .50. The World Values Survey contains questions about informal networking, however since they are only included in the 2000 wave, these are not included in the civic engagement index. The civic engagement index is built from World Values Survey indicators. This survey does not include measures for the vote.

227 significant (at the .001 level). While this analysis establishes that the correlation between regime-types and the civic engagement index are valid and not purely due to chance, it does not give us any information about whether these differences are statistically significant.

Accordingly, we must rely on a one-way analysis of variance (ANOVA) - a test designed to account for the amount of variance in a dependent variable for two or more groups. Using the civic engagement index scale, the one-way ANOVA results indicate that the difference in mean civic engagement levels between countries are statistically significant (at the .001 level).

Figure 5.1 Cross-National Civic Engagement Levels

Sweden 85%

United States 78%

Netherlands 77%

Denmark 73%

Finland 67%

Canada 62%

Belgium 52%

Germany 45%

Austria 45%

Great Britain 42%

Italy 40%

France 36%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

% of Civic Engagment

Source: World Values Survey, (1990-2000). Note: The relationship is significant (p <.001). Differences between overall welfare regime- types are statistically significant (p<.001). N = 17663.

In short, this analysis confirms what the few comparative studies that exist on

associational life have established: that significant difference persist between countries in

228 levels of civic engagement. Common explanations for such cross-national levels in civic engagement variation have been tied to the religious legacy of countries, the degree to which societies are homogenous, and the importance placed on family ties. These are briefly expanded on below.

1) Civic Engagement and the Religious Legacy Thesis Religion is understood to play an important role in the vibrancy of civic life because

of its ability to encourage norms of trust and cooperation, as well as social ties (Smidt, 1999;

Lam, 2006; Uslaner 2002). For instance, Tocqueville (Tocqueville, 2000) believed religion

has the capacity to act as a social glue, bonding together otherwise autonomous and disparate

individuals, keeping them from withdrawing from the community (Tocqueville, 2000, p.444-

5). He contended that religion can temper the harmful effects of individualism by imposing

“...on each man some obligations towards mankind, to be performed in common with the rest

of mankind, and so draws him [the individual] away, from time to time, from thinking about

himself” (Tocqueville, 2000, p.444-5). Church attendance has been said to encourage the

devout to network with those of their congregation and the wider religious community (Smidt

1999, p.177). Moreover, those active in a church can develop important civic skills, such as

letter writing, organizing meetings, and public speaking which can be used in a variety of

different types of associations, leading citizens to become increasingly active (Verba,

Scholzman and Brady, 2002, ch. 11). Religious values and beliefs are also said to have the

capacity to generate civic action, particularly if devout citizens believe they have a moral

obligation to help others (Uslaner, 2002, p.239; Smidt, 1999, p.177). Perhaps not surprisingly

then, scholars theorize that religious activity has a ‘spillover’ effect, leading to dense social

ties beyond the religious community, for instance, through charity work, welfare services, and

even encouraging citizens to become engaged in the political arena (Schmidt 1999, p.177;

229

Lam, 2006, p.178; Wilson and Janoski, 1995; Hwang, Andersen and Grabb, 2007, p.224).

However, one of the most common, long-standing theses on how religion can encourage participation, which can be traced back to Weber’s ‘Protestant Ethic,’ argues than that it is religious denomination that matters.

According to this approach, while Protestantism can work to foster civic activity,

Catholicism is inclined to curb it (Bermeo and Nord, 2000, p.xxiv; Jackman and Miller, 2005, ch.2; Curtis, Baer and Grabb, 2001, p.785-6). Distinctive values characterizing each religion are said to explain differences in levels of civic participation. As such, high levels of civic activity may be attributed to the Protestant values that promote self-sufficiency and individualism, in contrast to the dependency on church and state leaders which Catholicism supposedly breeds (Curtis, Baer and Grabb, 2001, p.785; Curtis, Grabb and Baer, 1992, p.148-9; Lam, 2006, p.178-9; Jackman and Miller, 2005, ch.2; Wuthnow, 1991, p.130, 322, note #11). The institutional structure of the church is also brought forth as a possible explanation for differences in the associational habits of Catholic and Protestant societies. The

Protestant church’s horizontal structure and sectarianism, it is proposed, creates an environment within which civic activity is allowed to flourish, as opposed to the stifling, all- encompassing confines of the Catholic church’s vertical structural hierarchy, which reduces opportunities for members to develop connections (Lam, 2006, p.178-9; Hwang, Andersen and Grabb, 2007, p.213; Curtis, Baer, and Grabb, 2001, p.785).

Some cross-national comparative studies have revealed that Protestant countries such as the U.S. and Sweden do display higher levels of civic activity than Catholic countries like

France and Italy (Curtis, Grabb and Baer, 1992, p.148-9; Curtis, Baer and Grabb, 2001, p.800;

Jackman and Miller, 2005, p.34; Bermeo and Nord, 2000, p.xxiv-xxv). It seems the ideal case

230 studies vindicating the thesis, however, are countries that have a long tradition of adherence to one dominant faith. It is less clear how this thesis fares in explaining the civic habits of countries with a mix of faiths, or multinational states such as Canada, where the historical legacy and survival of competing cultural groups within one country are tied to distinctive religious legacies.

In order to test the effect of Catholicism and Protestantism, a religious denomination dummy variable was created from the World Value Survey’s (1990, 2000) religious denomination variable, filtering out all other religious categories to isolate their distinctive effects. Figure 5.2 shows a breakdown of the percentage of respondents who identify as

Catholic or Protestant across 12 Western advanced democracies. It demonstrates that a majority of the countries that were found to be highly civically active in figure 5.1 do tend be countries where a majority of respondents identify as Protestant, including Sweden, Denmark, and Finland. There are a few cases, such as the US, Canada and the Netherlands, which challenge the validity of the ‘Protestantism-as-civic-engager’ hypothesis.

Figure 5.2 Cross-national Level of Catholics and Protestants

120%

99% 99% 100% 99% 99% 98% 100% 93%

80% 80% 70% 64% 60% 52% 53% 48% 47% % who are Catholic 40% 36% 30% % who are Protestant 20% 20% 7% 1% 1% 0% 1% 1% 2% 0%

231

Source: World Values Survey, (1990-2000). Note: The relationship is significant (p<.001). N = 24018.

Nevertheless, the results of an OLS regression, outlined in table 5.1, confirm that

Protestants are more active than Catholics. Religious denomination is found to have a significant effect on civic engagement, although the model only explains 4% (R² = .04) of the variation in civic engagement and is therefore quite weak.

Table 5.1 Predicting Civic Engagement by Religious Denomination: OLS Regression Results

Variable B SE B Beta

Religious Denomination .48*** .02 .20

R² .04

Source: World Values Survey, (1990-2000). Note: *** Significant (p<.001). N = 11210.

Thus, Protestants do seem to network more than Catholics. However, while the notion that the

institutional structure and values of churches shape behaviour and norms may, at one time,

have been especially convincing by virtue of the enormous influence of the church over

society, for the most part, this power has since waned. Across countries, the separation of

church and state, the secularization of societies, and rise and growth of the state and

bureaucracy has destabilized the church’s authority. For example, since the post-war rise of

the welfare state, there has been a systematic shifting of some of the church’s key roles to the

secular public sphere – namely in the area of poor relief and welfare. Moreover, depending on

church-state relations, religious cleavages, and the configuration of political parties - in particular, the strength and presence of Conservative or Christian Democratic parties - this process of secularization and the great ascendance of the state has had very specific implications for the shape of the welfare state and citizenship (Huber and Stephens, 2001;

Esping-Andersen, 1990; Morgan, 2006). For example, in Nordic countries such as Denmark

232 and Sweden, a relatively peaceful process of secularization took place reflecting a legacy of smooth church-state relations. With comparatively little polarizing conflict or religious cleavages to speak of which might have given Christian or Conservative parties leverage, a smooth shift of church responsibilities to the secular state transpired during the eighteenth and nineteenth centuries (Morgan, 2006). In this case, a relatively conflict-free political context facilitated the state’s increasing role and responsibility for social welfare in areas such as family and children - historically a central focus of the church.

In contrast, in states with an étatist legacy, a bitter struggle often took place between the church and state, as was the case in France and, to a lesser extent, Italy and Belgium

(Morgan, 2006). For instance, as previously outlined in chapter 4, a fervent anti-clericalism led the French state to usurp the Catholic church’s power through what scholars have referred to as statist familialism – transferring responsibilities for the care and welfare of families from the church to the secular state, often resulting in generous social provisions such as the public crèches and écoles maternelles systems in France, and the scuala materna system in Italy

(Revillard, 2006). Finally, in the United States and, to a lesser extent, Britain and English

Canada, religious diversity, Protestant sectarianism, and anti-statism resulted in a process of secularization that led to a decentralization of welfare and family matters to local communities and the third sector, rather than an ambitious or interventionist welfare state

(Morgan, 2006).5

5 The process of secularization was delayed in Quebec until the 1960s.

233

Figure 5.3 Cross-National Religious Attendance Levels 70%

60% 60%

52% 50% 44% 39% 40%

32% 30% 28% 27% 22% 20% 14% 12% 11% 10% 10%

0%

% Who Attend Religious Services on a Regular Basis

Source: World Values Survey, (1990-2000). Note: The relationship is statistically significant (p<.001). N = 36521.

The church’s weakening influence over today’s society is reflected in the dwindling

levels of church attendance, highlighted and supported by the findings in figure 5.3. It shows

that, aside from the predominantly Protestant United States and Catholic Italy, respondents in

the majority of countries do not regularly attend religious services, with Sweden - the most

civically engaged (see figure 5.1), displaying the lowest religious attendance score. As such, it

remains unclear, exactly how the church’s institutional structure, or the values it transmits,

can continue to have such a hold over the behaviour and norms of increasingly secularized

societies where individuals have little contact with the church. Thus, while the numbers

suggest that respondents who identify as Protestant generally also tend to network and trust

more than those who identify as Catholic, the process of secularization throws a serious blow

234 at the explanatory power of a thesis that rests on the influence of confessional forces – when those forces no longer hold great sway.

2) Civic Engagement and the Diversity Thesis As has already been underlined, dominant civic engagement scholars generally have

not focused on the effect of difference on participation and trust levels. In lauding American

civic life, for instance, Tocqueville’s account does not problematize the question of race

which shaped American participatory democracy. Instead, he presents a distorted portrait of

nineteenth century Jacksonian America by greatly exaggerating the level of equality in society

(Whittington, 1998). However, he was very mindful of the importance of building solidarity

and social cohesion to create an ideal context for trust to flourish and networking to take

place. Scholars continue to highlight the importance of norms of trust, cooperation, and

reciprocity – all of which contribute to social cohesion, and any factor that poses a threat to

solidarity is flagged for caution. There is a growing literature which identifies diversity as

such a factor, weakening civic engagement because of its allegedly damaging effect on

solidarity. A tradition also exists which understands diversity as a positive non-statist

attribute, linking it to stable democracy or economic growth (Anderson and Paskeviciute,

2006, p.784-5). The contact hypothesis, for instance, proposes that highly diverse societies are

desirable because they encourage greater tolerance, an important ingredient for producing

interpersonal trust and cohesion (Allport, 1954). However, for the most part, scholars focus

on how diversity can shrink civic life.

Countries such as Canada, the US and Australia, which have long been popular

immigration destinations and have important ethnic, linguistic, and native minority

populations, have very different histories with respect to race and ethnicity compared to

European countries. However, examining the role of diversity on social cohesion seems

235 particularly relevant now as international immigration levels have risen in Western and

Northern Europe as well. Figure 5.4 shows data from the UN database on international migration flows from 1990 to 2010 across 12 advanced Western democracies. This descriptive analysis provides a snapshot of where countries stand in terms of immigration levels.

Figure 5.4 indicates that there has been some increase in the level of immigration across all countries over the last twenty years. It suggests that some of the top civic engager countries, outlined in figure 5.1, are also popular immigrant destinations, making them increasingly diverse. Thus, by 2010, 21% of the population of Canada is made-up of foreign- born individuals, while it is 16% in Austria, 14% in Sweden, and just slightly less than 14% in the U.S. With this rise in immigration, those who propose that heterogeneity is harmful to social cohesion point to ethnic or racial clashes as evidence, including: conflicts in Brixton,

UK in the early 1980s, as well as in Burnley, Lancashire and Oldham, Greater Manchester in

2001; the L.A. and Toronto riots of 1992; as well as clashes in the highly diverse suburb

North-West of Paris in 2005 (Collins, Noble, Poynting and Tabar, 2000, p.84-89).

There are a number of different theories about how diversity erodes civic engagement and its causal mechanisms, but the most common contends that any rise in diversity increases ethnocentrism because it tends to enhance distinctions and mistrust between groups instead of fostering cohesion via cross-cutting bridging networks (Putnam, 2007; Forbes, 1997;

Johnston and Soroka, 2001; Lehning, 1998). Examples in support of this hypothesis usually come from the conflict, or group threat literature which argue that in light of scarce resources, dominant groups will feel threatened by minority groups, and that this will translate into

236 distrust, segregated networks, and hostility towards minority groups (Blumer, 1958; Blalock

Jr., 1967; Giles and Evans, 1986).

Figure 5.4 Cross-National Immigration as a Percentage of the Population

20

US

15 UK

popluation Sweden

the

Netherlands of

%

Italy a

as

Germany 10 France migrants Finland Denmark Canada

International Belgium 5 Austria

0 1990 1995 2000 2005 2010

Source: United Nations, 2009 (United Nations database, POP/DB/MIG/Stock/Rev.2008).

Although he rules out ‘white flight’ - a version of conflict theory positing that white

Americans pulled out of civic life as a result of civil rights gains culminating in legal

desegregation in the 1960s (Putnam, 2000, p.280), Putnam does find that diverse communities

have weaker social ties, as well as lower political efficacy and norms of trust (Putnam, 2007).

Diversity in America leads to what Putnam refers to as a ‘hunkering down’ effect – a general

withdrawal from civic life. Countless studies exist beyond Putnam’s which uncover a similar

237 correlation between diversity and lower levels of civic engagement (Taylor, 1998; Bobo,

1999; Bobo and Tuan, 2006; Alesina and La Ferrera, 2002; Costa and Kahn, 2003).

However, some scholars have challenged the theoretical claims and empirical findings of this scholarship, arguing that by presenting homogeneity as a prerequisite for civic engagement, this approach fails to adequately problematize the concept of diversity and inequality in society in the first place. For instance, the conflict literature does not account for the way race, space and socio-economic factors interact, and how these shape conflict, power dynamics and cleavages in society (Hero, 2007; Stolle and Hooge, 2005; Anderson and

Paskeviciute, 2006). Many scholars, moreover, have found that when studying the effects of heterogeneity on associational life, inequality and deprivation were often associated with, and sometimes even had a greater effect on, participation and trust than race and ethnicity

(Alesina and La Ferrera, 2002; Laurence and Heath, 2008; Letki, 2008; Kesler and

Bloemraad, 2010). In fact, when race and ethnicity are appropriately included in an analysis,

Hero finds that their impact on American social capital outcomes almost always surpasses that of social capital indicators themselves, suggesting that the effects said to stem from social capital, such as lower crime and high school drop-out rates, economic growth, effective and responsive institutions, may actually be attributable to racial factors (Hero, 2007, p.154). Such results reinforce the findings outlined in Part I which indicates that participation is closely tied to citizenship status. According to this line of reasoning, privileged positions in the social matrix usually translate to greater access to resources and networking opportunities which, in turn, lead to a more vibrant associational life. Thus, the work of scholars who challenge the conflict literature bring to light the racial implications of this approach: on the one hand, civic engagement theory claims that norms of trust and dense, bridging social ties can lead to better

238 social, political and economic outcomes; on the other hand, the conflict literature suggests that diversity leads to undesirable social, political and economic outcomes. From this perspective, civic engagement is not actually about better outcomes for everyone, but rather better outcomes for some - namely, wealthier, white, homogenous communities, a point that is rarely elaborated on in the conflict or dominant civic engagement literatures.

According to conflict theory logic, it follows that countries with historically high levels of diversity, such as Canada, and the US, should have particularly low levels of trust in those who are ‘different,’ viewing them as competition for already scarce resources (such as jobs, schooling, benefits); alternatively, countries which have historically been homogenous should have particularly high levels of trust because they have fewer ‘differences’ in society which may create tension or threaten the majority. Figure 5.5 examines this proposition by showing the results of a bivariate analysis between advanced Western democracies, and a

“trust in ‘different’ others” dummy variable, created by calculating the sum of values for two dummy variables: “trust in those of a different nationality,” and “trust in those of a different religion.” Figure 5.5 demonstrates that societies which have historically been homogenous, but where immigrants make-up an important percentage of the population today, have relatively high levels of trust in ‘different’ others, such as Sweden (94%) and France (89%).

However, societies with legacies of diversity, such as the US (87%) and Canada (86%), also have high levels of trust in ‘different’ others – a finding which contradicts conflict theory. The relationship between the various countries and levels of trust are strong, positive, and significant (at the .001 level). Countries with the lowest levels of trust in ‘different’ others also do not form a clear pattern, with some having been historically homogenous such as Italy

239

(54%), and others having long experiences with diversity such as the Netherlands (54%) (in terms of deep confessional and ideological cleavages).

Figure 5.5 Cross-National Levels of Trust in ‘Different’ Others 100% 94% 89% 87% 90% 86% 85% 83%

80%

70%

60% 54% 54% 53%

50%

40%

30%

20%

10%

0% Sweden France United States Canada Great Britain Finland Italy Netherlands Germany

Source: World Values Survey, (2005). Note: The relation is statistically significant (p<.001). N = 10220.

In order to delve deeper into the link between diversity and civic engagement, an OLS

regression is conducted. The process of looking for suitable data with which to test the effect

of diversity on civic engagement brings to light the difficulties encountered when measuring

race or ethnicity. Either surveys include limited civic engagement indicators, or do not

contain suitable information on race or ethnicity. Nevertheless, the ISSP Social Networks II

(2001) survey includes data on voluntary membership levels in seven types of organizations

(political party, neighbourhood, charitable, sports, religious, union, and the category ‘other’)

which is turned into a scaled membership index.6 The dataset also contains categories for

ethnicity, however these are recoded into a simple ‘race’ dummy (white/non-white) because

6 The alpha score for the membership index is .63. N=7804.

240 of the insufficient number of cases for the multitude of ‘ethnicity’ categories across the six advanced Western democracies included in the analysis (Canada, the US, the UK, Finland,

Denmark and Germany). The results of the OLS regression with race as the predictor and the voluntary membership index as outcome variable indicate that race has no effect on participation whatsoever, and the model is not statistically significant (F = 1.807, p = .179,

N= 4671). These findings cast doubt on one of the central arguments used in the conflict literature; that diversity disrupts social cohesion because it increases mistrust, and fails to foster tolerance of those who are different than us.

Thus, while there seems to be a sizeable number of studies documenting the negative relationship between diversity and civic engagement, it can hardly be considered closed for debate. Firstly, a disproportionate amount of evidence for a negative relationship between diversity and civic engagement stems from studies on American inter-ethnic or -racial group relations, with just a handful of analyses testing this in other countries or on a comparative basis. Although some findings from the Netherlands, the UK and Australia support the conflict hypothesis (Lancee and Dronkers, 2008; Leigh, 2006; Pennant, 2005; Fieldhouse and

Cutts, 2010), results from other studies on diversity in Canada, the UK, as well as across

Europe, contradict these, challenging the explanatory power of this approach (Hooghe,

Reeskens, Stolle, and Trappers, 2009; Aizlewood and Pendakur, 2005; Letki, 2008; Johnston and Soroka, 2001). Soroka and Johnston, as well as Banting, for instance, do not reach the same fatalistic conclusions about the effect of diversity on Canadian social cohesion, despite this country’s multinational, multi-ethnic legacy (Johnston and Soroka, 2001; Banting, 2005).

Secondly, the relationship between the state, citizenship, and civic engagement is rarely explored within most of this diversity literature. Perhaps one of the best examples of

241 this is the literature postulating a link between diversity and a decline in support for the welfare state because of its corrosive effect on social cohesion and solidarity. According to this perspective, an influx of immigrants into a community often leads the majority group to hold the welfare state at fault for enticing newcomers to settle there, exploiting the system

(McCarty, 1993; Alesina, Baqir and Easterly, 1997). Even aside from the many studies which propose evidence to refute such claims (Kesler and Bloemraad, 2010; Banting, Johnston,

Kymlicka and Soroka, 2006; Hero and Preuhs, 2006; Taylor-Gooby, 2005; van Oorschot and

Uunk, 2007), considering the effects of diversity and inequality on decision-making and social security spending without accounting for the impact that these have long had on racial and ethnic groups seems misguided. Studies which find a negative relationship between diversity, social cohesion, and support for welfare generally do not account for the role of the state in mediating conflict, inequality and race relations, and do not examine how questions of race or ethnicity may have shaped the development and restructuring of the welfare state in the first place (Quadagno, 1994; Mink, 1990; Amott, 1990; Neubeck and Cazenave, 2001).

For the most part, the scholarship tends to treat race, ethnicity and inequality as though they operated in isolation from the broader power dynamics of a polity, absent the role of the state or institutions. As such, it fails to consider how inequality is connected to citizenship, and structured by the state. Race and ethnicity have long shaped, and remain an important part of how the social and the political is structured (Hero, 2007, p.169). Moreover, the way the state receives immigrants and addresses inequality, reflected for instance in multicultural and social policies, has an important bearing on citizenship status and, hence, participation (Kesler and

Bloemraad, 2010).

242

3) Civic Engagement and the Family Ties Thesis Aside from religious tradition and diversity, another explanation for national-level

variations in civic engagement is linked to family ties. Dominant civic engagement scholars,

including Coleman and Fukuyama, have suggested strong, successful civil societies are

dependent upon the traditional nuclear family unit and that disruptions to its fundamental

structure has detrimental effects on civic engagement (Herd and Meyer, 2002; Field, 2005;

Edwards, Franklin and Holland, 2003; Coleman, 1998; Fukuyama, 1999). However, the dominant civic engagement literature also contends that when family ties are too strong, they can suffocate associational life. For instance, Banfield, whose work influenced Putnam’s social capital studies, argued that ‘amoral familism,’ a bonding strategy with kin he found was thriving in Montegrano, Italy, worked to undermine norms of generalized trust and networking opportunities outside of one’s immediate family (Banfield, 1958). Echoing this finding, Putnam argued that communities which rely heavily on close, tight-knit family ties, as in Southern Italy, tend to have weak civil societies, with low levels of generalized trust, fewer bridging networks, and less effective, more corrupt institutions (Putnam, 1993;

Fukuyama, 1995).

Scholars continue to highlight the potentially negative effects of familial bonds on associational life. According to this view, strong family ties stimulate strong bonding networks and personalized trust amongst members of the same family. Such network homophily limits the building of a web of diverse social connections and trust with those in the wider community. Moreover, it is argued that states with strong family ties tend to have

weak welfare states, reinforcing the family as the main source of support, while states with

weak family ties tend to have more developed welfare states (Alesina and Giuliano, 2007;

Reher, 1998; Pichler and Wallace, 2007; Alesina and Giuliano, 2009; Curtis, Baer and Grabb,

243

2001; Saraceno, 1994; Kääriäinen and Lehtonen, 2006; Norris and Inglehart, 2006).

These differences are often rooted in cultural explanations, linking back to a country’s religious tradition. For example, historically Protestant countries are said to encourage individualism and, eventually, de-familialization through the welfare state; Catholic countries are said to encourage close-knit bonds through familialism and corporatism, reinforced through the principal of subsidiarity – a mode of social organization which valorizes the family as primary social unit, dictating that the state and community should not interfere in the affairs of the family (Esping-Andersen, 1990; Reher, 1998). As such, scholars suggest that it is possible to distinguish between countries which have a legacy of strong or weak family ties. Amongst Western advanced democracies, Southern European or Mediterranean countries are said to have underdeveloped welfare states and strong family traditions, while North

American and Northern European societies typically have more ambitious welfare states and display weaker family networks (Kääriäinen and Lehtonen, 2006; Curtis, Baer and Grabb,

2001; Reher, 1998).

Yet, while it may best explain civil society and the lower levels of social security found in Mediterranean countries, the family ties theory cannot adequately explain the case of

Western Europe. As will be shown in the subsequent chapter, countries such as France,

Germany, Belgium and, to a lesser extent, Italy, generally spend a larger percentage of their

GDP on welfare compared to liberal or social democratic countries, and have developed important welfare states – despite their strong family legacies.

244

Figure 5.6 Cross-National Level of Family Ties

Italy 96%

Denmark 90%

Austria 89%

Germany 83%

Great Britain 83%

United States 83%

Canada 83%

France 79%

Finland 76%

0% 20% 40% 60% 80% 100% 120%

Source: ISSP Social Networks II, (2001). Note: The relationship is statistically significant (p < .001). N = 1539.

Studying how familial bonds shape associational life proves tricky because of the

limited indicators available in surveys which measure such networks. Nevertheless, as figure

5.6 demonstrates, it is possible to get a good sense of how countries rank in terms of family

ties. Figure 5.6 shows the results of a bivariate analysis using the ISSP Social Networks II

(2001) survey between nine Western advanced democracies, and a ‘frequent networking with

kin’ index dummy variable built from eight different ‘family networks’ categories (how often

do you visit with: father; mother; brother or sister; aunts and uncles; cousins; parents-in-law;

brothers- and sisters-in-law; nieces and nephews).7 According to the family ties hypothesis,

countries with stronger familial bonds should not only have underdeveloped welfare states,

but also lower levels of civic engagement since they mostly network with a limited number of

kin. As figure 5.6 indicates, however, this is not always the case. In accordance with the

familial ties theory, Italy, which has a Catholic history and low levels of civic engagement,

7 The alpha score for the kin index is .50.

245 has strong family ties, with 96% saying they frequently visit with family. However, while

France tends to display some of the lowest levels of civic engagement (for instance, see figure

5.1), French respondents also scored low on family networks, with 79% saying they frequently visit with family; this finding betrays both France’s strong familialist legacy, as well as the family ties hypothesis. 83% of respondents in the U.S. - which has notoriously high levels of civic engagement - said they frequently visit with family; thus Americans say they network 4% more than the supposedly familialist French. Moreover, countries such as

Denmark (90%), Austria (89%) and Germany (83%), which Esping-Andersen has highlighted have ambitious welfare regimes, all also have stronger family ties than the more meagre liberal regimes of Canada, Great Britain or the US (all at 83%), contradicting the family ties hypothesis. In short, this analysis does not provide support for the family ties theory since, contrary to this approach, many countries with ambitious welfare states and active societies are also found to have strong family ties.

Finally, in order to examine the family ties theory further, an OLS regression is conducted on the kin networks index scale and a voluntary membership index dummy which includes seven organizational categories (political party, neighbourhood, charitable, sports, religious, union, and the category ‘other’) across the Western advanced democracies found in figure 5.1 that are included in the survey (Canada, the US, the UK, Finland, Austria, Italy,

France, Denmark, Germany).8 However, the analysis shows that family ties have no effect on

organizational involvement, and the model itself is not statistically significant (F = .735, p =

.391. N = 1258). In light of this, the family ties theory is not empirically vindicated here, and its theoretical claims do not fare better.

8 The alpha score for the kin index is .63. N= 7804.

246

In seeking to gain a better understanding of the factors which shape civic engagement at the country-level, this section reviewed prevailing theories offering explanations for cross- national variations in associational life. However, these theories miss an important piece of the puzzle in disregarding the role of power structures both at the individual-level, through factors such as gender, race and class, as well as at the country-level through the role of the state in shaping citizenship and participation. Indeed, the state’s impact on inequality and civic involvement is evident but remains unexplored in each of the three theories. The religious legacy theory, for instance, does not consider how the Church’s power defused and shifted to the state in the area of welfare and the voluntary sector through the rise and development of the welfare state, and the implications this might have had on levels of disparity as well as participation. Scholars who posit that diversity hinders social cohesion and threatens civic bonds do not pay attention to how the state might shape associational life in their approach to social stratification, disparity and tension between racial and ethnic groups through immigration policy as well as the rights and entitlements which stem from the welfare state. Finally, supporters of the family ties hypothesis fail to account for the gender implications of relying on homogenous, tight-knit family ties for welfare and support, and overlook the way the state affects the family-state-market nexus by establishing social and family policies which either ignore, encourage, or reinforce traditional gender roles. In light of this, and despite the paucity of research in this area, the following section provides a review of the most prominent theories on the role of the state in associational life.

247

THE ROLE OF THE STATE IN CIVIC ENGAGEMENT 1) The Dominant Literature and the State as a Hindrance One of the most long-standing views on the role of the state in civic engagement

suggests that strong states weaken social ties. For our purposes here, “strong” or “large” states

can be defined as: “… those with well-developed bureaucracies and long traditions of

governmental intervention” (Koven and Michel, 1990, p.1079-80). According to this

approach, when the state is small and weak, the needs of the collectivity are met by the

community – either through informal social networks, or with the help of more formal

voluntary groups. Under such conditions, civic engagement is said to flourish. However, as

the state grows, it encroaches upon activities otherwise led by organizations, suffocating

associational life. Espousing such a view, Norton contends, “as government has stepped in to

provide welfare benefits, there has been that much less need for private provision of welfare,

whether this be through the family, networks of friends, or formally organized charitable

bodies. The incentive to invest in these social relationships declines, and so social capital is

lost” (Norton, 1998, p.1). Thus, according to this ‘crowding-out’ logic, Anglo-Saxon liberal

welfare regimes, which tend to limit the state in favour of private market solutions, should

enjoy a more vibrant civic life compared to Western European corporate-conservative and

Northern European social democratic welfare regimes, which have historically accepted much

more interventionist states.

A cross-national overview of the mobilization efforts by early women reformers can

be used to illustrate how strong, large states are said to crowd-out civic life. During the nineteenth and early twentieth century, when they lacked full citizenship rights, early women reformers in Europe and North America actively campaigned for greater women’s rights, as

248 well as for advances in maternal and child health and welfare (Pedersen, 1995; Kovens and

Michel, 1993; Allen, 2005; Bock and Thane, 1994; Sainsbury, 2001). Scholars often refer to this movement as ‘’ because activists used their roles as mothers and wives in the private sphere as a springboard for action in civic groups and politics, shaping the development of the welfare state, and challenging the boundaries of the public and private divide (Orloff, 2006; Koven and Michel, 1990). States either modeled welfare programs based on the voluntary work that maternalists had undertaken in the community, or recruited them for their expertise to directly manage or deliver state-run welfare programs (Koven and

Michel, 1990).

However, in comparing these movements on a cross-national basis, scholars uncovered a trend: women in countries with comparatively small states, such as the United

States and Britain, were more successful not only in building strong, diverse networks, but also in seeing more women get promoted to important administrative positions in which they could play a greater role in shaping welfare policy (Skocpol and Ritter, 1991; Skocpol, 1995;

Koven and Michel, 1990). In contrast, early women reformers in strong states such as France and Germany were much weaker; they were excluded from prestigious positions in the welfare state administration because intense professionalization of the public service, popular among strong states, formally disqualified them. Thus, as Koven and Michel state:

[A] comparative examination of maternalists’ achievements… leads to an awkward and disconcerting conclusion: the power of women’s social action movements was inversely related to the range and generosity of state welfare benefits for women and children. “Strong states”… allowed women less political space in which to develop social welfare programs than did “weak states,” where women’s voluntary associations flourished (Koven and Michel, 1990, p.1079-80).

249

Therefore, the experience of early women reformers across various countries suggests a direct and negative relationship between more ambitious, interventionist states and the vibrancy of civic engagement.

Dominant civic engagement scholars do not dwell on the role of the state in civic engagement because they endorse the crowding-out hypothesis, which conceptualizes the state as a hinderance to civic life. Tocqueville, whose work continues to greatly inspire the dominant civic engagement literature, subscribed to this view after having encountered an actively engaged nineteenth century American citizenry burgeoning alongside a relatively weak state structure (Tocqueville, 2000, p.514-15; Whittington, 1998, p.22). It suggested to him that a negative relationship existed between the relative strength of the state and civic engagement, particularly compared to his native France, where associational life was compromised by an overpowering strong state. According to this reasoning, the ebb and flow of associational life is largely dictated by the degree to which social needs are met. When the state is small and weak, citizens must work together to meet their needs, generating a great deal of activity; but as the state develops, citizens lose their motivation to form social networks because their needs are increasingly, and more effectively, met by the state

(Whittington, 1998, p.23-4). Thus, this perspective conceives of the state as a threat to the civil society. For instance, Tocqueville states:

I know that many of my contemporaries…claim that as citizens become weaker and more helpless, the government must become proportionately more skillful and active, so that society should do what is no longer possible for individuals... but I think they are mistaken…. It is easy to see the time coming in which men will be less and less able to produce, by each alone, the commonest bare necessities of life. The tasks of government must therefore perpetually increase, and its efforts to cope with them must spread its net ever wider. The more government takes the place of associations, the more will individuals lose the idea of forming associations and need the government to come to their help. That is a vicious circle of cause and effect…. The morals and intelligence of a democratic people would be in as much danger as its commerce and

250

industry if ever a government wholly usurped the place of private associations (Tocqueville, 2000, p.514-15).

The preference for a small state is consistent with a central idea inherent to the dominant civic engagement literature, and rooted in the liberal and civic citizenship traditions: that citizens are independent, active participants in both self-governance

(independent, autonomous, rational agents) and in governance of the community. As

Tocqueville conceives it, there is a direct link between democracy, self-governance, and governance for the common good (community); the freedom and equality that comes with democracy transforms what were once political subjects into political citizens, able to actively participate in governance of the self and the community (Cruishank, 1999, p.96).

Paradoxically, this very state of democratic freedom and equality can lead to isolation and powerlessness rather than solidarity and efficacy, as the old hierarchical bonds from the previous social order give way to a new caliber of individualism. As such, what allows free, isolated, democratic individuals to gain power is to work together in great numbers. In other words, the self-governance that comes from democracy requires people combine to get things done and build trust and reciprocity – a level of participation which inevitably serves to strengthen democracy (Cruishank, 1999, p.96-7; Tocqueville, 2000).

Since Tocqueville’s time, dominant civic engagement scholars have remained uninterested in exploring the role of the state in participation, or considering its potentially positive influence on associational life, despite the development of the modern welfare state which was accompanied by an unprecedented and exponential growth in the role and size of the state. Thus, Putnam does not find any correlation between ‘big’ government and social capital in America, nor does he pursue the subject further (Putnam, 2000, p.281; Putnam,

2002, p.414). Instead, he approaches the state very narrowly, as a measure of social capital

251 outcomes. According to Putnam, the ultimate social capital outcome is enhanced democracy, reflected in effective and responsive state institutions. Coleman, on the other hand, links the state to the family and civil society, lamenting what he sees as the linkage between decreases in social capital, and the downfall of the traditional family. According to this view, and in line with the crowding-out theory, the duties and functions that were originally the responsibility of the family have increasingly been replaced by the modern welfare state and, as such, the welfare state is seen as contributing to the weakening of traditional families, rather than encouraging parental obligations to the family and community (Coleman, 1994, p.606-610).

Additionally, Fukuyama endorses the notion that the size of the state is negatively correlated with social capital. For instance, echoing Tocqueville, he points to France as an example of the harmful effect that the state can have on civic life, arguing,

…[S]tates can have a serious negative impact on social capital when they start to undertake activities that are better left to the private sector or to civil society. The ability to cooperate is based on habit and practice; if the state gets into the business of organizing everything, people will become dependent on it and lose their spontaneous ability to work with one another. France had a rich civil society at the end of the Middle Ages, but horizontal [social, generalized] trust between individuals weakened as a result of a centralizing state that set Frenchmen at each other through a system of petty privileges and status distinctions (Fukuyama, 2000, p.15).

Finally, the ‘crowding-out’ approach has been adopted by those who support market-based neo-liberal logic to justify ‘rolling back’ the welfare state and, at an international-level, placing restraints on World Bank structural adjustment programs. It is argued that, rather than expect a hand-out in the form of aid, the poor and disadvantaged must be encouraged to pull themselves up by the bootstraps via self-help efforts (Skocpol, 1996; Harriss, 2002, p.7; Fine,

2006; Schuurman, 2003; Fine, 1999).

252

2) Shortcomings of the Dominant Approach Scholars have highlighted various drawbacks of the crowding-out hypothesis – most

notably, that there is little empirical support for this approach either at the individual or

aggregate level, and any evidence brought forth usually comes from the American experience

(van der Meer, Scheepers and Grotenhuis, 2009; van Oorschot and Arts, 2005). Furthermore, the crowding-out hypothesis continues to be critiqued for inadequately defining what is meant by, or how to operationalize, ‘large’ or ‘strong’ states. As Novak states,

What do we mean when we talk about a state as weak or strong? Are we talking about the organization of officialdom or the reach of public authority? Are we talking structure or function, process or substance, inputs or outputs, scale or scope, law- making or law enforcement, people or policies? At precisely the point where an interesting substantive discussion of the multifaceted nature of governance should begin, the weak vs. strong state debate usually devolves into a rather vapid (and endless) discussion of the proper unit of measurement (Novak, 2008, p.762; Krasner, 1984).

The most common measure for state size or strength is the level of GDP spent on social

welfare provisions. However, merely calculating levels of social expenditures does not tell us

much about the underlying purpose of policies or their outcomes. The trade-off for such a neat

and easy measure for state size and strength is that it provides limited information. Raw social

spending numbers give us a sense of how much a state spends in the area of social welfare,

but it does not include information about how this money is spent, which can help distinguish

states based on how their social systems vary (Esping-Andersen, 1990, p.19). This lack of

information can lead to false conclusions because a state may spends less on social welfare –

meaning it would be interpreted as a weak state; however, a low-spending state may

nevertheless intervene in the market and family – which would be interpreted as a strong

state. As Esping-Andersen contends,

By scoring welfare states on spending, we assume that all spending counts equally. But some welfare states, the Austrian one, for example, spend a large share on benefits

253

to privileged civil servants. This is normally not what we would consider a commitment to social citizenship and solidarity. Others spend disproportionately on means-tested social assistance. Few contemporary analysts would agree that a reformed poor-relief tradition qualifies as a welfare-state commitment (Esping- Andersen, 1990, p.19-20).

Moreover, proponents of the crowding-out hypothesis only conceptualize the welfare state in terms of the degree to which the provision of welfare usurps private responsibility and self-sufficiency. As such, it presumes that the state has the will and capacity to eliminate the

‘need’ for associational life in the first place. In proposing that by effectively meeting social needs, all-encompassing, generous welfare states inevitably eliminate the incentive to be civically active, this approach reduces people’s motivation for networking to satisfying their social welfare needs and, as such, implies that a weak-state, active civil society would be dominated by service-based groups whose central purpose is to deliver social services including those related to health care, housing and education (Salamon, 1996). However, this downplays another important ‘social’ component inherent in networking: that individuals often want or need to interact with others for no other reason than leisure. Citizens often seek out membership in ‘expressive’ organization, which focus on the expression of values, interests and believes of members, including professional associations, cultural and recreation groups, spiritual or religious networks (Salamon, 1996). The prospect that people’s incentives for networking go beyond seeking to address their social welfare needs poses a problem for the ‘size or strength of the state’ approach because it leaves open the possibility that a vibrant civil society could exist alongside a strong, large welfare state.

A comparative survey of the dominant fields in the nonprofit sector, outlined in table

5.2, hints at a link between the size of the welfare state, and the types of organizations individuals belong to.

254

Table 5.2 The Cross-National Voluntary Sector, by Field

Regime-type Service-based Organization (%) Expressive-based Organization (%) Liberal Australia 53 44 Canada 75 22 UK 57 38 US 64 31 Liberal mean 62 34

Corporate-Conservative Belgium 60 40 France 32 65 Germany 23 58 Italy 45 52 Corporate-Conservative mean 40 54

Social Democratic Finland 22 78 Netherlands 46 52 Norway 19 77 Sweden 14 81 Social Democratic mean 25 72 Source: Adapted from Salamon, Sokolowski, and List, 2003, p.58. Data for Canada from Hall, Barr, Easwaramoorthy, Sokolowki and Salamon, 2005, p.12. The categories of ‘other,’ ‘international’ and foundations’ labelled under the field ‘other’ were omitted here to focus solely on the service-expressive aspect of the nonprofit sector. As such, not all country percentages add up to 100%.

It confirms that, on the one hand, liberal regimes have a strong service-based voluntary sector

- presumably to make up for a weak welfare state (Salamon, Sokolowski, and List, 2003, p.34). On the other hand, social democratic and corporate-conservative regimes, which have more generous, interventionist welfare states, have relatively smaller service-based nonprofit sectors, with the exception of Belgium. However, contrary to the crowding-out hypothesis, the sizeable score for expressive voluntary groups in social democratic and corporate- conservative regimes demonstrates that people want to be civically active irrespective of state intervention levels. Moreover, alongside strong service-oriented nonprofit sectors, liberal regimes generally also have an important base of expressive associations. As table 5.2 shows,

31% of voluntary organizations in the United States qualify as expressive, while 44% do so in

Australia.

Since it views the state as a threat to civil society, the crowding-out hypothesis posits that for associational life to flourish, there should be minimal state intervention. For instance,

255 the US is said to owe the strength and vigour of its civil society to a deliberately weak state, historically and systematically restrained from interfering in the lives of citizens. However, scholars have shown that this is largely American civic folklore; in reality, this picture of civil society as superior to, and separate from the state, was not self-evident until after the late nineteenth century. According to Salamon (1996), great social, economic and political shifts which transpired between the end of the American Civil War and the start of the Great

Depression led to a change in ; one which elevated the private for-profit and voluntary nonprofit sectors as superior to and at odds with the state (Salamon, 1996). In other words:

...[T]he concept of a distinct nonprofit sector was a late nineteenth century invention in America, and one with powerful – and ultimately highly effective – political overtones. By creating a potent ideology of ‘voluntarism’ and investing it with mythic status as the best and true expression of the American character, conservative forces effectively held the state at bay for two generations despite deteriorating social and economic conditions, and despite the clear inability of private philanthropy to live up to the expectations claimed for it (Salamon, 1996, p.6). Skocpol has also worked to dispel the ‘weak state-strong civil society’ myth by detailing the way in which the voluntary sector and the state in America have developed and functioned concurrently over time, rather than as contradictory opposites. (Skocpol, 1996). More recently, and particularly with the welfare state restructuring that has taken place cross- nationally since the 1970s, other scholars have also highlighted the blurring of lines between the state and voluntary sector, such as with quangos (or quasi-NGOs) in countries like

England (Hall, 1999, p.440-3; Worms, 2002), and arrangements between the state and voluntary sector that characterize the ‘social economy’ in countries like France and Italy

(Warin, 2002; Seibel, 1990; Ullman, 1998; Barbetta, 1997).

In fact, the relationship between the state and voluntary sector can be scrutinized by examining sources of nonprofit sector revenue, such as those highlighted in table 5.3.

256

Intuitively, the crowding-out hypothesis would predict that strong and large welfare states such as Sweden should receive the majority of their nonprofit funding from public sources due to their highly interventionist states; alternatively, smaller, weaker states such as the US should rely more heavily on private sources of funding. Yet, the findings in table 5.3 challenge this logic, with private sources of funding making up the dominant source of funding for nonprofit organizations in non-interventionist and interventionist countries alike.

Table 5.3 Sources of Nonprofit Sector Revenue, (2003-2005)

Private Sources Public Sources Regime-type (fees and donations) (%) (%)

Liberal Australia 69 31 Canada 48 51 UK 64 36 US 74 26 Liberal mean 64 36

Corporate-Conservative Belgium 34 66 France 67 33 Germany 57 43 Italy 70 30 Corporate-Conservative mean 57 43

Social Democratic Finland 75 25 Netherlands 54 46 Norway 80 20 Sweden 86 15 Social Democratic mean 74 26 Source: Adapted from Salamon, Sokolowski, and List, 2003, p.32. Data for Canada from Hall, Barr, Easwaramoorthy, Sokolowki and Salamon, 2005, p.15. Note that some percentages may not add up to 100%.

In fact, the countries with the highest level of private funding are not solely liberal, but include social democratic regimes - countries that proponents of the crowding-out approach might characterize as strong and large. For instance, 86% of Swedish nonrprofit revenue stems from fees and donations, while 80% of nonprofits in Norway and 75% in Finland get funding from private sources (see table 5.3). In addition, the investment of public money into the voluntary sector on a cross-national basis also poses a problem for the crowding-out approach. Contrary to the hypothesis, social democratic regimes, despite their strong, large

257 welfare states, have some of the lowest levels of government funding for nonprofit organizations, with only 15% in Sweden, and 20% in Norway; this, compared to the 26% in the US, or even the 51% in Canada.

Another weak-spot with the crowding-out approach, and related to previous points, is its tendency to falsely presume the presence of an alternative source of welfare; if people are not getting their welfare from the state, they are getting it elsewhere – presumably from social networks. What is not elaborated upon, however, is that this alternative source of welfare has historically been the product of women’s unpaid work in the home. Private welfare is not new; it predates and persists alongside the welfare state, and has seen a resurgence more recently as a result of the privatization and downloading of health and welfare that has characterized cross-national welfare state restructuring (Christopherson, 2001). As such, the crowding-out approach rests on traditional gender roles, and completely discounts the impact that the development of the welfare state has had on women’s citizenship, providing them with some relief from their roles as unacknowledged care and welfare providers. This can be seen in Coleman, who argues that, in the wake of the breakdown of the traditional family, the welfare state has stepped in where the original duty and function of the family once lay – in caring for children, the sick, the infirm and the elderly (Coleman, 1994). What is overlooked by Coleman is the deeply gendered implications of his account of welfare provision (Risman and Ferree, 1995; Edwards, Franklin and Holland, 2003, London, p.5). In addition, the crowding-out hypothesis also neglects the empowering effect that welfare states have had on women, for instance through family policies, facilitating their entrance into the paid labour market. This point is never considered by those who support the crowding-out approach,

258 however, because of their failure to consider the outcomes of the welfare state beyond the degree to which they substitute for private obligations.

While concerned with the vibrancy of associational life, the crowding-out hypothesis takes a very narrow understanding of what constitutes an ‘association;’ it presumes that social networks only serve private goals and meet private needs, restricted to the individual members in the immediate network or community. As such, it fails to recognize the existence of other types of collective action, such as adversarial groups like the women’s movement that, in seeking social change on a large-scale, have often called for greater state intervention in the market and family. The crowding-out hypothesis, therefore, tends to exclude certain types of networks, such as those that call for major social change, and greater state intervention.

Furthermore, it is only concerned with one part of the democratic process – civic activity. Outcomes in terms of reducing or reinforcing inequality, or gaining access to and influence over decision-makers – are ancillary. For instance, as mentioned above, delving deeper into the story of early women reformers, many of whom were mobilizing for increased state action to address women’s poverty and maternal health and hygiene issues, reveals an uneasy paradox: women in strong, interventionist states which tended to exclude them and where networks were limited generally benefited from better outcomes by way of more comprehensive and generous social welfare provisions. Conversely, women in weaker states, such as the US and Britain, while being more successful in forging diverse, powerful networks, and gaining access to influential positions in the state itself, were unable to make such gains (Koven and Michel, 1990). Thus, irrespective of the impact of the size or strength of the state on maternalist mobilization, the policy outcome is undeniable:

259

In the development of maternal and child welfare legislation… Germany and France repeatedly led the way in the range and amount of benefits they provided mothers and children…British women received fewer benefits, and what they had often took the form of limiting [e.g. maternity leave without pay] rather than redistributive protective legislation…The United States provided neither federal maternity benefits nor medical care for mothers and children… (Koven and Michel, 1990, p.1105).

Thus, what matters to the crowding-out hypothesis is that women’s groups in America were more successful in building stronger networks than women in large states like France or

Germany because of the small state in liberal regimes – whether or not those networks were effective. According to this approach, it is of no consequence that early women reformers in liberal welfare regimes were unable to achieve the gender equality outcomes they sought which would have required intervention into the ‘private’ realm of the market or family; something that is not tolerated by weak states.

In sum, the crowding-out theory is blind to power structures in society, and the capacity that welfare states have to alleviate poverty and build solidarity in the community. It takes for granted that all individuals in the community start off with equal citizenship status and have the same ability and capacity to network; as such, it suggests that the comprehensiveness, redistributive capacity, and reach of the welfare state can easily be replicated through private social networks between members of the community.

Consequently, this approach tends to value networks for their ability to encourage social cohesion and self-sufficiency, and does not recognize groups or networks which challenge or oppose existing power structures. Moreover, in its preoccupation with participation, the crowding-out hypothesis does not theorize the role of trust in civic engagement, or consider how the state can shape such norms and build solidarity. This leads to an alternative view of the role of the state in relation to civic engagement – one that is expanded upon in the following section.

260

An Alternative View: The State as Catalyst An alternative to the crowding-out hypothesis has developed more recently, one that

takes issue with presenting the state as a threat to the civil society. Proponents of this approach challenge the conventional wisdom of the dominant civic engagement literature which envisions the causal link as stemming from the citizen to the state. In emphasizing a top-down approach, scholars illustrate how the state and civil society have developed in tandem, and investigate how the degree of state intervention (e.g. universal versus residual welfare states) might affect levels of civic engagement (Skocpol, 1996; Rothstein, 1998;

Halpern, 2005, p.187). Two approaches seem to have come out of this line of inquiry: one approach focusing mostly on the structural side of civic engagement – namely, the link between the state and social networks; the second focusing on the cultural side of civic engagement – namely, the link between the state and norms and values such as trust and cooperation.

1) The State as a Catalyst for Networking Skocpol’s work on the evolution of civic engagement and democracy stands as a

prime example of the former approach, centered on a networks-based account of the way the

state influenced participation. In the case of America, for instance, she shows that factors

including a political opportunity structure that rewarded associations organized at the

national-level, as well as organizational limitations resulting in federally-organized groups

which could coordinate activities across localities or regions, led to the development of a

relationship between the state and voluntary sector that was much more symbiotic than described by the crowding-out hypothesis (Skocpol and Fiorina, 1999; Skocpol, Ganz and

Munson, 2000). While not excluding it from the roster of possible causal variables, Skocpol’s

structural approach refrains from viewing norms of trust and social cohesion as the panacea

261 for associational life. Instead, in considering the road to democracy, she takes note of the historical struggle for power between groups along the way, and pays attention to the crucial role that organizations played in the struggle for greater equality. As such, she does not overlook the role that conflict has played in the formation of democracy in the first place

(Skocpol and Fiorina, 1999, p.14). According to this account, then, “…voluntary associations matter as a source of popular leverage, not just as facilitators of individual participation and generalized social trust” (Skocpol and Fiorina, 1999, p.15).

Thus, contrary to the division between state and civil society that is central to the crowding-out hypothesis, supporters of this alternative approach acknowledge that civic engagement – even in so-called ‘weak’ states like the United States – were shaped by political events, and functioned in tandem with an active government and democratic political system

(Skocpol, 1996; Schlesinger, 1944; Skocpol, Ganz and Munson, 2000). Tracing the evolution of voluntary associations alongside the growth of the American bureaucratic and welfare state, Skocpol contends:

Twentieth-century voluntary federations were often built from the top down, deliberately structured to imitate and influence the three tiers of U.S. government, and encouraged by parts of the federal government itself. Thus the American Legion was launched from the top by World War I military officers and later nurtured by the Veterans Administration. And the American Farm Bureau Federation was encouraged by the U.S. Department of Agriculture. The PTA itself, now romanticized as a purely local voluntary group, did not originally bubble up from below. It was founded in 1897 as the National Congress of Mothers (renamed the PTA in 1924)… From its inception, the Congress of Mothers/PTA was actively involved in public policymaking and the construction of a distinctively American version of the welfare state (Skocpol, 1996, p.20-25).

In a similar vein, other scholars have documented the historical partnership between state and voluntary sector that has evolved in different countries.

262

Figure 5.7 Annual Registration of New Associations and Laws Extending the French Welfare State Annual Registration (Thousands) (Thousands) Annual Registration

Laws Extending the Welfare State (Year of Implementation)

Source: Worms, 2002, p.144. Édith Archambault, Le secteur sans but lucrative en France et dans le monde, Economica, Paris, 1996, p.2. Note: Some corresponding laws include: law granting hunting and fishing associations the privilege of delivering permits (1965); law on adult professional education opening new opportunities for associations (1971); new subsidies for services for the elderly rendered by associations (1977); new legal possibilities for the development of local independent non-commercial radio stations run by associations (1981); transfer of optional school sports activities to associations (1986); legal recognition and financial help to intermediary associations for the unemployed (1989); inner-city programs heavily leaning on local associations (1992).

As figure 5.7 shows, studies of civic engagement in France have highlighted the prominent

role that the state has had on the development of voluntary associations (Archambault, 2001;

Worms, 2002). Despite a long history of étatism, “[t]he millennial fight of the central state

against local power in any form,” the development of an ambitious post-war welfare state did

not crush the voluntary sector (Archambault, 2001). In fact, the French nonprofit sector came

to play a particularly important role starting in the 1970s, to address the rise of ‘social

exclusion’ stemming from economic crisis. Indeed, over 40,000 new groups were created in

the thirty-year period between 1969 and 1999 (Archambault, 1996, p.2). Thus, in spite of its

statist legacy, French associational life actually blossomed in tandem with a strong welfare

state, leading Worms to conclude: “[i]n my view, the growth of the welfare state stands out as

the single most obvious explanation of the exceptional growth of the associative sector since

263

1960” (Worms, 2002, p.144). Echoing Worms, Hall points to the ever-blurring line between the public service and associations with the rise of quasi-non-governmental organizations, and to the great impact that public policy has had on associational life (Hall, 2002, p.40-1;

Worms, 2002, p.144-5).

While Skocpol, Worms and Hall use voluntary membership data to trace activity levels over time, they focus on single-country cases based on national-level findings. It becomes evident from the cross-national-level evidence, however, that vast differences exist between different nations in terms of organizational involvement (Curtis, Grabb and Baer,

1992; Curtis, Baer and Grabb, 2001; Schofer and Fourcade-Gourinchas, 2001). Cross-national examinations of the way the state shapes civic engagement tend to rely on comparative survey data, such as the World Values Survey (WVS) or the International Social Survey Programme

(ISSP). Moreover, some have begun to employ hierarchical modelling techniques, capable of controlling for both individual-level and national-level effects, as well as the interaction between variables operating at different levels of analysis (Curtis, Baer and Grabb, 2001;

Schofer and Fourcade-Gourinchas, 2001). Results from the few cross-national studies that have been undertaken find support for a positive relationship between individual-level factors such as education and religion and networking. However, even when controlling for these, the influence of country-level variables on participation remains strong (Schofer and Fourcade-

Gourinchas, 2001; Curtis, Baer and Grabb, 2001). Overall, results from comparative studies reveal that Nordic countries such as Sweden lead in membership levels in voluntary associations, followed, and sometimes rivalled by, Anglo-Saxon countries such as the United

States and Canada. In fact, liberal regimes lead in active participation in groups, followed by

Nordic countries (Curtis, Baer and Grabb, 2001, p.797). Western European countries such as

264

France, Italy and Germany, tend to have the lowest scores in voluntary membership and activity.

Variation in civic engagement across nations is typically attributed to the long- standing historical legacy and the cultural context of groups of countries that tend to correspond to Esping-Andersen’s typology. For instance, in testing for the effects of various factors, including religious tradition, industrialization or level of economic development, and democratic stability, Curtis, Baer and Grabb (2001), also consider the impact effect of divergent ‘political systems’ on civic engagement. Building on Janoski (1998) who uses welfare regime-typology, they predict that social democratic and liberal regimes will lead in the forming of voluntary organizations but for very different reasons.9 In the case of liberal

countries, their explanation echoes the crowding-out hypothesis, arguing that countries such

as the US and Canada have high levels of activity to compensate for a weak welfare state

(p.786-7). However, contrary to the crowding-out logic, they contend that social democratic regimes like Sweden will also demonstrate high levels of civic engagement, largely due to the enduring prominent role that unions and political organizations play in these societies (Curtis,

Baer and Grabb, 2001, p.786). In fact, Curtis, Baer and Grabb (2001) argue that, in social democracies, the state actually strengthens the voluntary sector by providing groups with a stable, steady source of public funding which, it is argued, frees up the time of volunteers from fund-raising duties, “…which means that they [voluntary associations] are often more independent from and critical of state activities than in the case of liberal regimes” (Curtis,

Baer and Grabb, 2001, p.786). Moreover, still drawing on Janoski (1998), Curtis, Baer and

9 Curtis, Baer and Grabb (2001) depart from Janoski (1998) and Esping-Andersen (1990, 1999) by adding a fourth welfare regime category – to include countries from the former eastern bloc (Bulgaria, East Germany, Estonia, Hungary, Latvia, Lithuania, Romania and Russia). While it is beyond the scope of this dissertation to include these countries in the analysis, Curtis, Baer and Grabb find that countries from this regime-type have the lowest score, below that of corporate-conservative countries.

265

Grabb (2001) argue that, because of their elitist legacies and segregated institutional arrangements, corporate-conservative regime-types such as France and Italy suffer from low civic engagement scores (p.786-7).

Schofer and Fourcade-Gourinchas (2001) also conduct a cross-national voluntary membership study that links variation in civic engagement levels to long-standing cultural and institution country-level differences. However, while they anticipate results similar to Curtis,

Baer and Grabb (2001), Schofer and Fourcade-Gourinchas draw on the work of Jepperson

(Jepperson, 2002; Jepperson, 2000) rather than Janoski (1998), and therefore base their analysis and interpretation of findings on a different theory. Rather than welfare regime-types acting as indicators for different political systems, Schofer and Fourcade-Gourinchas test for the effects of concepts developed by Jepperson, including high or low levels of ‘statism’ on participation.10 Statism can be understood as a continuum describing the relationship between the state and civil society over time, ranging from low-statist, or ‘non-statist,’ countries where political authority and legitimacy resides with the people, to statist countries which have a

history of authoritarian rule and a clear boundary between the state and civil society (Schofer

and Fourcade-Gourinchas, 2001, p.811). Non-statist countries generally consist of those that

approximate Esping-Andersen’s Anglo-Saxon liberal regimes – which have a tradition of a self-governing, autonomous civil society. However, non-statism, as defined by Jepperson, should not be confused with a ‘weak state’ since the Nordic social democratic regimes are also considered non-statist. According to this view, even though social democratic countries have a long history of state intervention, the state does not work to usurp power from the

10 While Schofer and Fourcade-Gourinchas (2001) examine two of Jepperson’s concepts (2000, 2002), statism and corporateness, the purposes of this research, I only include ‘statism’ here. Statism is best at describing variation in levels of civic engagement, whereas corporateness is best at explaining the modalities – which types of organizations are favoured by which types of countries (old organizations versus new social movements). It is elaborated on in the subsequent section.

266 people and there is little or no boundary between the state and civil society (Schofer and

Fourcade-Gourinchas, 2001, p.812). Thus, non-statist countries such as Sweden, Norway,

Denmark, the US, Canada and Great Britain, should see high levels of voluntary activity, compared to statist countries. In statist countries such as France, Germany and Italy, absolutist pasts have entrenched the notion that decision-making is centralized, apart from and above of the people, resulting in a weaker civil society that is not prone to participation

(Schofer and Fourcade-Gourinchas, 2001, p.811-2).

2) The State as a Catalyst for Trust While such structural accounts of the impact of the state on participation tend to be

preoccupied with networking, another set of scholars concentrate on how states shape norms

of trust. According to this view, best captured in the work of Rothstein, differences in the

institutional design of welfare states can stimulate generalized trust and confidence in

government (Kumlin and Rothstein, 2005). Specifically, scholars posit that social welfare

provided on a universal basis tends to build greater trust than when it is offered on a selective

basis (Rothstein and Ulsaner, 2005; Rothteinand Stolle, 2003).11 According to Rothstein and

Uslaner, universal programs enhance trust in three ways:

First, they are more redistributive than means-tested programs and thus create more economic equality. Second, since they are based on the principle of equal treatment and minimize bureaucratic discretion, they increase the sense of ‘equality of opportunity’ in the population. Third, means-tested programs exacerbate class and often racial division within society – and thus lead to less generalized trust and social solidarity and the perception of shared fate among citizens (Rothstein and Uslaner, 2005, p.42-3).

Therefore, those who espouse this view tend to place great focus on the importance of welfare state egalitarianism in shaping values and norms, and how this necessarily affects

11 Rothstein and Ulsaner (2005) examine the level of economic equality and equality of opportunity to find that states that promote greater equality tend to have higher levels of trust. Larsen’s (2006) findings suggest that high levels of social trust in Nordic states might be a result of the absence of a culturally distinct ‘underclass.’

267 civic life. While it does not conceptualize or even mention ‘citizenship,’ this is the only approach that provides a plausible theory explaining the relationship between inequality, civic engagement and the state developed in Part I. According to this logic, the more generous and all-encompassing a welfare state, reflected in a preference for universal programs - the more egalitarian it is because it strives to apply the same rules to all claimants and treats all citizens equally. For instance, scholars typically measure the more egalitarian, social democratic,

Nordic welfare regimes, which tend to opt for universal programs, against the least generous and most selective welfare state type – that of Anglo-Saxon, liberal welfare regimes, which use means-tests to rank citizens according to ‘deservingness’ in allocating welfare provisions.

This process of ranking or stratifying citizens, it is argued, has several deleterious effects on civic life. More generous, extensive benefits, that characterize the egalitarian welfare regimes, exemplified in countries like Sweden, allow the poor access to the same neighbourhoods, child care facilities, schools and hospitals, as those who are better off. In ensuring that the poor maintain a quality of life that closely approximates that of the average citizen, all- encompassing welfare regimes reduce socio-economically-based segregation and foster cohesion and trust among citizens (Larsen, 2008, p.155). Alternatively, means-tests are notoriously meagre, often failing to provide claimants with a standard of living that meets or surpasses the poverty line. Moreover, it is argued that means-testing requires a greater degree of street-level bureaucratic discretion (as opposed to the much more straightforward eligibility requirements required for universal programs), often leading to a system that is perceived as dishonest, biased and corrupt (Rothstein and Stolle, 2008; Rothstein and Uslaner, 2005). It is argued that this breeds distrust – both by administrators who suspect recipients will try to cheat the system, leading them presume recipients behave like criminals; and by criminalized

268 recipients, who feel they are being treated unfairly. This stratification of citizens, proponents assert, builds an unhealthy, lasting, wedge between the ‘underclass’ (claimants), and the rest of the population (Rothstein and Uslaner, 2005; Rothstein and Stolle, 2008; Larsen, 2006,

2008). Thus, it is believed that the degree of inequality – which is dependent upon the type of institutional structure which characterizes a welfare state - shapes trust; the more egalitarian a welfare regime – the more trust there will be in society, which stimulates civic activity.

3) Shortcomings of the State as Catalyst Approach These network and norm-based hypotheses about the role of the state in fostering civic

engagement are useful because they delve beyond simple state size and strength arguments, to consider how particular institutional logics and structures can work to encourage

participation. However, both approaches have inconsistencies and lacunas which cannot be

ignored, the first being that these studies are too varied to form a theory proper, and too few to

be considered a ‘literature’ per se.

One of the major problems stemming from the network-focused approach is its lack of

interest in the role of norms in civic activity to the point where trust and reciprocity either

appear as secondary to networking, or are overlooked altogether. It is unclear, for instance, which of Janoski’s political systems used by Curtis, Baer and Grabb (2001), or Jepperson’s statist/non-statist countries used by Schofer and Fourcade-Gourinchas (2001), build or erode trust, and why. Yet, a brief review of civic engagement literature reveals that trust is an

important component of civic engagement; among other things, it is said to reduce the risks

involved in collective action, and acts as an incentive to cooperate with others. Thus, while

scholars disagree on and remain undecided about the exact relationship between norms and

networks of sociability, most agree that trust is a key ingredient in building a vibrant civil

269 society and, as such, a solid understanding of civic engagement requires some attention be paid to trust.

Moreover, aside from some single-country case studies such as that of Skocpol in

America, and Worms in France, networking studies do not adequately account for the role of the state in shaping equality and citizenship through the rights and entitlements of the welfare state. For instance, Curtis, Baer and Grabb (2001) consider the impact of individual-level variables such as education and gender but do not elaborate on how these factors structure networking habits and opportunities. They invoke Janoski’s ‘political systems’ approach, which draws upon Esping-Andersen’s welfare regime typology, but fail to examine how states reinforce or alleviate inequalities depending on the type of ‘political system’ in question, and the implications this might have on participation (Curtis, Baer and Grabb, 2001, p.786).

In addition, the tendency of Curtis, Baer and Grabb (2001) to attach different explanations to account for networking patterns in divergent political systems leads to confusing conclusions which contradict their own hypotheses. This is the case, for example with the explanation given for the high levels of civic activity in two opposing political systems – the small, weak, non-interventionist Anglo-Saxon countries, and the large, strong, interventionist Nordic countries. The crowding-out hypothesis is said to explain civic engagement patterns in the former, with high levels of voluntary membership being the result of a lack of welfare state. However, rather than concede that the case of generous Nordic welfare states automatically poses a serious challenge to this theory, they appeal to a different explanation - revolving around the power of unions and political organizations in these countries, as well as the state’s allegedly hefty contributions to the third-sector (Curtis, Baer

270 and Grabb, 2001, p.786). While it is unclear how they determine that the vibrancy of the social democratic civil societies can be attributed to the generous public funding of voluntary groups, results from the comprehensive John Hopkins Nonprofit Sector Study simply does not support this claim. In fact, overall, table 5.3 suggests that between 2003-2005, just 26% of voluntary sector funding in social democratic countries came from the state, compared to 36% in liberal and 43% in corporate-conservative countries. Moreover, it turns out that corporate- conservative regimes, which Curtis, Baer and Grabb’s argument (2001) argue should have relatively high levels of activity by virtue of their tendency to provide generous third sector funding, in fact display some of the lowest scores in terms of membership and participation in voluntary organizations.

While network-focused explanations downplay the role of trust, norm-driven accounts which centre on how different types of states and institutional arrangements can build trust, tend to downplay the role of networks. Indeed, scholars such as Rothstein tend to place trust before networking in the causal chain and, as such, suggest that encouraging participation will be fruitless if levels of trust among members in the community are low (Rothstein and

Uslaner, 2005, p.44-5; Larsen, 2006, 2008). In fact, scholars such as Rothstein who focus on the role of the state in building trust, reverse the causal order between political and social trust that is presumed in the dominant social capital literature, revitalized by Putnam. Putnam

(1993, 2000) suggests that societies with high social capital produce greater amounts of social trust and more efficient institutions, which increases political trust. However, Rothstein overturns this logic, contending that vertical, political trust is a predictor for horizontal, social trust (trust in others). For instance, Rothstein and Stolle conclude that the most important institutions which can generate social trust are those that offer cues to citizens about a state’s

271

commitment to fairness and efficiency, including social service bureaucrats, the judicial

system, and police services (together, referred to as ‘order institutions’) (Rothstein and Stolle,

2008, p.445-6). According to this view,

...[I]f citizens think that these order institutions do what they are supposed to do in a fair, reasonably efficient, and unbiased manner, then they also have reason to believe that the chance people will get away with treacherous behaviour is relatively small. If so, citizens believe that most people have good reason to refrain from acting in a treacherous manner and thus conclude that most people can be trusted (Rothstein and Stolle, 2008, p.445-6).

In light of this, the results from a basic analysis testing the relationship between social

and political trust, as well as civic engagement, is found in table 5.4. In addition to the social

trust and civic engagement index scale already explained at the beginning of this chapter,

table 5.4 includes a political trust index built from three variables in the World Values Survey

(1990, 2000) which measure confidence levels in specific institutions.

Table 5.4 Predicting the Effect of Social and Political Trust: OLS and Logistic Regressions

Model 1 Model 2 Model 3

Independent Social Trust (Logistic) Political Trust (OLS) Voluntary Membership (OLS) Variable

B SE Exp B SE Beta B SE B Beta B (B) B Model 1 Political Trust .21** .01 1.23

Model 2 Social Trust .22** .01 .10

Model 3 Political Trust .22** .03 .04 Social Trust .77** .02 .22

R² --- .01 .05 Nagelkerke R² .02 ------

Sample Size 30554 30554 29625 Source: World ValuesSurvey, Four-wave Integrated Survey, waves 1990 and 2000. Note: Model 1 is a logistic regression, with logit coefficients, standard errors and odds ratio. Model 2 is an OLS regression with regression coefficients, standard errors, and standardized coefficients. ** Significant at the .01 level. Tolerance measures are above .20 indicating there is no multicollinearity problem.

272

The three variables measure for what Rothstein and others have referred to as ‘order institutions,’ assigning them particular importance in building cohesion. These include: the justice system, the police, and the social security system.12 The logistic regression between

the political trust predictor and the dependent variable, social trust, outlined in model 1,

reveals a positive and statistically significant relationship, suggesting that those with political

trust are more likely to also have trust in others. Indeed, calculating the probabilities from a

logistic regression, we find that political trusters are just slightly more likely (38%) to have

social trust than non-political trusters (33%).13 However, when reversing the hypothesis, and

testing for the effect of social trust on confidence in order institutions, we find another positive and statistically significant relationship, as outlined in the OLS regression results in model 2 of table 5.4. Moreover, as model 3 indicates, when testing the effect of both political and social trust on levels of voluntary membership, both are found to have a positive and statistically significant effect on changes in the dependent variable. Ultimately, social trust has a significantly greater effect on membership levels than political trust, suggesting there is little support here for Rothstein’s hunch that political trust comes before social trust.

The analysis in table 5.4 serves as a reminder that the relationship between social and political trust, and where these fit in the civic engagement causal chain remain, as of yet, unresolved and open for debate (Uslaner and Badescu, 2004; Gibson, 2001; Mishler and

Rose, 2001; Newton, 2001; Levi, 2000; Brehm and Rahn, 1997; Hall, 1999; Kaase, 1999).

While all these components – social and political trust, just institutions, and participation – are

12 Countries included in the analysis are: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Sweden, Great Britain, and the US. For the US, data on political trust is only available for 1990. The alpha score for the political trust index is .60. 13 Probabilities were calculated from the odds from the logistic regression output as follows: for non-political trusters: ODDS /1 + ODDS = 0.487/1.487 = 0.328 (or 33%). For Political trusters: ODDS /1 + ODDS = 0.599/1.599 = 0.375 (or 38%).

273 understood to be requirements for a healthy democracy, scholars continue to disagree on whether political trust spills over into social trust or vice versa; whether egalitarian, all- encompassing institutions encourage trust in others and political institutions, or trusting societies established fair, all-encompassing institutions; whether those who network develop greater trust in others and political institutions – or already-trusting societies are naturally more prone to being civically engaged. The analysis in table 5.4 found support for the dominant understanding of the relationship between social and political trust. Ultimately, however, theorizing a causal relationship between these factors appears to be a tricky endeavour at best.

Furthermore, the trust-centered approach to studying the state’s impact on civic activity offers too narrow a conceptualization of institutional design, contrasting states crudely in a simple universal-selective binary (Rothstein, 1998). For instance, building on the work of Titmuss (Titmuss, 1969.), Korpi (Korpi, 1980, p.303) and Sainsbury (Sainsbury,

1991, p.4), who distinguish between ‘marginal/residual’ and ‘institutional’ welfare models,

Rothstein makes a distinction between ‘selective’ and ‘universal’ welfare state. As table 5.5 shows, among other things, selective welfare states are those which favour minimal state involvement, spend little on social security, and tend to have means-tested, limited benefits which cover a small proportion of the population (Rothstein, 1998, p.20). Universal welfare states are those in which the state plays an important role, spends more generously on social security, and tends to have universal benefits with adequate levels which cover a majority of the population (Rothstein, 1998, p.20). However, all welfare states contain elements of selective and universal welfare states. Needs-tests are applied to claims to social assistance benefits – even in the ‘universal’ Nordic welfare state model, while social insurance schemes

274

and universal benefits exist in both corporate-conservative and liberal regimes in areas such

as family benefits, health care and pensions. Thus, Rothstein himself contends that he draws

“…a fairly coarse distinction… between universal and selective welfare policies…. This

means simplifying somewhat that benefits and services are intended to cover the entire

population throughout the different stages of life, and on the basis of uniform rules”

(Rothstein, 1998, p.19). Like Esping-Andersen then, Rothstein employs an ‘ideal-type’

model, where countries are categorized as ‘universal’ or ‘selective’ based on the prevalence of

one mode over the other. While this approach might adequately explain variation in civic

engagement levels between all-encompassing social democratic regimes such as Sweden and

selective liberal regimes such as Canada or the United States, it is less effective at explaining

where the corporate-conservative countries of Western Europe fit in this binary.

Table 5.5 Dimensions of Variation Between Selective and Universal Welfare Policy

Selective or Universal or Welfare State Dimension Marginal/Residual Institutional Model Model

Proportion of GNP devoted to social purposes Small Large

Benefit levels Minimal Adequate

Range of statutory benefits and services Limited Extensive

Proportion of population covered by benefits and services Minority Majority

Dominant type of Program Means-tested Universal

Role of programs preventing need for assistance in the first Nonexistent Critical place

Financing Fees Taxes

Significance of private organizations Great Small

Favored role of state direction Minimal Optimal Source: Rothstein, 1998, p.146.

Although Rothstein does suggest that generalized trust will not be able to take root in

countries with Catholic legacies and hierarchical cultures (Rothstein and Ulsaner, 2005, p.47,

275

49), traits which regime theorists happen to attribute to corporate-conservative regimes like

France, Italy and Germany, his recourse to a ‘universal/selective welfare state’ binary precludes any analysis of this complex set of countries. Consequently, determining where corporate-conservative regimes fall in the selective/universal model, outlined in table 5.5, becomes a confusing exercise, largely dependent upon which set of countries are being compared and which dimensions of social programs are analyzed (e.g. pensions, income security/social assistance, family policy). This can be demonstrated by contrasting the cases of social democratic Sweden, corporate-conservative France, and liberal Canada already reviewed in chapter 4. Like Sweden, France has historically opted for social insurance schemes that cover a majority of the population. Yet, unlike social democratic countries, the

French corporate-conservative legacy has led to a system which differentiates between citizens based on occupational-status, limiting redistribution to similarly positioned individuals. As such, the French scheme reinforces inequalities, which is said to weaken levels of trust (Rothstein, 1998, p.145; Esping-Andersen 1990). Thus, when compared to

Sweden, France appears to be a ‘selective’ welfare state. However, compared to a liberal regime like Canada, French social provisions are much more ‘egalitarian,’ offering a greater variety of programs and benefits that cover a greater proportion of citizens over a longer period of time. Comparative welfare state and feminist scholars have already highlighted the complexity, relative extensiveness, and generosity of corporate-conservative countries in the area of family and social policy (Huber and Stephens, Development 2001; Korpi, 2000; Korpi and Palme, 1998; Moller, Misra and Budig, 2007). Thus, while they may be less ambitious and more stratifying than social democratic countries like Sweden, the French welfare state tends to be more egalitarian and more closely resembles the ‘universal’ model when

276 contrasted to liberal welfare regimes like Canada. In light of this, an approach like Esping-

Andersen’s welfare regime typology, which accounts for such distinctions, is better equipped to explain civic engagement in corporate-conservative countries such as France or Germany, compared to a binary ‘selective/universal’ approach.

Finally, comparing the findings from network-based analyses and trust-centered approaches reveals a glaring inconsistency. Scholars who place great emphasis on norms, such as Rothstein, have argued that the more egalitarian and universal a welfare state’s design, the more political and social trust grows, encouraging citizens to become civically active. This argument is used to explain high levels of trust in universal social democratic countries, such as Sweden, compared to selective liberal countries such as the US. However, according to this argument, the high trusting social democratic countries should also lead in networking, while the less trusting liberal countries should lag. Yet, networking-based studies show that the meagre, means-testing liberal regimes have high levels of networking, in some cases, rivalling social democratic regimes. Instead, it is the corporate-conservative countries that lag in participation levels – a regime-type that, while selective in its own way, may nevertheless be considered more egalitarian than its liberal counterpart. This suggests that the degree to which welfare regimes are egalitarian might not explain overall civic engagement, or that social trust is disconnected either from institutional design or civic engagement in ways that are not adequately explained by existing theories.

To gain a better understanding of the factors which shape associational life, it is important to account for the role of the state. In view of this, this chapter examined the literature theorizing how the state affects civic engagement and uncovered two different

277 schools of thought. The first approach, espoused by dominant civic engagement scholars, views the state as a hindrance to civic life, linking minimal state intervention with higher levels of activity. The second approach views the state as a catalyst for participation and trust.

With respect to participation, single-country studies suggest that the state and voluntary organizations have long existed and developed in tandem, while cross-national studies reveal patterns based on political systems indicating that Nordic and Anglo-Saxon countries lead in networking compared to Western European countries. With respect to the state fostering trust, scholars argue that all-encompassing welfare states reduce social stratification, increase social cohesion, and foster the trust that is integral to civic engagement. Ultimately, these existing approaches are not only too scarce and varied to form a literature, but are also fraught with inconsistencies and contradictions. What is needed is an approach which can address some of the shortcomings of these models, while bringing order to the multiple theories and concepts.

To this end, the following chapter draws on Esping-Andersen’s welfare typology and World

Values Survey data (1990, 2000) to test for the impact of three dimensions of the state on civic engagement: interventionism, egalitarianism and statism.

278

Chapter 6 Interventionism, Egalitarianism, and Statism: The Impact of Three State Dimensions on Associational Involvement

In the pursuit to gain a better understanding of the factors which shape civic engagement, the previous chapter reviewed literature which sought to explain the effect of the state on associational life. One school of thought viewed the state as a hindrance to civic activity, crowding organizations out via comprehensive benefits and services. A second school of thought views the state as a catalyst for participation, with political systems and welfare states designed to promote civic activity and trust. However, the preceding chapter highlighted that these approaches contain internal inconsistencies, and their findings contradict one another.

Indeed, scholars disagree on which aspect of the state to focus on, how it should be measured, and how it actually affects civic engagement. Most civic engagement studies conducted until now have overwhelmingly focused on single-country case studies, examined on one aspect of civic engagement (i.e. participation or trust) and one dimension of the state (i.e. size and strength), and overlooked important regime-type differences.

Since no adequate theory of state exists which can adequately explain comparative civic engagement levels, this chapter outlines and tests theories which hypothesize how the state shapes associational life, using Esping-Andersen’s welfare regime-types as a case study.

Borrowing from the literature and approaches reviewed in the preceding chapter, three theories are developed and tested to determine which are most effective at explaining the role of the state in shaping networking and trust. Each theory, including interventionism, egalitarianism and statism, posits how particular dimensions of the state might lead to regime- specific participation outcomes. The interventionist approach proposes that countries which refrain from intervening in the market or family tend to encourage participation. The

279 egalitarian approach contends that comprehensive welfare states promote norms of trust and networking by treating benefit claimants equally and striving to reduce inequalities. Finally, the statist approach proposes that states in which political authority has historically been legitimized and located in society, as opposed to an authoritarian tradition, will foster civic involvement. Since Esping-Andersen’s regime typology takes political structure, institutional design and historical tradition into account, his model can be used to distinguish between interventionist/non-interventionist, egalitarian/non-egalitarian, and statist/non-statist regimes.

The findings support the egalitarian theory, with states which actively promote equality through generous, comprehensive social policies, such as in social democratic Nordic countries, leading in levels of trust and networking. In fact, highly egalitarian social democratic regime-types are found to significantly reduce the impact of SES on civic engagement. The results also lend support for the statist theory, with countries with a tradition of popular , such as in liberal Anglo-Saxon and social democratic Nordic regimes, demonstrating higher levels of civic involvement than corporate-conservative Western

European countries, where political power is centralized and kept apart from the people.

Ultimately, this chapter draws attention to the role of the state as a factor that can shape civic engagement. It shows that by implementing egalitarian policies and creating opportunities for participation, certain state-types have fostered civic engagement. THREE HYPOTHESES ON THE ROLE OF THE STATE IN CIVIC ENGAGEMENT In seeking to identify and understand the factors which shape civic engagement, the last

chapter surveyed the literature looking for theories and evidence on the effect of the state on

associational life. This analysis revealed that no adequate approach currently exists. These

approaches tend to be flawed, not only because they lead to contradicting conclusions, but also

280 because they do not provide an approach which: a) explains both fundamental aspects of civic engagement, including networks and norms of trust; b) accounts for the impact of different attributes of the state; and c) considers the experience of all regime-types. Building on the literature review in the preceding chapter, this section develops three theories on the impact of the state on civic engagement which resolve these issues, including: interventionism, egalitarianism and statism. These theories hypothesize the way in which specific dimensions of the state (interference in the market and family; welfare state egalitarianism, and the organization of political authority) affect networking and trust based on the characteristics of welfare regime- types.

Before outlining and subsequently testing each approach in greater detail, they are briefly reviewed here. Interventionism builds on the literature that views the state as a hindrance outlined in Chapter 5. According to the interventionist approach, non-interventionist states such as liberal Anglo-Saxon regime-types, reflected in low net total social spending measures, generate civic engagement because the minimal social provision they provide force citizens to meet their needs by forging associational ties, rather than relying on the state. Drawing on the literature that views the state as a catalyst, the egalitarian theory suggests that welfare states with all-encompassing benefits, treating all claimants equally, tend to reduce inequality, reflected in low post-tax and transfer Gini coefficients and poverty rates. According to this approach, egalitarian welfare states, such as that found in social democratic Nordic countries, foster civic engagement by building trust. Finally, drawing on the literature on cross-national variation in networking, the statist theory contends that non-statist regimes, measured through a qualitative review of the historical legacy of countries from scholars such as Jepperson and Esping-

Andersen, create a polity that invites feedback from civic groups, whether through pluralism, as

281 in Anglo-Saxon liberal countries, or the neo-corporatism found in social democratic Nordic regimes. The next three sections give a more detailed overview of each theory, how they are measured and defined, and what role they posit the state plays in shaping civic engagement.

1) The Impact of State Intervention on Civic Engagement The first hypothesis on the impact of the state on civic engagement is referred to as the

interventionist approach. It draws on the crowding-out theory, but distinguishes between

states based on the degree to which they intervene in the market, family and the everyday

lives of citizens – rather than their size or strength. From this perspective, non-interventionist

welfare regimes generate active civil societies because the lack of social programs leads

people to turn to each other to meet their needs whereas interventionist states deter voluntary

activity by encouraging citizens to rely on the state instead.

Esping-Andersen’s typology makes it possible to identify interventionist regime-types

relatively easily, as they are the ones which interfere in the market and family; these include

regimes that tend to implement de-commodifying policies (for instance a guaranteed basic

income, generous parental leaves, sick leaves, unemployment policies), as well as those which

de-familialize (for instance, providing elaborate childcare services) and familialize (for

instance, providing generous and extensive family benefits) via the welfare state. Thus, as

table 6.1 demonstrates, since liberal Anglo-Saxon countries are the least de-commodifying

and de-familialize via the private market, they are considered the least interventionist regime-

type, and should have higher levels of participation and trust compared to its interventionist

social democratic and corporate-conservative counterparts.

282

Table 6.1 Hypothesizing the Effects of Interventionism on Civic Engagement Across Welfare Regime-Types

Welfare Regime-types State Dimension Liberal Corporate-Conservative Social Democratic (Anglo-Saxon countries) (Western European countries) (Nordic countries)

High civic engagement due to Low civic engagement due to Low civic engagement due Interventionism minimal state intervention high state interventionist to high interventionist state

Historically, state interventionism has been measured by indicators of government

spending. In the case of the welfare state, this has typically meant accounting for gross public

social expenditure, which can be defined as “...the provision by public and private institutions

of benefits to, and financial contributions targeted at, households and individuals in order to

provide support during circumstance which adversely affect their welfare…” (OECD, 2007b,

p.6).1 Based on this measure, variation between welfare states is clear, as suggested by the

descriptive statistics in the first column of table 6.2. Overall, countries that fall under the

‘liberal’ rubric, namely, Anglo-Saxon countries, are inclined to spend less on social programs

and services than both continental Western European ‘corporate-conservative’ and Northern

European ‘social democratic’ countries (with the exception of Norway and the Netherlands).

Canada, for instance, spent 19% of its GDP on social expenditure in 2005, compared to

France’s 34% and Sweden’s 35% - a sizeable 15 to 16% difference. In light of their lower

gross social expenditure levels, proponents of the interventionist hypothesis would conclude

that liberal countries are smaller or weaker than their corporate-conservative and social

democratic counterparts, and should therefore benefit from greater levels of civic

engagement.

1 OECD social expenditures measures go beyond strictly public spending for a social purpose, to include private spending as well. Examples of private spending can include tax relief, as well as private health and pension plans which are “... often introduced by government regulation or fiscal intervention. Governments may force individuals and/or employers [for instance,] to take up protection provisions regardless of their risk-profiles or the prevailing market prices” (OECD, 2007b, p.8).

283

However, gross expenditures figures can be misleading, since they fail to account for the impact of the varied tax schemes put in place by different governments cross-nationally.

For instance, some states may tax cash benefits, while others use tax breaks as the preferred source of social provisions (Adema, 2000, p.191; Adema and Ladaique, 2009). Gross social expenditure measures also exclude other types of transfers which, while targeted at social protection, are not calculated because they come from the private sector, such as employer- provided sickness benefits and any occupationally-related contribution plan (Adema, 2000, p.191; Adema and Ladaique, 2009). Accounting for private social expenditures is important since some welfare regime-types favour market solutions over state intervention, as is the case with liberal regimes. As the second column in table 6.2 shows, aside from a few exceptions

(such as New Zealand, and the Netherlands), overall, in liberal countries, a larger share of social expenditure comes from private provisions. For instance, 25% of Canada’s total social expenditures come from private spending, compared to the 9% in France and Sweden – forming a 16% difference.

Thus, net social expenditure is a more accurate reflection of government spending.

When accounting for different tax schemes and private social expenditures, cross-national variation narrows substantially, but nevertheless endures; liberal countries, with their preference for private market solutions and minimal state intervention, continue to demonstrate a lower commitment to social expenditure as a % of GDP compared to social democratic and corporate-conservative welfare regimes. As the third column in table 6.2 reveals, liberal Canada spent 23% of GDP on net social expenditures in 2005, in contrast to corporate-conservative France’s 34%, and social democratic Sweden’s 29%.

284

Table 6.2 Gross, Net and Share of Private Social Expenditure Across Welfare Regimes

Gross public Share of Private Net total social social in total (private expenditure expenditure + Public) social 2005 (in % of Regime-type 2005 (in % of spending 2005 GDP at market GDP at (% of GDP) prices) market prices)

Liberal Australia 19 18 22 Canada 19 25 23 New Zealand 21 2 19 UK 24 25 30 US 17 39 27 Liberal mean 20 22 24

Corporate-Conservative Austria 31 7 27 Belgium 30 15 30 France 34 9 34 Germany 30 10 30 Italy 29 8 27 Corporate-Conservative 31 10 30 mean

Social Democratic Denmark 32 9 26 Finland 30 4 24 Netherlands 24 29 26 Norway 24 9 22 Sweden 35 9 29 Social Democratic mean 29 12 25 Source: Adema and Ladaique, 2009. Note: ‘Gross Social Public Expenditure’ is before-tax social spending; it does not include tax schemes or private spending. ‘Share of Private in Total’ is the amount of private spending out of the total of public and private social expenditures. ‘Net Total Social Spending’ is after-tax social spending, inclusive of tax schemes and private spending.

The findings in table 6.2 therefore suggest that countries’ social spending habits behave in accordance with Esping-Andersen’s welfare regime-based expectations, with the interventionist tendencies of social democratic and corporate-conservative regimes reflected in higher social expenditures levels compared to non-interventionist liberal countries.

However, a key drawback of such measures is that they do not consider the underlying goal of policies, or consider, as Esping-Andersen’s typology does, how states rank citizens differently through the design and distribution of social provisions. For instance, a high spending state

285 may, on the one hand, create programs and benefits that perpetuated gender roles by encouraging women to withdraw from the labour market to be full-time caregivers, or facilitate women’s dual roles as mothers and paid workers – or even encourage men to play a larger role in caregiving, as father’s quota systems in some Nordic parental leave schemes are said to do (Haas and Hwang, 2008; Ray, Gornick and Schmitt, 1996, p.192).

In light of its comprehensiveness, and since countries of the same regime-type generally cluster around similar social spending levels, ‘interventionism’ will be operationalized using Esping-Andersen’s typology, grouping welfare regimes according to the degree to which they are interventionist. Esping-Andersen makes it clear that corporate- conservative and social democratic countries have long accepted a great deal of state intervention into the market and family, reflected in their high levels of de-commodification, the degree to which citizens are free from market dependency (Esping-Andersen, 1990, p.21-

22), and generous family policies. The liberal countries’ preference for the market over the state is indicative of this regime-type’s non-interventionist legacy which, according to Esping-

Andersen, is reflected in their low levels of de-commodification, and de-familialisation via the market. As table 6.1 shows, in order for the results to lend support to the interventionist theory, liberal, Anglo-Saxon countries, with a tradition of minimal state intervention, would have the highest civic engagement scores, while the more interventionist corporate- conservative Western European and social democratic Northern European countries would have lowest scores.

2) The Impact of Welfare State Egalitarianism on Civic Engagement The second theory about the effect of the state on associational life that will be

examined here draws on dimensions of both the network- and norm-based analyses. On the

286 one hand, network-based studies largely overlook norms of trust, thought to be so central to civic engagement (Schofer and Fourcade-Gourinchas, 2001; Curtis, Baer and Grabb, 2001).

Moreover, they tend to downplay the effect of inequality and the welfare state in shaping participation. On the other hand, norm-based analyses typically put little emphasis on networking, and instead focus on norms which emphasize the importance of trust, equality and social cohesion (Rothstein, 1998; Rothstein and Ulsaner, 2005; Rothstein and Stolle,

2008). Network- and norm-based studies are in agreement on some key points; 1) both approaches acknowledge the important impact of long-standing historical legacies on states, institutions, and individual behaviour, and 2) both are interested in understanding the prominent role of the state in civic engagement. However, these approaches begin from different premises (i.e. that trust or networks are the most important explanatory factor for variation in civic engagement), rely on dissimilar theories and concepts, and reach contradicting conclusions.

Yet, an ideal study of civic engagement should include measures for both networks and trust, while paying attention to the impact of inequality and the role of the state in alleviating or reinforcing disparity. Focusing solely on trust can lead to an approach which conflates vibrant civic life with a high degree of social cohesion and lack of conflict, overlooking the experience and mobilizing efforts of groups struggling for equality and a host of adversarial movements. Focusing solely on networking can lead to a theory that fails to distinguish between the types of social ties that build norms of social and political trust, and those that do not. For instance, Rothstein notes:

...[M]any voluntary organizations and networks are actually built to instill mistrust of other people in general and of members of other organizations in particular.... All voluntary associations are not like the parent-teacher association (PTA) or bird

287

watching club; their raison d’être may be criminality or other forms of deviation that hardly generates interpersonal trust... (Bo Rothstein, 2005, p.101).

Thus, according to this view, taking norms into account is crucial to being able to make a distinction between groups that contribute to building social and political trust, versus those that do not. Ultimately, the most effective way to account for the effect of trust and networking is to include both in the analysis. Bringing together these two types of studies results in a new approach that places importance on both networks and norms, while focusing on the impact of the degree to which regimes work to address inequality and social stratification.

In light of this, the second theory about the impact of the state on civic engagement, developed here based on the literature reviewed in chapter 5, is the egalitarian approach. This model suggests that states which are not committed to promoting equality, and tend to stratify citizens, make it difficult to foster the generalized trust required for networking. Accordingly, egalitarian welfare states, such as those that avoid ranking citizens and offer the majority of the population a variety of social provisions with relatively high benefit amounts, will encourage civic engagement. This approach presumes that individual-level factors have an important impact on participation because disparity based, for instance, on gender, income or education level, and which can be alleviated or reinforced by the state, tends to erode trust and thus poses a challenge to collective action.

As the policy analysis in Chapter 4 confirmed, Esping-Andersen’s typology suggests that liberal regimes are least egalitarian, reflected in an approach to social security which tends to commodify and stratify citizens through meagre, means-tested benefits. Corporate- conservative regimes are moderately egalitarian, with more generous benefits which de- commodify and offer support to families – but are inclined to encourage traditional gender-

288 roles. Finally, social democratic regimes are understood to be the most egalitarian since they favour the use of universal policies which reduce social stratification, and offer generous social provisions which de-commodify and de-familialize. Thus, as table 6.3 shows, by combining the egalitarian theory and Esping-Andersen’s typology, it is possible to hypothesize how different welfare regimes will fare in terms of civic engagement. The more egalitarian social democratic Nordic regimes should lead in networking and trust, followed by the moderately egalitarian corporate-conservative Western European countries, with the meagre, selective, liberal Anglo-Saxon regimes having the weakest civil societies.

Table 6.3 Hypothesizing the Effects of Egalitarianism on Civic Engagement Across Welfare Regime-Types

Welfare Regime-types State Dimension Liberal Corporate-Conservative Social Democratic (Anglo-Saxon countries) (Western European countries) (Nordic countries)

Low civic engagement due to Moderate civic engagement due High civic Egalitarianism low levels of egalitarianism to a moderate levels of engagement due to egalitarianism high levels of egalitarianism

Even when using common indicators for welfare state ‘egalitarianism,’ which

typically centre on outcomes such as the effect of policies on inequalities between citizens,

countries belonging to the same regimes-type generally cluster together. This is certainly the

case with Gini coefficients, a standard statistic frequently used to measure inequality in

income distribution, which varies from 0 to 1, with 0 representing perfect equality, and 1

representing perfect inequality (Rothstein and Stolle, 2008; Rothstein and Uslaner, 2005;

Larsen, 2006; Van Oorschot and Finsveen, 2008). According to the first column of table 6.4,

which indicates post-tax and -transfer Gini coefficients across various countries, we can see

that countries generally cluster together based on regime-types. Overall, liberal countries,

which tend to have meagre, means-tested benefits, have the least equitable distribution of

289

Table 6.4 Income Inequality and Poverty Rates by Regime-type

Post Post Tax/Transfer Post-Tax/Transfer Poverty Post-Tax/Transfer Poverty Rate: % of Tax/Transfer Poverty Rate: % Rate: % of people living people living below 50% median income, Gini of people living at below 50% median income, by Gender and Family Type Coefficient, below 50% by Gender (early-mid-2000) Regime-type (mid-2000) median income (early-mid-2000) value (mid-2000) M Fem Differe Non-single Single Differe ale ale nce Mother Mother nce Liberal Australia 0.30 12 11 13 2 11 37 26 Canada 0.32 17 10 13 3 9 38 29 New Zealand 0.34 17 ------UK 0.34 8 9 11 2 9 40 31 US 0.38 17 11 14 3 11 40 29 Liberal mean 0.34 14 10 13 3 10 39 29

Corporate-Conservative Austria 0.27 7 6 7 1 5 18 13 Belgium 0.27 9 5 7 2 5 26 21 France 0.28 7 5 6 1 5 27 22 Germany 0.30 11 5 7 2 4 33 29 Italy 0.35 11 11 12 1 13 18 5 Corporate-conservative 0.29 9 6 8 2 6 24 18 mean

Social Democratic Denmark 0.23 5 4 3 -1 2 6 4 Finland 0.27 7 5 3 -2 2 9 7 Netherlands 0.27 8 3 5 2 3 33 30 Norway 0.28 7 5 4 -1 2 10 8 Sweden 0.23 5 5 5 0 2 10 8 Social Democratic mean 0.26 6 4 4 0 2 14 11 Source: Gini coefficient based on OECD, 2008, (Datafile available online through OECD Statistics). Poverty rate based on OECD.Stat Extract database, Income Distribution – Inequality. Gender poverty rates adapted from Gornick and Jäntti, 2010.

290 income out of all three regime-types, with a mean score of .34. Alternatively, the more ambitious Nordic social democratic countries and, to a lesser extent, corporate- conservative Western European countries, which offer a broader range of social security, covering a wider proportion of the population and with higher benefit levels, have lower

Gini scores and should therefore fare better for the most part. Corporate-conservative regimes have a mean score of .29, while social democratic regimes have the lowest level of inequitable income distribution, with a mean Gini score of .26.

Another common indicator used to test the degree to which states are egalitarian is the poverty rate. This calculation provides us with the number of people whose income falls below a certain income-level threshold, known as the poverty line. For our purposes here, the poverty rate is the score for any post-tax and -transfer income that falls below

50% of the median household disposable income. As the second column of table 6.4 shows, welfare regimes tend to score similarly in poverty rates, and do so in a rather predictable way, with Anglo-Saxon liberal regimes having the highest overall poverty rate of 14%, while corporate-conservatives have the slightly lower score of 9%, and with social democratic regimes have the lowest mean poverty rate of all at 6%. The other columns in table 6.4 show that welfare regimes also score similarly when we focus on the poverty rates of certain disadvantaged groups, in this case, women and single mothers. In fact, we see that the poverty rates for women and single mothers follow the same pattern as the total poverty rate, with liberal countries showing the highest levels of disparity, followed by corporate-conservative and social democratic countries. Thus, welfare regimes, for the most part, provide a consistent and reliable ‘egalitarianism’ ranking system, with social democratic regimes leading, followed by corporate-conservative

291 countries and liberal regimes lagging. However, the advantage of using regimes over single indicators of equality is that in addition to measure of equality that look at outcomes

(such as poverty levels), they also include information on inputs via regime-type design and logic (such as whether a regime is inclined to promote familialism, commodification or social stratification, reflected in eligibility rules and benefit levels).

3) The Impact of Statism on Civic Engagement Thus far, each hypothesis on the role of the state in shaping civic engagement, constructed from the literature reviewed in the preceding chapter, has focused on specific qualities of the state which characterize divergent welfare regime-types; the first centres on interventionism, and the second on egalitarianism. Borrowing from Jepperson (2000, 2002) and Schofer and Fourcade-Gourinchas (2001), the final theory developed here is referred to as the statism approach. This model links patterns in associational life with the way in which political authority has historically been located and legitimized. According to this view, countries may be distinguished based on their experience with ‘statism’ - the degree to which authority is vested in the state or civil society.1 In statist countries, authority is concentrated within a highly unified state structure, and activism is seen as unproductive and unwelcome

(Spencer, Murtha and Lenway, 2003, p.3). The state is understood as having the duty and ability to direct the public realm and non-statist activity – in the name of promoting a well- defined national interest, and there are clear demarcations between state and society (Janoski,

1998, p.109-111; Jepperson 2000, 2002; Esping-Andersen 1990; Schofer and Marion

Fourcade-Gourinchas, 2001). According to this theory, the way that political authority is

1 This approach is an adaptation of Jepperson since, in actuality, he looks at two areas of structuration – the state or ‘statism’ (the degree to which authority is vested in the state or society), and the society or ‘corporateness’ (the degree to which society is organized along corporate or associational lines). However, for the purposes of explaining social capital across regime-types, it seems statism provides the key. ‘Corporateness’ may be more useful in explaining intra-regime variations – such as differences between corporate-conservative France and Germany (Jepperson, 2002).

292 structured in statist countries should discourage civic engagement as an effective, legitimate form of participation, and emphasize the sovereignty and supremacy of the state in decision- making (Schofer and Fourcade-Gourinchas, 2001, p.813).

Conversely, in non-statist countries, authority and the legitimacy of state power rests in the civil society. Since it is understood to be an instrument of society, the state is guided by pressures from various interests in society who work to shape policy, whether it is through a competitive, adversarial pluralist system, or a bargaining, neo-corporatist system with groups formally integrated into decision-making (Janoski, 1998, p.108-111; Jepperson 2000, 2002;

Esping-Andersen 1990; Schofer and Fourcade-Gourinchas, 2001). Additionally, in non-statist countries, boundaries between the state and society are less defined, and “[s]tate structures are envisioned as coordinating and administering, rather than directing” (Jepperson, 2002, p.74).

Thus, according to this theory, non-statist countries should benefit from higher levels of civic engagement than statist countries because they “...have a more fluid demarcation between the public and the private spheres that allows ‘private’ actions to be more legitimate in the ‘public context’” (Schofer and Fourcade-Gourinchas, 2001, p.813).

Although this hypothesis does not focus on social policy or the extent to which policies alleviate or sustain inequalities, it is still closely tied to welfare regimes because the particular structure of a country’s political authority has an important bearing on the development and design of the welfare state. For example, statism, corporatism, occupationally-based social insurance and familialism are not the hallmarks of the corporate- conservative regime-type by happenstance. Rather, these strategies were put in place by historically Catholic, conservative countries at a time when capitalism and industrialization were eroding the traditional system held so dear by corporate-conservative countries,

293 including institutions (e.g. the Church and family), forms of social order (e.g. ), and modes of social security (e.g. guilds, corporations and mutual aide). Corporate-conservative regimes relied on statism to temper what were understood to be the harmful, commodifying excesses of the free market economy. This form of corporate-conservative political authority, exemplified by the cases of Germany and France, encouraged:

...[D]irect state intervention to grant social rights in order to enhance the integration of hierarchical society, forge a bond between workers and the state, maintain traditional relations of authority, and provide an opposing power to the modernist forces of liberalism and socialism. This led to the principle of ‘monarchical socialism’: ‘an absolutist model of paternal-authoritairan obligation for the welfare of its subject (van Kerbergen and Kremer, 2008, p.78). Esping-Andersen’s regime typology takes political structure, institutional design and historical tradition into account and, as such, his model can be used to distinguish between statist and non-statist regimes. Before outlining how different regimes fit in each category in more detail, it is helpful to first provide a brief description here. Specifically, liberal Anglo-

Saxon and social democratic Nordic countries are understood to have non-statist traditions, while one of the hallmarks of the corporate-conservative regime-type is a history of statism in the form of authoritarian rule. Thus, as table 6.5 indicates, by combining the statism theory with Esping-Andersen’s model, it is possible to hypothesize welfare regime-based patterns of civic involvement, with both liberal and social democratic regimes expected to lead in levels of networking and trust due to their respective traditions of popular sovereignty.

Table 6.5 Hypothesizing the Effects of Statism on Civic Engagement Across Welfare Regime-Types

Welfare Regime-types State Dimension Liberal regime Corporate-Conservative Regime Social Democratic Regime (Anglo-Saxon countries) (Western European countries) (Nordic countries)

High civic engagement Low civic engagement due to High civic engagement due to low Statism due to low levels of high levels of statism levels of statism statism

294

Just how each regime-type qualifies as statist or non-statist deserves further explanation. Liberal and social democratic regimes fall into the non-statist category, and as a consequence of political systems in which authority has never historically been hierarchically structured, or concentrated in the state, they should benefit from higher levels of civic engagement. For instance, in liberal countries such as Canada, the United States and, to a lesser extent, Great Britain, “…the political culture remains firmly centred on the idea of a self-governing society, largely autonomous from the state… [and where] the state derives its legitimacy from its function as the representation of civil society which is considered to be the principal locus of public life. The public bureaucracy is much less elaborate and rationalized than in statist systems” (Schofer and Fourcade-Gourinchas, 2001, p.811-2). This is reflected in the American tradition of championing , kept in check by a strong and vibrant civil society. Moreover, scholars have highlighted how social groups were able to shape decision-making early on – even as the state developed (Koven and Michel 1993;

Skocpol, 1995). As relatively new nations, the United States and Canada did not have a history of absolutism, a legacy of elite-driven corporatist estates, nor a strong “mythic elaboration of an authoritative centre” (Jepperson, 2002, p.71). Furthermore, Britain shifted away from elite-driven corporate estates to a more class-based system comparatively early and, despite its enduring monarchy, embraces the conception of Parliament as a representative of society (Jepperson, 2002, p.70).

It might seem contradictory to lump social democratic regimes - which gave rise to the

Nordic Model, alongside liberal countries – notorious for their preference for limited state, laissez-fair economics, and non-interventionism. After all, the Folkhemmet (the People’s

Home) - the Swedish welfare state, serves as a prime example of a typically statist project:

295 government rationalizing Swedish life in the name of national interest (Rojas, 2005).

However, ambitious, interventionist social democratic states, such as that of Sweden and

Denmark, are counter-balanced by an equally strong civil society. For instance, unlike many of its European neighbours where the held great sway, Sweden did not experience feudalism or despotic rule to the same extent. Although the Swedish state was constituted in the thirteenth century and was one of the largest early political entities in Western Europe, it was also one of the least populated, and the power of its limited aristocracy was stunted

(Rojas, 2005, p.8-9, p.49-50). Sweden’s early embrace of popular liberty can be traced as far back as the early fourteenth and fifteenth centuries, reflected in the inclusion of peasants in early deliberative assemblies, as well as the Council which elected Prince Magnus Eriksson as king of Norway and Sweden and had authority over taxes and the appointments of certain officials (Rojas, 2005). Thus, in Nordic countries, political authority and legitimacy have long been located in both the state, and the civil society. The historically close relationship between the state and the associational sector in Nordic countries has been documented, as has the potency of social movements which flowered during the nineteenth and twentieth centuries (Hirst and Bader, 2001, p.107-8). These movements managed to wield tremendous influence over the state and shaped decision-making - whether it be through Denmark’s alternative grass roots movements, or the popular movements and formal integration of divergent interests into the formal Swedish political process (Eyerman and Jamison, 1991, p.35).

In light of this, while the state in social democratic regimes may be described as extensive and interventionist, Jepperson argues that it would be misleading to equate its ambitious welfare states with the social engineering of statist corporate-conservative regimes.

296

In fact, scholars have argued that calls for greater state intervention in Denmark have typically come ‘from below’ (Siim, 2000, p.111), while others suggest an ‘absence’ of state altogether in social democratic regimes, highlighting that in countries like Sweden and Norway, the term for ‘state’ and ‘society’ may be used interchangeably (as in the term ‘welfare society’)

(Jepperson, 2002, p.73-74, and note #19; Trägårdh, 2007, p.259). Finally, various mechanisms have been entrenched into the decision-making process in social democratic regimes such as Sweden and Denmark for the purpose of fostering collaboration and consensus between various interests, including: corporatism which strives to integrate major groups into the policy-making process, encouraging bargaining and compromise; seeking out organized views to devise more widely accepted, non-controversial policies; and

Parliamentary procedures that seek out the views of interested parties and the opposition

(known as remiss in Sweden) (Castles, 1975, p.175-6; Rothstein, 1991).

The Dutch polity followed an evolutionary path distinct from its Nordic counterparts, which may explain why, in relation to the welfare worlds typology, it is considered an

‘imperfect fit.’2 Associational life in the Netherlands was stunted by measures such as the Le

Chapelier law which placed restrictions on associations until 1872, and political fragmentation which were entrenched in a system of pillarization, segregating life along the class- and confessionally-based cleavages of Calvinism, Catholicism, Socialism, and

Liberalism (Ertman, 2000, p.165-8). However, Dutch pillarization weakened dramatically in the 1960s, which not only worked to curb clientielism, but also created a political climate that was much more open, affording citizens with greater opportunities for participation (Blom

2 When focusing on income maintenance policies, Esping-Andersen (1990) concludes that the Netherlands belongs to the social democratic welfare world. However, he admits (1999) that when considering service delivery and the role of the family, it falls into the corporate-conservative category (p.87). Korpi and Palme (1998) categorize the Netherlands as a ‘basic security,’ liberal model, while many other scholars conclude that it is more in line with the corporate-conservative regime-type (Arts and Gelissen, 2002, p.151).

297 and Lamberts, 2006, p.483). Thus, according to Jepperson (2000, p.29-30), the Netherlands should be considered a low-statist country or, more precisely, a hybrid of the social democratic and liberal family of regimes.

While liberal and social democratic regimes are defined as non-statist and should therefore display high levels of civic engagement, corporate-conservative countries are described as having a long absolutist, étatist legacy which is expected to result in low levels of participation. For instance, France has a legacy of staunch absolutism that located authority and legitimacy in the state, rather than the civil society. According to this view, French associational life should be understood as the result of a series of oppressive regimes aimed specifically at curbing the power of other institutions such as the Church, as well as the civil society. For instance, following the French Revolution, the state suppressed associations, and did so for more than a century (from 1791 to 1901). Appealing to Rousseauian theory that suggested that the state should have monopoly over public interest, unimpeded by intermediary bodies such as associations, the 1791 Décret d’Allard banned guilds

(compagnonnage) and the loi Le Chapelier outlawed all voluntary associations. Later, the

Code Penale of 1810, suppressed groups that had more than twenty members without government consent (Archanbault, 2001, p.205; Edith Archambault, 1996). More recently,

‘foreigners’ associations’ were restricted in 1939 during WWII, only to be legalized again in

1981 (Raissiguier, 2003). According to this theory, it is likely that this unyielding challenge of associational life not only delayed the growth of the French civil society, but also shaped how that regime understands collective forms of participation, and constrained generalized trust (Vassallo, 2004, p.328; Worms, 2002; Archambault, 1996, 2001). Today, the interactions between the state and the civil society in corporate-conservative regimes such as

298

France continue to be comparatively limited, mediated through peak institutions (i.e. centralized labour and capital organizations) and their experts.

Similarly, corporate-conservative Italy’s historically repressive state also actively worked to limit associational life, positioning civil society as a threat to legitimate political authority – concentrated in the state. In a study of the development of civil society and democracy in Italy in the nineteenth century, Banti confirms: “[d]uring the Napoleonic era, the inclusive clubs were subject to inspection by the governments and later were banned outright… The picture did not change much in the fifty years preceding the unification of

Italy. The states that were re-established after the fall of Napoleon accepted the regulations on associations…” (Banti, 2000, p.46). Scholars have documented the historical divide between the Italian state and civil society (Banti, 2000, p.43, 51; Jepperson, 2000, p.31; Lyttelton,

2000, p.61-81), reflected in the highly exclusionary elite control of key associations, and the tradition of ‘localistic clientelism’ that reinforces ‘rigid vertical hierarchies’ (particularly in the south), compromising both political and general social trust. Compared to France and

Italy, Belgium has had less experience with a repressive state but it nevertheless falls in this statist category. Unlike the Dutch experience, Belgian pillarization endures, acting as an intermediary between the state and civil society. It segments social and political structures in the civil society including educational, economic, welfare, leisure, mass communication and political institutions along subcultural group lines. In the case of Belgium, not only does pillarization takes place on a class and confessional basis, but also via linguistic-regional lines of Flanders and Wallonia as well (Blom and Lamberts, 2006, p.483; Ertman, 2000, p.155).

Under such a system, it is suggested that “people are less citizens than clients” (Jepperson,

2000, p.30).

299

It is important to note that statism should not be equated with necessarily regressive or meagre social policies. Indeed, countries that can be slotted in the statist corporate- conservative regime-type, such as France, Belgium and Germany, have a tradition of promoting a wide range of social provisions with moderate-to-generous benefit-levels. This is because, according to corporate-conservative logic, state intervention is required to preserve important traditions and institutions, and help cushion citizens from the harmful effects of the capitalist economy by offering social rights through policies that will de-commodify workers, maintain long-standing corporatist hierarchies through occupationally-based insurance schemes, and promote traditional gender roles through familialism (van Kerbergen and

Kremer, 2008).

In some instances, the outcome of corporate-conservative state intervention has been surprisingly, if unintentionally, progressive. For instance, as early chapters have shown, despite familialist legacies, corporate-conservative countries often boast ambitious social provisions which help women balance work and family life, such as the highly effective

French and Belgian childcare systems, or German, Italian, and Austrian parental leave with relatively generous income replacement rates (Morgan, 2003; van Kerbergen and Kremer,

2008). While in the case of corporate-conservative regimes, state intervention is usually triggered by a desire to preserve social ordering rather than to promote equality as in social democratic tradition, this does not preclude the development of generous social provisions.

Finally, since the statism hypothesis places all the emphasis on how political authority is legitimized and where it is located, individual-level factors such as gender, race and class are not expected to have any significant effect on participation.

300

Table 6.6 Hypothesizing the Effects of Three State Dimensions on the Civic Engagement of Welfare Regime-Types

State Dimensions Welfare Regime-types Interventionism Egalitarianism Statism

High civic engagement Low civic engagement due High civic engagement Liberal regime due to minimal state to low levels of due to low levels of (Anglo-Saxon countries) intervention egalitarianism statism

Low civic engagement due Moderate civic Low civic engagement due Corporate-Conservative Regime to high state engagement due to a to high levels of statism (continental Western European interventionist moderate levels of countries) egalitarianism

Low civic engagement due High civic engagement High civic engagement Social Democratic Regime (Nordic to high interventionist due to high levels of due to low levels of countries) state egalitarianism statism

Thus, three theories positing the effect of distinctive dimensions of the state

(interventionism, egalitarianism and statism) on civic engagement have been outlined, attached to specific welfare regime-based participation outcomes. In fact, table 6.6 provides a clear overview of how each theory developed in this chapter and state dimension examined is expected to shape civic engagement across welfare regimes. The next sections employ cross- and multi-level analyses to test the validity of these theories. Exploring the Effect of the State on Civic Engagement: A Cross-National Analysis In testing the effects of the state on associational life, a descriptive analysis is conducted to get a sense of how welfare regimes rank across civic engagement indicators.

Data is drawn from the World Values Survey (1990, 2000), using the same twelve Western advanced democracies as in chapter 5, but recoded into a regime-type variable with three values, including liberal (Canada, the United States and Great Britain), corporate-conservative

(France, Austria, Belgium, Italy and Germany), and social democratic (Sweden, Finland,

Denmark and the Netherlands). Figure 6.1 shows the results of an analysis using the civic engagement index scale which incorporates four different indicators (voluntary membership

301 and activity, political action and social trust) to give us a preliminary, overall sense of civic engagement levels across all three regime-types. Details about these variables and the civic engagement index have already been outlined in chapter 5. The findings, detailed in figure

6.1, indicate that social democratic regimes lead, with 76% of respondents saying they are civically engaged, followed by liberal regime-types with 57% who are involved, and corporate-conservative countries scoring last with less than half of the respondents, 44%, saying they are civically engaged.

Figure 6.1 Overall Civic Engagement Levels by Welfare Regime-Type 80% 76%

70%

60% 57%

50% 44%

40%

30%

20%

10%

0% Liberal Regime Total Corporate‐Conservative Total Social Democratic Total

Source: World Values Survey, (1990-2000). Note: The relationship is significant (P < .001). Differences are statistically significant (p<.001). N = 17663.

The relationship between welfare regime-types and the civic engagement index is quite strong, positive, and statistically significant (at the .001 level). While this analysis establishes that the correlation between regime-types and civic norms and behaviour are valid and not purely due to chance, it does not give us any information about whether these differences are

302 statistically significant. Accordingly, we must rely on a one-way ANOVA - a test designed to account for the amount of variance in a dependent variable for two or more groups. The results indicate that the variation in levels of civic engagement between regime-types is indeed statistically significant (at the .001 level).

Although this is a very basic, preliminary result, thus far, it appears that examining networks and norms cross-nationally along welfare regime-lines leads to interesting conclusions about the validity of the three theories we are testing. The strong score held by the highly interventionist, ambitious, Nordic social democratic regime-type - even surpassing the non-interventionist liberal regimes – poses a challenge to the interventionist theory, which predicts that such strong, large welfare states will debilitate civic life. Moreover, the fact that the comparatively more generous Western European corporate-conservative regime-types come in last place, surpassed by meagre, means-testing and socially stratifying Anglo-Saxon liberal regimes, casts doubt on the explanatory power of the theory that egalitarian welfare states necessarily foster civic engagement. As such, cross-national civic engagement habits seem to lend credence to the theory that associational life is highly dependent on the political structure of regime-types. According to this view, corporate-conservative regimes lag in networks and norms of trust because they have strong statist legacies, while social democratic and liberal regimes lead in civic engagement because they have non-statist traditions.

While figure 6.1 presents data on welfare regime totals across overall civic engagement indicators, figure 6.1 breaks the index down into its various components, including a trust dummy variable, as well as a voluntary membership, unpaid work, and political index dummies. In addition to this, a four-category informal networks index

(network with friends; network with work colleagues; network with people from church,

303

mosque or synagogue; network with people from cultural or community group) which is not

included in the original civic engagement index because it is only available for the 2000 wave

of the survey, is added to figure 6.2.3

Figure 6.2 Five Civic Engagement Indicators by Welfare Regime-Type

85% % who hold a membership in any voluntary 54% organization 65%

44%

% who do unpaid work 27%

45%

32%

% who take part in political action 29% Social Democratic Total 29% Corporate‐Conservative Total Liberal Regime Total 93%

% who engage in informal networks 83%

89%

61%

% who say most people can be trusted 32%

43%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

% Civic Engagement

Source: World Values Survey, (1990-2000). Note: All relationships are significant (p <.001), and one-way ANOVAs (and z-scores for the trust dummy) show that all differences between indicators are statistically significant (p<.001). For voluntary membership, N = 35702; for unpaid work, N = 21483; for political action, N = 31834; for informal networks, N = 12366, for trust, N = 34707.

All relationships are positive and statistically significant (at the .001 level), with the

strongest relationships being between regime-type and voluntary membership, followed by

trust, unpaid work and with a weak relationship for informal networks and political action).

Moreover, one-way ANOVAs for the scaled indices, and z-test scores for the trust dummy, all

3 The alpha score for the informal networks index is .44.

304 reveal that differences between regime-types on the various civic engagement measures are statistically significant (at the .001 level).

Figure 6.2 indicates that social democratic regimes lead in four of the five categories, including voluntary membership, political action, informal networking, and trust.

Respondents in liberal regimes are actually most active, reflected not only in a high unpaid work score, but also in the fact they lead in membership and activity volume as well, as highlighted in figure 6.3. Organizational ‘volume’ variables, like the unpaid work variable, only considers the behaviour of those who are involved in associations, but they differ from other indices in that they account for the number of groups respondents belong to.

Figure 6.3 Membership and Activity Volume, by Welfare Regime-Type 18%

16% 16%

14% 14%

12%

10% 9% Liberal Regime Total

8% 7% Corporate‐Conservative Total Social Democratic Total 6%

4% 4% 3%

2%

0% % who are active in 5+ voluntary organizations % who hold 5+ memberships in any voluntary organization

Source: World Values Survey, (1990-2000). Note: The relationship is statistically significant (p < .001). Differences between welfare regime-types are statistically significant (p < .05). N = 23277.

Membership and activity volume variables were recoded into dummies to distinguish between those who belong to fewer than five networks, and those who are involved in at least five groups. The relationship between unpaid work, as well as membership and unpaid work

305 volumes on the one hand, and welfare regimes on the other hand, is weak, but positive and statistically significant (at the .001 level), and one-way ANOVAs reveal that differences between regime-types are also statistically significant (at the .001 level).

Thus, 16% of respondents in liberal regimes say they belong to five or more networks with 9% saying they are active members – this, compared to 14% in social democratic regimes who say they belong with 4% being active in those networks, and 7% of in corporate- conservative regimes saying they hold a membership in at least five voluntary organization with just 3% saying they are active in as many groups.

Figure 6.4 Organizational Involvement by Group-Type, Across Welfare Regimes

22% % Membership in Social or Service Groups 13% 24%

47% % Membership in Leisure Groups 26% 33%

33% % Membership in Employment‐related Groups 19%

25% Social Democratic Total

48% Corporate‐Conservative Total % Membership in Politically Inclined Groups 19% Liberal Regime Total 23%

30% % Membership in Religious Groups 13% 32%

14% % Membership in Other Groups 6% 12%

0% 10% 20% 30% 40% 50% 60%

Source: World Values Survey, (1990-2000). Note: All relationships are significant ( p< .001) Differences in proportions test scores show differences between welfare regime-types are statistically significant (p < .001). For membership in politically-inclined groups, N = 35702. For all other categories, N = 36739.

306

Finally, figure 6.3 shows that corporate-conservative regimes consistently have the lowest score across all civic engagement indicators. In sum, the findings show that the civic engagement patterns of welfare regimes provide support for the statism theory which posits that countries in which political authority is located and legitimized by the people will foster greater involvement than those with a history of authoritarian rule. Here, we see that across civic engagement indicators, social democratic and liberal regimes – opposites in egalitarian and interventionist terms – are both most active, and both have a non-statist popular sovereignty tradition.

In our pursuit to gain a greater understanding of the role of the state in civic engagement, we delving deeper into the formal networking habits of welfare regimes, considering membership levels in voluntary organizations divided into six group-types based on their goals or inclinations. As figure 6.4 shows, the findings reveal that social democratic regimes lead in four categories, including leisure networks, employment-related groups which include unions, politically-inclined groups and the open category of ‘other.’ Corporate- conservative regimes score last across all categories. Perhaps not surprisingly, liberal regimes lead in religious organizations as well as social or service-based associations. According to the interventionist theory, since liberal regimes are less interventionist, we should expect them to have a larger voluntary sector, with an especially large number of groups devoted to meeting social needs and providing services to the community (see chapter 5). Yet, while liberal regimes do lead in service and social groups, as outlined in figure 6.4, this is hardly the most popular type of network that liberal respondents say they join - nor is it the second or third most preferred group. In fact, contrary to interventionist theory expectations, social and service networks only rank as the fourth most popular type of network in liberal countries,

307 after leisure, religious and employment-related groups. Thus, figure 6.4 does not provide a ringing endorsement for the interventionist theory because it demonstrates the importance of types of networks that this theory largely dismisses such as leisure groups.

A similar analysis was conducted, this time to see how active welfare regimes are in different group-types. Figure 6.5 confirms that, once again, liberal countries are most active, leading in four of the five categories, including social and service, religious, politically- inclined and employment-related groups. Social democratic countries, however, are most active in leisure organizations, posing a challenge to the interventionist theory’s argument that ambitious, interventionist welfare states would erode associational life by replacing the role of organizations with public programs and provisions.

Figure 6.5 Unpaid Work by Group-Type, Across Welfare Regimes

12%

% Unpaid Work in Social or Service Groups 8%

21%

20%

% Unpaid Work in Leisure Groups 12%

17%

7% Social Democratic Total % Unpaid Work in Employment‐related Groups 4% Corporate‐Conservative Total 8% Liberal Regime Total

11%

% Unpaid Work in Politically Inclined Groups 7%

14%

8%

% Unpaid Work in Religious Groups 6%

19%

0% 5% 10% 15% 20% 25%

Source: World Values Survey , (1990-2000). Note: All relationships are significant (p < .001). Differences in proportions test scores show differences between welfare regime-types are statistically significant at the .001 except for ‘employment-related groups,’ which is significant at the .05 level. For membership in politically-inclined groups, N = 21483. For all other categories, N = 36739.

308

The relative activity levels of highly interventionist social democratic countries in contrast to the low score of highly interventionist corporate-conservative regimes suggests that the interventionist theory does not explain civic engagement effects here. Moreover, the high activity scores of the least egalitarian liberal regime-type, surpassing the more generous corporate-conservative countries, also suggest that the egalitarian hypothesis does not adequately explain cross-national associational outcomes. Rather, welfare regime-based civic engagement patterns found here once again appear to validate the political authority theory, suggesting that an authoritarian legacy has had a negative, lasting impact on corporate- conservative civic life.

Figure 6.6 Membership in Religious Organizations and Labour Unions, by Welfare Regime-Type 50%

45% 43%

40%

35% 32% 30% 30% Liberal Regime Total 25% Corporate‐Conservative Total 20% Social Democratic Total 14% 13% 15% 12%

10%

5%

0% % who belong to a religious organization % who belong to a union

Source: World Values Survey ,(1990-2000). Note: The relationship is statistically significant (p < .001). Differences in proportions test scores show differences between welfare regime-types are statistically significant (p < .001). N = 36739.

Before taking a closer look at cross-national levels of informal social ties, we consider a common argument encountered in the literature (Curtis, Grabb and Baer, 1992; Curtis, Baer and Grabb, 2001) which suggests that certain countries may be predisposed to having higher

309 levels of participation by virtue of the religious and social mobilization legacies which unduly skews comparative results. More precisely, it is argued that religious networks in Anglo-

Saxon, liberal regimes – particularly the U.S., and union memberships in Nordic social democratic countries, are so rampant that they ought not be considered voluntary networks.

In order to get a sense of how regime-types actually compare in terms of membership in religious organizations and unions, figure 6.6 shows the results of two separate bivariate analyses between regime-type and involvement in religious organizations and unions dummy variables. According to these findings, social democratic regimes do have a high rate of membership in unions, and liberal regimes lead in religious organization involvement, although social democratic countries also have surprisingly high levels of affiliation with religious associations.

Figure 6.7 Measuring the Impact of Unions and Religious Organizations on Membership and Unpaid Work Levels Across Welfare Regimes

77% Voluntary Membership Levels (%), Excluding 49% Unions 63%

82% Voluntary Membership Levels (%), Excluding 51% Religious Organizations 59%

71% Voluntary Membership Levels (%), Excluding 46% Unions and Religious Organizations 56% Social Democratic Total 43% Corporate‐Conservative Total Unpaid Work Levels (%), Excluding Unions 27% Liberal Regime Total 45%

40% Unpaid Work Levels (%), Excluding Religious 25% Organizations 40%

39% Unpaid Work Levels (%), Excluding Unions and 24% Religious Organizations 40%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

Source: World Values Survey, (1990-2000). Note: All relationships are significant (p<.001). For membership levels, N = 35702, and for unpaid work levels, N = 20483.

310

The relationships are strong, positive and statistically significant (at the .001 level) for both religious organizations and unions. Furthermore, the 29% and 31% gaps separating corporate- conservative and liberal regimes from social democratic levels are all statistically significant

(at the .05 level). Yet, despite the great role these associations play in their respective regimes, figure 6.7 demonstrates that removing unions and religious organizations from the unpaid work and voluntary membership indices does not affect the overall regime-based rankings.

Moreover, it is important to note that while compulsory union membership was a cross-national phenomenon ranging from European, North American and Antipodean countries, such practices have long been abolished (Visser, 2006, p.39). Thus, despite the exclusion of unions, social democratic countries still lead in voluntary membership levels, and despite the exclusion of religious groups, liberal regimes continue to lead, or be rivalled by social democratic regimes in the area of unpaid work.

Not only does including unions and religious groups in the analysis not appear to have a great impact on results, but perhaps more importantly, removing them from civic engagement indicators would mean omitting crucial facets of these regime-types. Unions and religious groups were often instrumental in offering poor relief and social provisions including unemployment and sickness benefits. Thus, since exploring the interconnection between the state, the civil society and welfare is a central goal of this analysis, it is precisely because of the tremendous influence these organizations have had on the design and institutional logic of welfare regimes that they must be included in the analysis (Schnabel and

Wagner, 2005, p.7; Esping-Andersen, 1999, p.16-18; Pestoff, 2009, p.42; Madsen, 2009;

Wadensjö, 2009).

311

One advantage of the civic engagement approach compared to traditional ways of

studying democratic participation is its consideration of social ties such as informal networks

with friends and colleagues as important to associational life. Informal networks are those

which require no membership dues or cards, and in groups that are not formally structured

(such as an established organizational or governance structure). Table 6.8 shows the results of

an analysis between welfare regimes and four different informal networks categories,

including associating with friends, work colleagues and people from places of worship and

other groups.

Figure 6.8 Informal Social Ties by Network-Type, Across Welfare Regimes

65%

% Informal Network with People from 57% Cultural or Community Group

64%

38%

% Informal Network with People from 40% Church, Mosque or Synagogue

52% Social Democratic Total Corporate‐Conservative Total 83% Liberal Regime Total

% Informal Network with Work Colleagues 66%

72%

99%

% Informal Network with Friends 97%

98%

0% 20% 40% 60% 80% 100% 120%

Source: World Values Survey, (1990-2000). Note: Differences in proportions scores show that all differences are statistically significant (p <. 05 level), except for the 1% differences between liberal and social democratic regimes in networking with people from cultural or community group. For informal networks with friends, N = 17223; with work colleagues, N = 14076; with people from church, mosque or synagogue, N = 15063; and people from cultural or community group, N = 15272.

312

As with formal networking, the findings demonstrate that social democratic countries have the highest levels of informal social ties, leading in three of the four categories – although there is very little variation across regime-types in this last category. This analysis suggests that Nordic countries are greatly involved in a variety of networks beyond just formal groups.

This might be interpreted as the expected outcome of long-standing egalitarian policies which have fostered high levels of trust among citizens for each other and their institutions.

However, if this were the case, corporate-conservative regimes, which are also relatively egalitarian, should have, at the very least, moderate levels of engagement; but the results show otherwise. Indeed, meagre, means-testing liberal regimes rival social democratic regimes in informal networks, and even lead in the category of networking with membership in one’s church, mosque or synagogue. Therefore, once again, the findings support the political authority hypothesis, suggesting that it is the tradition of popular sovereignty in

Nordic social democratic, as well as Anglo-Saxon liberal regimes - where political authority is located in and legitimized by the people - which best explains the role of the state in cross- national civic engagement levels.

Alternative forms of participation are increasingly included in civic engagement measures, particularly in the more recent studies that are more mindful of how socio- economic inequality can structure networking opportunities and capacity. By broadening the definition of ‘engagement’ to include such forms of collective action, there is a better chance of capturing the associational habits of a larger set of individuals, including those who favour alternative forms of involvement. Figure 6.9 shows the results of bivariate analyses between regime-type, and the five components which make up the political action index. We can

313

immediately see that all regimes are most active in ‘signing a petition,’ a form of participation

which requires relatively little effort.

Overall, liberal regimes are highly active in signing petitions and also lead in joining

boycotts, acts which may be considered highly individualistic. Moreover, they rival, and are

only barely surpassed by social democratic regimes in the category of ‘unofficial striking’ - a

form of participation that is much more collective in nature as the decision to strike is

typically tied to union membership and rules. This close score between liberal and social

democratic regimes is surprising considering the great power of and high membership levels

in unions in Nordic countries. What is most interesting, however, is that, for once, corporate-

conservative regimes lead in two categories which happen

Figure 6.9 Political Action by Type of Activity, Across Welfare Regimes

60%

% Who Have Signed a Petition 57%

76%

18%

% Who Have Joined a Boycott 10%

19%

26% Social Democratic Total % Who Have Attended a Lawful Demonstration 31% Corporate‐Conservative Total 18% Liberal Regime Total 8%

% Who Have Joined an Unofficial Strike 5%

7%

2%

% Who Have Occupied a Building or Factory 4%

3%

0% 10% 20% 30% 40% 50% 60% 70% 80%

Source: World Values Survey, (1990-2000). Note: All relationships are significant (p < .001). Differences between overall welfare regime-types are statistically significant (p < .05). For petitions, N = 35573; for boycotting, N = 34440; for demonstrating, N = 35205; for striking, N = 34450; for occupying a building, N = 34565.

314 to require a great deal of involvement on behalf of respondents: attending a demonstration, and occupying a building or factory. Yet it seems that even this engagement by corporate- conservative respondents can be attributed to the statist legacy in corporate-conservative countries. For instance the frequent and fervent opposition to the state in France can be interpreted as an inevitable reaction by citizens who feel they have largely been shut out of the decision-making process.

Thus far, the analysis has examined how welfare regimes fare across various civic engagement indicators, revealing that, overall, social democratic countries tend to lead in informal networks, membership levels, political action and trust, followed closely by liberal regimes which tend to be particularly active in organizations, with corporate-conservative regimes lagging. However, delving deeper into the relationship between the state and civic engagement is possible. OLS regressions offer a way to predict the impact of each hypothesized relationship between regime-type and civic engagement. While the dependent variable is the civic engagement index scale, independent variables represent each of the three hypotheses, consisting of countries recoded into three dummy variables according to whether they are interventionist, egalitarian or statist. Thus, for model 1 in table 6.16, the interventionism dummy includes corporate-conservative and social democratic countries as

‘interventionist’ based not only on traditional measure of social spending but also the degree to which regime-types interfere in the market and family as outline by welfare state scholars such as Esping-Andersen. Alternatively, liberal regimes are labelled non-interventionist. For model 2, egalitarian polytomous dummies categorize the more generous, universal, social democratic regimes, which also happen to have the most favourable Gini score and lowest poverty rate, as ‘most egalitarian.’ Corporate-conservative regimes are labelled as

315

‘moderately egalitarian,’ while the reference category, liberal regimes, are the least egalitarian

because of their preference for meagre, means-tested benefits, and high levels of inequality –

based on Esping-Andersen’s analysis, which also corresponds to the result of Gini

coefficients and poverty rates. Finally, for the statist dummy in model 3, corporate-

conservative countries are labelled as ‘statist’ while social democratic and liberal regime-

types qualify as ‘non-statist’ - drawing on Jepperson and the work of Schofer and Fourcade-

Gourinchas (Schofer and Fourcade-Gourinchas, 2001; Jepperson, 2002).

According to table 6.7, the negative coefficient in model 1 indicates that

interventionist countries have lower levels of civic engagement than non-interventionist

countries, lending support to the interventionist theory. This finding is surprising considering

how welfare regimes have ranked thus far on the various civic engagement indicators in the

descriptive analysis, and suggests that it may be the effect of the consistently low

participation scores of corporate-conservative regimes, weighing down the otherwise highly

engaged social democratic regimes.

Table 6.7 OLS Regressions Predicting the Effect of Welfare Regime Dimensions on Civic Engagement Model 1 Model 2 Model 3

Independent Variable Interventionist Theory Egalitarian Welfare State Theory Statist Theory B SE B Beta B SE B Beta B SE B Beta Model 1 Interventionism -.06** .02 -.023

Model 2 Egalitarianism Most Egalitarian .511*** .02 .178 Moderately Egalitarian -.341*** .02 -.141

Model 3 Statism -.56*** .02 -.231

R² .00 --- .05 Adjusted R² --- .08 --- Sample Size 17663 Source: World Values Survey, Four-wave Integrated Survey, waves 1990 and 2000. Note: Countries include: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Netherlands, Sweden, Great Britain, and the US . Each model includes unstandardized coefficients with standard errors and standardized beta values. *** p>.001; ** P > .01. Tolerance measures are above .20 indicating there is no multicollinearity problem.

316

Moreover, while the F-test indicates that the model as a whole is significant, it does not explain a great deal of the variation on the dependent variable (R² = .00). Model 2, which explains 8% of variance on associational life (adjusted R² = .08), indicates that social democratic regime-types have high levels of civic engagement and have the greatest impact on civic engagement levels, while liberal regimes surpass the more egalitarian corporate- conservative countries. Although this is in line with the findings outlined above, it is contrary to the egalitarian theory, which predicts that the least egalitarian liberal regimes should have difficulty fostering trust, and encouraging the development of associational ties. Finally, the negative coefficient for statism in model 3 demonstrates that statist corporate-conservative countries have lower levels of civic engagement than liberal and social democratic countries.

Not only is this consistent with the above analysis, but it also vindicates the statism hypothesis, which predicts that non-statist countries encourage a vibrant civil society because political authority is located and legitimized in society. In sum, the OLS analysis provides greatest support for the statist theory of the role of the state in shaping civic engagement.

While the findings provide some evidence for the interventionist theory, the model itself is too weak to be considered reliable. Moreover, the findings do not endorse the egalitarian theory. The following section involves a hierarchical analysis to control for both individual- and country-level factors, and determine whether any cross-level interaction takes place. In other words, it seeks to gain more insight into the relationship between SES indicators and welfare regimes, in light of the interventionism, egalitarianism and statism theories. Exploring the Effect of the State on Civic Engagement: A Comparative, Multi-Level Analysis Thus far, we have only tested for individual- and country-level effects separately. We saw for instance that, at the individual level, being a woman has a negative, significant effect

317 on participation, while SES has a significant positive effect on associational life. Such findings demonstrate the extent to which inequalities in civil society are replicated in participation. While it is individuals who engage in networks and do the trusting, shaped by their particular socio-economic context, we also saw how all of this takes place in countries with long-standing historical legacies and institutional logics which work to shape individual behaviour and norms. Thus, it is appropriate and useful to test for the effect of both individual- and country-level factors at once, and consider the possibility of any cross-level interaction. As such, hierarchical linear modelling is used to analyze civic engagement, accounting for individual and country-level effects (Raudenbush and Bryk, 2002; Luke,

2004). Such an approach is used when data and/or hypotheses involve multi-level or hierarchical structures, such as students nested in schools or, in this case, individuals nested in countries.

Drawing once again on the 1990 and 2000 waves of the World Values Survey, HLM 6 is used to build equations at two levels: level 1 which is the individual or within-country level, and level 2 which is the country-level.4 The individual-level variables included in the analysis are factors that have been found to structure associational life in the gender literature and in previous chapters, including: a gender dummy, a marital status dummy, a dummy which accounts children, and an SES index scale which incorporates income, education and profession (e.g. manual versus professional labour).5 The country-level variables are the

4 Especially large datasets such as that of the World Values Survey allow for the use of different methodological possibilities. For instance, an alternative to hierarchical linear modeling would be to adopt an approach similar to that used by Jusko and Shively (2005) who draw on meta-analysis, a variant of hiarchical linear modeling, to study cross-national public opinion data. For our purposes here, such a method would involve running independent regressions for each country, subsequently plotting the coefficients to reveal the effects of individual-level indicators (e.g. gender, SES), then running a regression to determine whether or not differences in coefficients between welfare regime-types are satistically significant. Given the more conservative nature of HLM estimates, however, I suspect that the results found here would hold in the alternative method as well. 5 The alpha score for the SES index is .50.

318 twelve advanced Western democracies, this time recoded into polytomous dummies according to regime-type. The liberal regime consists of Great Britain, Canada and the US; the social democratic regime is composed of Sweden, the Netherlands, Finland and Denmark; and finally, the reference category, the corporate-conservative regime, is made up of France,

Italy, Belgium, Austria and Germany. Determining how the state shapes civic engagement from this analysis involves assessing the findings in light of the interventionist, egalitarian and statism theories developed earlier (see table 6.6).

As a preliminary step, an unconditional baseline model is created, which only has an intercept (β0j). The central purpose of model 1 is to establish whether there is any significant variance in civic engagement between countries which warrants further investigation. The baseline system of equations is:

Level 1: CIVICENGij = β0j + rij

Level 2: β0j = γ00 + u0j

As Model 1 in table 6.8 demonstrates, there is significant country-level variation in civic engagement. The intra-class correlation (ICC), which considers the proportion of variance in the outcome (in this case, civic engagement) variable that is accounted for by groups, shows that countries are responsible for 14% of the variability of civic engagement, suggesting that we should pursue country-level effects further.6 We proceed by building a multi-level explanatory model, beginning with model 2, the random intercept and slopes model, which introduces individual level predictors (gender, marital status, having children, and SES). In model 2, the intercept and slopes are allowed to vary so that we can account for individual

6 2 To calculate the ICC, we turn to random effects: ρ = τ00/ (τ00 + σ ) . So, for model 1: .21711/(.21711 + 1.28936) = .1441183, or 14%.

319 level factors on civic engagement, and assume these might vary from country to country. The equations system for model 2 is:

Level 1: CIVICENGij = β0j +β1j (GENDER) + β2j (MARITAL) + β3j (CHILDREN) + β3j (SES) + rij

Level 2: β0j = γ00 + u0j β 1j = γ10 β 2j = γ20 β 3j = γ30 β 4j = γ40

Here, the intercept (γ00 = 1.029) is the grand mean of civic engagement across countries, controlling for gender, marital status, children and SES when their values equal 0 (in other words, for childless, unmarried, low-SES males). The negative coefficient for the gender slope (γ10 = -.099) indicates that women have lower levels of civic engagement, while higher levels of SES (γ40 = .025) lead to higher levels of civic engagement. Being married seems to have a negative effect on civic engagement while those who have children are more involved than those without any children – but these are not statistically significant. Examining components reveals, moreover, that there continues to be variation in civic engagement levels across countries (u0j = .218). It also confirms that the effect of gender (u1j) and SES (u4j) on civic engagement also varies significantly across countries. However, this is not the case for the marital (u2j) and children (u3j) slopes.

We proceed by building a means-as-outcomes model, model 3 in table 6.8, which predicts civic engagement using country-level variables – in this case, welfare regime-types.

The system of equations for model 3 is as follows:

Level 1: CIVICENGij = β0j + rij

Level 2: β0j = γ00 + γ01 (LIB) + γ02 (SOCDEM) + u0j

320

Table 6.8 Hierarchical Linear Models Evaluating Country-Level Factors as Predictors of Civic Engagement, with Cross-Level Interaction

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Intercept-only Conditional Model Means as Means as Full Model: Non-randomly Model with random Outcomes Model Outcomes with L1 Intercepts and Slopes Varying Slopes intercept and Covariates and as Outcomes slopes Random Effects FIXED EFFECTS For Intercept (β0j) Intercept (γ00) 1.865*** (.13) 1.029*** (.14) 1.48*** (.13) .064*** (.13) .719*** (.14) .719*** (.12)

Regime-type Liberal (γ01) .398 (.20) .484* (.165) .158 (.22) .143 (.19) Social Democratic (γ02) .897*** (.16) .801*** (.16) .865** (.21) .859** (.18)

For Gender Slope (β1j) Gender (γ10) -.099* (.04) -.100* (.04) -.176** (.05) -.182 *** (.03)

Regime-type Liberal (γ11) .119 (.08) .116* (.05) Social Democratic (γ12) .124 (.08) .114* (.05)

For Marital Slope (β2j) Marital (γ20) -.039 (.03) -.038 (.03) -.045 (.04) -.046 (.04)

Regime-type Liberal (γ21) .034 (.06) .036 (.05) Social Democratic (γ22) -.006 (.07) -.012 (.06)

For Children Slope (β3j) Children (γ30) .056 (.03) .058 (.03) .014 (.05) .008 (.04)

Regime-type Liberal (γ31) .079 (.08) .079 (.04) Social Democratic (γ32) .074 (.08) .061 (.07)

For SES Slope (β4j) SES (γ40) .025*** (.01) .025*** (.01) .261*** (.02) .263*** (.01)

Regime-type Liberal (γ41) .009 (.03) .011 (.02) Social Democratic (γ42) -.051 (.03) -.048 * (.02)

RANDOM EFFECTS Intercept ( u0j) .217*** .218*** .080*** .098*** .079*** .057***

Gender Slope (u1j) .009** .009** .008*

Marital Slope ( u2j) .002 .002 .003

Children Slope ( u3j) .003 .004 .004

SES Slope ( u4j) .001* .001* .001

Level 1 (e1j) 1.289 1.15 1.289 1.15 1.153 1.157 Source: World Values Survey, (1990, 2000), 12 Countries. For fixed effects: coefficients are L1 and L2 fixed effects, with standard errors in parentheses. For Random Effects, coefficients are variance components. Except for the Level 1 random effects component, which has no p-value: *** p > .000; ** p > .01; * P > .05.

The intercept (γ00 = 1.48) is now the average civic engagement score for the reference

category, corporate-conservative regimes. Although liberal regimes have a positive effect on

321

civic engagement (γ01), it is only social democratic regimes which show a significant increase in civic engagement over corporate-conservative regimes (γ02 = .897). The random effects indicate that significant variance between countries remains (u0j = .080). We can also see that the error term for the level-2 intercept is much lower than it was in the baseline intercept-only model (u0j = .217); this reduction in between-country civic engagement variance demonstrates the great impact of controlling for regime-type. Thus, in addition to finding that social democratic regimes lead as they have in most of the descriptive analysis (figures 6.1-6.9), in comparing the baseline model – without any controls - with model 3, we can conclude that

65% of country-level variance in civic engagement is explained by regime-type.7

The next step is to add individual-level covariates as random effects (allowing level-1 factors to vary across countries) to this means-as-outcomes model, outlined in model 4 of table 6.8. The system of equations for model 4 is:

Level 1: CIVICENGij = β0j +β1j(GENDER) + β2j(MARITAL) + β3j (CHILDREN) + β3j (SES) + rij

Level 2: β0j = γ00 + γ01 (LIB) + γ02 (SOCDEM) + u0j β 1j = γ10 + u1j β 2j = γ20 + u2j β 3j = γ30 + u3j β 4j = γ40 + u4j

Looking first at the fixed effects, the intercept (γ00 = .064) is the civic engagement value for level-1 when predictor values are at 0 (low-SES, unmarried, childless, males) in corporate- conservative regimes. The civic engagement levels for predictors at 0 in liberal regimes (γ01) is a significant .484 units higher than in corporate-conservative regimes, while social democratic regimes (γ02) have a significant .801 advantage in associational life over corporate-conservative regimes. In other words, even controlling for individual-level factors,

7 The raio is calculated as: the difference between the baseline intercept (u0j = .217) and the means-as-outcomes intercept (u0j = .080) = .14, divided by the unconditional variance: .14/.217 = .6451612, or 65%.

322 liberal and social democratic regimes have the greatest impact on civic engagement, contrary to the egalitarian and interventionist theories – and in accordance with the statism theory (see figure 6.6). Again, we find that women have lower levels of participation, those who have higher socio-economic status fare better in associational involvement, and the marital and children slopes continue to have no significant effect. The random effects results suggest that there continues to be variance between countries, indicating that gender and SES do not have the same effect across countries.

Finally, we fit the full intercept-and-slopes-as-outcomes model, model 5, which aims to explain this cross-national variation in gender and SES, and examine whether these can be attributed to differences in regime-types. The system of equations for model 5 is:

Level 1: CIVICENGij = β0j +β1j(GENDER) + β2j(MARITAL) + β3j (CHILDREN) + β3j (SES) + rij

Level 2: β0j = γ00 + γ01 (LIB) + γ02 (SOCDEM) + u0j β 1j = γ10 + γ11 (LIB) + γ12 (SOCDEM) + u1j β 2j = γ20 + γ21 (LIB) + γ22 (SOCDEM) + u2j β 3j = γ30 + γ31 (LIB) + γ32 (SOCDEM) + u3j β 4j = γ40 + γ41 (LIB) + γ42 (SOCDEM) + u4j

In this case, the intercept (γ00 = .719) coefficient shows that the average civic engagement score is .719 when predictors are 0 in corporate-conservative regime-types. The coefficients for the liberal (γ01) and social democratic (γ02) intercepts indicate that liberal and social democratic regimes have higher civic engagement levels than corporate-conservative regimes, but only in the case of social democratic regimes are these effects significant. Once again, we find that women have lower levels of civic engagement than men, those with higher

SES are more civically involved, and that marital status and having children has no significant effect on associational life. The fixed effects for model 5 can be interpreted as interaction terms between welfare regime-types on the one hand, and level-1 variables on the other. For

323

example, for the liberal (γ41) and social democratic (γ42) coefficients, model 5 indicates that the slope for SES increases in liberal countries (by .009 units), and decreases in social democratic regimes (by .051 units). Thus, the relation of SES to civic engagement is stronger in liberal and corporate-conservative countries, than it is in social democratic regimes.

However, none of the cross-level effects are found to be significant.

Looking at the variance components under the random effects section in table 6.8 for model 5, we find that the error term for the level-2 intercept (u0j = .079) is significant, indicating that average civic engagement levels, controlling for level-1 effects, varies significantly between countries. However, here we see that the only slope for which significant variance between countries remains is for the relationship of gender to civic engagement (u1j = .008). Since all random effects of model 5 slopes, with the exception of gender, are found to be insignificant, indicating there is no variance between countries, a final non-random model is run. This analysis, outlined in model 6 of table 6.8, replicates the full model (model 5), except that here, slopes are constrained so that they vary non-randomly from country to country. The system of equations for model 6 is:

Level 1: CIVICENGij = β0j +β1j(GENDER) + β2j(MARITAL) + β3j (CHILDREN) + β3j (SES) + rij

Level 2: β0j = γ00 + γ01 (LIB) + γ02 (SOCDEM) + u0j β 1j = γ10 + γ11 (LIB) + γ12 (SOCDEM) β 2j = γ20 + γ21 (LIB) + γ22 (SOCDEM) β 3j = γ30 + γ31 (LIB) + γ32 (SOCDEM) β 4j = γ40 + γ41 (LIB) + γ42 (SOCDEM)

As in the last model, model 5, we see that liberal (γ01) and social democratic regimes (γ02) continue to have a positive effect on civic engagement, with levels that surpass that of corporate-conservative countries, although this is only a significant effect in the case of social democratic countries (γ02 = .859). Once again, the estimated mean slopes for gender (γ10 = -

.182) and SES (γ40 = .263) indicate that women continue to have significantly lower levels of

324 civic engagement, while those who are better-off socio-economically enjoy significantly higher participation levels. However, while this is not the case for marital status and having children, model 6 does offer significant cross-level interactions in the case of gender and SES.

In the case of gender, we find that liberal (γ11 = .116) and social democratic countries (γ12 =

.114) demonstrate a stronger relationship between gender and civic engagement than in corporate-conservative countries. Additionally, while liberal regimes do not have any significant inter-action with SES (γ41 = .011), the significant but negative coefficient for social democratic countries (γ42 = .-.048) reveals that the SES effect on civic engagement is stronger in corporate-conservative countries. This could be interpreted as support for the egalitarian hypothesis which positions social democratic regimes as the most egalitarian – above liberal and corporate-conservative countries; according to this view, since social democratic social policies narrow inequalities in these societies, SES should have less of an impact on civic engagement than in other regime-types. Finally, the random effects show that the error term for the level-2 intercept (u0j = .057) is significant, indicating that average civic engagement levels varies significantly between countries, however the much smaller intercept in model 6 compared to our first baseline model shows that the controls have a great effect on our outcome variable, civic engagement. In fact, comparing this to model 1, we find that model 6 explains 74% of country-level civic engagement.8 Considering the Findings Examining civic engagement across divergent welfare regime-types demonstrates that social democratic regimes are highly engaged, leading in norms of trust as well as diverse forms of participation, including informal networks, membership levels and political action.

8 This is calculated as follows: calculate the difference between the u0j for model 1 (.217) and model 6 (.057) (.217 -.057 = .16 / the unconditional model variance (.217) = .737327 or 74%.

325

Liberal regimes also tend to be highly involved, often rivalling social democratic countries, and surpassing them in active participation. Despite their penchant for certain forms of adversarial political action which require a great deal of effort on behalf of respondents, namely, joining a demonstration and occupying buildings or factories, corporate-conservative regimes consistently have the lowest scores in trust and associational life.

What does this analysis tell us about the role of the state in civic engagement, particularly with respect to the three theories we seek to test? It confirms that welfare regimes do shape civic engagement – in some cases, even when controlling for individual-level factors; however this relationship is quite complex. In seeking to explain cross-national civic engagement levels, the above analysis uncovers some support for both the egalitarian and the political authority theories. The civic engagement regime-type rankings that are found pose a serious challenge to the interventionist theory which cannot explain the vibrant associational life of strong, large and highly interventionist Nordic social democratic countries. While the

OLS regression findings in model 1 table 6.7 suggest that non-interventionist states explain a greater amount of variation on the dependent variable, the model was too weak to be considered reliable. Moreover, the great popularity of expressive organizations and adversarial forms of participation goes against the logic of this theory which focuses on socially cohesive aspects of associational life, and presumes that citizens’ prime motivation for networking is to meet their needs with minimal help from the state by forming service organizations. The descriptive (figures 6.1-6.9) and regression analyses of civic habits also fail to provide a forceful endorsement of the egalitarian theory in light of the strong level of organizational involvement by the liberal regime-type, and the weak performance by the relatively egalitarian corporate-conservative European countries. Contrary to the egalitarian

326 hypothesis, a historical preference for social stratification and de-commodification through modest, stigmatizing, means-tested provisions, as well as de-familialization via the market, does not appear to have thwarted civic life in Anglo-Saxon countries.

Rather, the political authority hypothesis seems to best explain how it is that two otherwise polar opposite regime-types - the highly interventionist and egalitarian social democratic countries, and the least interventionist and egalitarian liberal countries – can both be avid civic engagers. According to this view, the degree to which the states encroach on the market and family, or their commitment to building solidarity through egalitarian policies, does not fully or adequately explain comparative civic habits. Instead, emphasis should be placed on the way power is structured between the state and society. Specifically, countries with non-statist legacies such as that of the liberal and social democratic regimes, where the boundary between the state and society has historically been small, and offering opportunities to participate in the decision-making process (whether through pluralism or corporatism), generates a high level of movement among citizens. Even the one area in which the otherwise stunted civic life of corporate-conservative regimes thrives – protesting - appears to support this theory since such adversarial, antagonistic forms of political action can be seen as the inevitable outcome a political system where the long-standing concentration of power in the central state has led frustrated citizens to view the authoritative mythic state as rival

(Jepperson, 2000; Jepperson 2002). The finding that statism, which offers fewer avenues for citizens to voice their opinions and organize to solve problems through collective action, seriously impedes the development of networks and norms of trust, suggests that greater attention must be paid to the way political authority is structured.

327

Yet, in fact, it would be misleading to conclude from this that political authority explains the whole comparative civic engagement story. Most notably, the political authority account fails to consider the impact that inequality continues to have on participatory democracy. Indeed, both the individual- and multi-level analyses, which take into account how citizenship status structures associational life, illustrate that social and economic factors such as gender and class have a consistently strong, significant effect on civic life, even controlling for country-level regime-type. In fact, including inequality and regime-type concurrently in the analysis actually leads to different results. Thus, while both liberal and social democratic regimes are found to have a positive effect on civic engagement in the cross-national descriptive and OLS regression analyses as predicted by the political authority theory, this effect is only consistently significant for the latter in the multi-level analysis. In other words, when both individual- and country-level factors are considered as is the case in the hierarchical modelling approach, the findings lean more towards a partial endorsement of the egalitarian theory. This is not only reflected in the fact that only social democratic regime- type have a positive, significant effect on civic engagement levels (γ02), above both corporate- conservative regimes, as well as liberal countries – which largely have no effect, but also in the inter-action effect in model 6 of table 6.8, where the highly egalitarian social democratic regime is found to have a weaker relationship between SES and civic engagement (γ42).

According to this line of reasoning, the narrowing of inequalities in Nordic countries due to highly egalitarian policies would mean that SES is not as great a factor in civic engagement in these countries as it might in the less egalitarian corporate-conservative and liberal regimes.

In other words, accounting for both country- and individual-level factors at once brings to light the impact that generous, egalitarian social democratic policies have on participation in

328 these regimes; they are so effective at reducing inequalities, that the effects of social and economic conditions, which typically pose a challenge to civic activity, are significantly tempered in these countries.

It is only a partial endorsement of the egalitarian theory, however, because the same effect does not take place when we look at the surprising interaction between welfare regimes and gender in model 6 of table 6.8. As predicted by the egalitarian theory, we find a positive, significant interaction effect between the least egalitarian liberal regime and gender, indicating that gender matters more in these countries to civic engagement than they do in the moderately egalitarian corporate-conservative reference category. However, contrary to the theory, we find that the most egalitarian social democratic regime-type also has a positive significant effect where we would expect a negative coefficient, reducing the importance of gender differences to civic engagement because of the lower levels of inequality. Perhaps gender differences appear to have less of an impact on associational life in corporate- conservative countries because, in these Western European countries, everyone is equally withdrawn from civic life, or perhaps the very generous benefits targeting women in such countries, as outlined in the French case in Chapter 4, contributes to this effect. Nevertheless, this is a puzzling and unexpected result, particularly within the greater context of the analyses as a whole.

This analysis of three hypotheses of the state on civic engagement highlights that welfare regimes have distinctive impacts on associational life. It also demonstrates the value of including both individual- and cross-national factors at once. The findings suggest that states can encourage civic engagement depending on how they choose to address inequalities and

329 social stratification, and in the way political authority is structured and legitimized by society.

As such, implementing social policies which are aimed at narrowing disparities and traditional gender roles helps generate a more active, more trusting community. However, it is just as important that the state create opportunities for participation in the first place, to encourage civic engagement.

330

Conclusion A Recipe for Participatory Democracy

The dominant civic engagement literature has shown that participation lies at the heart of democracy. Being able to exercise some control over issues that matter to and affect us is integral to a system of governance that is defined as ‘rule by the people.’ Equality is an important component of a healthy democracy because the needs and interest of some should not be privileged over others. All citizens should have the chance to convey their concerns and decision-makers should have the opportunity to consider all concerns. Civic engagement theory proposes that civic involvement in networks of sociability builds norms of generalized trust and that, together, these create a virtuous circle in which networking and trust are reproduced. This virtuous circle is said to lead to beneficial outcomes for individuals, communities and systems of governance alike. The dominant civic engagement literature has devoted itself to documenting what civic involvement can do. It has found that, while withdrawal from the community can lead to social ills, civic activity allows average citizens to voice their concerns to decision-makers, develop important skills that are transferable to other forms of participation (such as organizational and communication skills), recruit others to their cause or in support for their community, and stimulate greater interest and efficacy in politics. It has been linked to stimulating economic growth, promoting good health, and lower high school dropout and crime rates. The literature has also found that institutions and communities function best when citizens are civically involved and trusting. Even aside from its beneficial outcomes, civic engagement is ultimately an end in itself. A life without social ties would be one of powerlessness, loneliness, and without dignity; but, having the ability to meet with others to exchange ideas, deliberate and provide and get support is empowering,

331 fulfilling and freeing. In light of its instrumental and intrinsic value, it is critical to understand what shapes civic engagement, to determine whether some have a participatory advantage over others and why, and learn how civic norms and behaviour can be fostered.

To this end, this study explores key factors which shape civic engagement at two levels of analysis. Part I examines individual-level factors, exploring the effect of social and economic conditions on participation and trust, using women as a test case. Part II examines country-level factors, exploring how the state impacts civic life, using Esping-Andersen’s welfare regimes as case studies. Using a combination of comparative and multi-level statistical analyses, extensive literature and document reviews, personal interviews, and grounded in a gendered citizenship perspective, this study has found that inequality and institutional structure, ranging from eligibility rules for social programs to how political authority is structured, have an important effect on associational life. An analysis focusing on gender indicates that social and economic inequalities are reproduced in levels of civic active and trust. This finding underscores the extent to which social and economic context structures networking, with those belonging to disadvantaged groups facing a disproportionate number of challenges to participation, including inadequate resources required to participate, and fewer opportunities to belong to a wide assortment of networks which can lead to new contacts or the development of civic skills. Some state-typess, have been more actively committed to reducing the inequalities which stunt organizational involvement by implementing egalitarian policies that minimize social disparity and stratification. However, it seems that addressing inequalities is not enough, as the story of Western European corporate-conservative welfare regime-types have shown. The findings of this study indicate that the manner in which political authority is structured is also of vital importance to

332 fostering active and trusting societies, signalling the importance of creating opportunities for participation and dismantling boundaries between the state and society. This study contributes to the existing civic engagement scholarship in its unique focus on citizenship and individual-level context, drawing on the case of gender differences in participation; and the value it places on women’s experiences and child- or care-networks; the link it draws between welfare regime-types and differences in citizenship status; and its emphasis on the impact of the state on civic life.

Citizenship, Context and Gender Differences in Participation This study highlights the extent to which the dominant civic engagement literature fails to account for the effect that inequality has on participatory democracy. This is reflected in their appeal to a narrow view of citizenship, grounded in formal equality, which presumes that individuals are autonomous agents, unattached and unaffected by social or economic factors. The proposition that associational life be considered from a feminist citizenship perspective, rooted in substantive equality, is an important contribution this study makes to the existing civic engagement literature, bringing to the fore the impact of difference, social and economic conditions, and power dynamics to civic activity and trust. It leads to an examination of civic life which better accounts for the way networking is constrained by social and economic factors, and reveals the extent to which inequalities and the public and private divide are political constructs that are capable of change.

The analysis of gender differences in participation, based on an analysis of ISSP and

World Value Survey data, indicates that women tend to belong to the types of groups that are commonly associated with them, namely expressive and informal networks such as kin ties, as well as social welfare and religious organizations. However, this analysis shows that

333 women are also surprisingly active in politically-inclined groups. Moreover, they tend to have a higher membership volume, and are more actively engaged in associations than their male counterparts. Yet, despite their religious activity, a form of participation that is linked to greater overall associational involvement, and despite their preference for politically-inclined groups, overall, the findings paint a grim picture of gender-based participatory differences. It finds that women are generally less trusting and less civically involved than men across an array of different indicators, and suggests that a negative relationship exists between class- and status-based inequalities, and civic engagement. In addition, the findings indicate that a complex relationship between gender, race and class exists indicating that, in some cases, race has a stronger effect and can better explain SES than gender. The analysis in Chapter 2 revealed that race explains away gender differences for non-white individuals when looking at key SES indicators, education and income. While quantitative studies tend to find that non- white citizens are less active and trusting that white citizens, qualitative studies, such as that of Runyon Heights, emphasize that racial and ethnic communities have long been highly active. However, these networks are often bonding ties – making it difficult to produce the benefits one might expect from civic engagement. Thus, the case of Runyon Heights serves as a reminder that networking alone is not enough; what is required are bridging ties with different others. For Black community members in Runyon Heights, a history of residential segregation translated into a highly involved but segregated associational life. However, those most interested in promoting the community’s needs and interest recognized that they would have to forge ties with white neighbours in order to accomplish their goals and secure resources.

334

Explanations for women’s comparatively narrower associational lives are inextricably connected to their citizenship. A depreciated citizenship status, sustained by enduring traditional gender roles and the compartmentalization of the polity into tidy but illusory public and private spheres of activity and influence, lead women to have lower stocks of resources and levels of interest in politics; to surround themselves with tight-knit, kin-based, limiting social ties; and even renders the types of networks they belong to, such as care or child- related ties, invisible from those who measure and study civic engagement. On the other hand, this study’s results also reveal that women with children appear to be more involved than those without children, suggesting that it is misleading to conceptualize family and civic life as two separate worlds, or assume that caregiving necessarily orients women away from politics and the community. Ultimately, the findings suggest it is time to acknowledge the value of child- and care-related networks, lending support to the argument that civic engagement studies should be broadened to include such networks in their analyses. Women’s Experiences and Child- or Care-Networks Feminist scholars have argued that the dominant approach to measuring civic participation tends to overlook certain important sources of women’s networking. Moreover, the impact of women’s dual roles as paid workers and primary caregivers on civic engagement remains understudied because the dominant literature’s understanding of citizenship takes the public and private divide for granted, excluding any activities that take place in the home or family. Moreover, such questions are also largely excluded from surveys.

Putnam speaks of a ‘double-edged sword’ when referring to women entering the workforce because, while this advancement created new networking opportunities for them, it simultaneously reduced the amount of time they had to participate (Putnam, 2000, p.196).

Alternatively, Herd and Meyer contemplate childcare as both a hindrance and catalyst to

335 women’s networking (Herd and Meyer, 2002, p.669-73). In order to gain some insight into how women’s experiences influence their networking habits, and take a closer look at informal and care networks, personal interviews were conducted with sixty single mothers in

Canada, France and Sweden.

While the findings indicate that respondents across all three countries are quite active, the interviews confirm what preceding chapters have also found; that liberal and social democratic welfare regime-types, represented here by the case of Canada and Sweden, are the most trusting and civically involved, while corporate-conservative countries, represented by the French case, are the least. Balancing work and primary cargiving does put a great deal of stress on single mothers across all countries, with respondents saying that care work and the resources it demands (time, money, energy) poses a challenge to participating in associations.

However, despite this, single mothers interviewed nevertheless found a way to remain highly engaged, often relying on various forms of informal networking for support and caregiving.

In fact, in a many cases, interviewees asserted that children are the origin of networks.

Respondents did not engage in networks merely to ‘get by,’ nor did they do so to ‘get ahead’

(Lowndes, 2004, 2006). Instead, most joined formal and informal groups to satisfy their desire for social interaction, whether rooted in a need for belonging, to satisfy a curiosity, gain support or information, or merely for leisure. Thus, the interview findings suggest that researchers seeking to explore how certain types of networks might better translate into political clout or engagement must take into account the underlying the motives and goals of network members.5 Moreover, they reveal that what might otherwise be relegated as the discrete lives of mothers in the private realm is actually rich in networks and reciprocal

5 The motive and goal of joining is of vital importance to this end. For example, interviewees were generally not at a stage in their life where they were eagerly setting out to climb run for office or even lead a campaign.

336 relationships. For respondents, informal social contacts form an important part of their associational lives and act as a social lubricant, propelling single mothers in Canada, France and Sweden into other networks. While respondents were quite active in formal and informal associations alike, particularly in Canada and Sweden – interviews reveal interesting differences in the types of networks single mothers gravitate to – namely other parents in

Canada, and family and friends in France and Sweden. The findings also show that familial bonds, when not accompanied by a range of other types of informal and formal networks, are found to be linked with lower levels of trust and networking.

The interviews also demonstrate the value of child- and care-related networks, building trust and reciprocity, and even facilitating further networking and community ties.

They also indicate that, contrary to the psychological, orientation-based account for gendered participation suggests, care work and networks do not seem to orient mothers’ mental energies away from participation. Rather, findings from this study lend credence to the methodological account of the gender gap in participation, illustrating the importance of parent-, child- and care-related groups so often discounted by dominant civic engagement studies and surveys. These types of networks turn out to be a fundamental and indispensable component of the interviewees’ associational lives and, as such, should be accounted for. As such, future civic engagement research which strives to capture a more complete picture of the civil society would benefit from a including such network categories in civic engagement analyses.

While the personal interviews with single mothers serve as a good starting point towards better understanding the networking habits of women, future studies should include results with data from interviews conducted with non-single mothers, as well as men – to test

337 how the pressures and experiences of single mothers vary and shape their associational lives.

Furthermore, it is important that subsequent cross-national civic engagement research include more case studies representing each regime-type, and measure networking and trust over a longer period of time – to see if the present findings might be replicated. However, the scarcity of comparable, time-series data for the various civic engagement indicators will make this a difficult task to accomplish.

A key insight that interview findings contribute to the study of civic engagement is that women’s role as caregivers and paid workers leave them with limited resources needed to participate. However, respondents alluded to the role that the state played in shaping civic life.

The state can implement effective policies, for example, which help women reconcile the work-family tensions which constrains their civic involvement. This is reflected in the

Swedish 50-50 custody arrangements, as well as in the assortment of generous benefits (i.e. family, housing and maintenance) and services (i.e. leisure centres and daycares) available to

Swedish and French respondents to compensate them for their low stocks of resources, such as time and money. In fact, in the case of Canadian respondents, a lack of state presence best explains respondents’ heavy reliance on networks with other parents and the high levels of activities revolving around their children. Linking Welfare Regimes to Differences in Citizenship Status While men tend to fare better than women in terms of citizenship status and practice generally, when comparing women living in different welfare states, it becomes evident that some fare better than others. In some countries, states have done little to address traditional gender roles and inequality more broadly, while in others, rights and entitlements of the welfare state have been designed to alleviate disparity and social stratification. This is aptly

338 demonstrated in an analysis of family, income security and housing policies in Canada,

France and Sweden. This analysis not only brings to light how states can shape citizenship status in divergent ways, but suggests there is some order or logic to how this takes place.

Esping-Andersen’s regime typology posits that, as a result of historical legacies which have moulded the political landscape and the relationship between the state on the one hand, and the market, family and the community on the other, groups of states cluster together in terms of institutional logic and design as well as policy choices and practices. According to this approach, Anglo-Saxon countries, represented by the case of Canada, form the liberal welfare regime-type which favours private market solutions over state intervention into the market or family. This leads to a system which commodifies workers by providing little relief from the market through extensive social provisions, and stratifies the very poor from other citizens by making benefits conditional upon means-testing. Moreover, liberal regimes implicitly encourage traditional gender roles failing to provide support for women’s employment.

Western European countries form the corporate-conservative regime-type, represented by

France, which favour state intervention into the market and family with the aim to preserve the traditional social order. Corporate-conservative countries thus demonstrate a preference for social provisions which reproduce the old corporatist modes of social organization via occupationally-based insurance schemes, each having their own distinctive rights and privileges. While they are highly de-commodifying, offering a wide range of generous benefits, they stratify citizens according to occupational status, and between those with employment-based coverage and those without. Corporate-conservative regimes, although highly supportive of women’s employment, most notably through a highly sophisticated early childhood education system, also reinforce traditional gender roles by encouraging large

339 families and enticing women to remain outside of the labour market through familialist policies such as lengthy childrearing leaves and generous family benefits. Nordic countries form the social democratic regime-type, represented by the case of Sweden, which have a clear preference for state intervention into the market and family with the aim to promote equality. Policies are highly de-commodifying and promote solidarity by making the vast majority of benefits and services available on a universal basis. Generous, flexible benefits are in place which not only support women’s employment, but in some cases, actively encourage men to participate in caregiving.

As the social policies and women’s status reviewed in Chapter 4 suggest, social democratic Nordic regimes are the most favourable for women, actively working to discourage traditional gender roles by supporting female labour market participation while encouraging parenting by both parents. The results of such egalitarian policies are reflected in the low poverty rates for women in Nordic countries, including the narrow povery gap between single and non-single mothers. Corporate-conservative regimes also offer a great deal of support for women’s employment and also have relatively low poverty rates for women, but their familialist tendencies and occupationally-based insurance schemes maintain a system of stratification that encourages old status divisions and traditional gender roles which are not beneficial to women. Liberal regimes are found to be the least beneficial to women, with fewer supports to in place to help women reconcile paid work and family responsibilities; as a result, the poverty rate for women is highest in these countries, as is the poverty gap between single and non-single mothers. Thus, this analysis of comparative social policies demonstrates that the way in which states choose to approach traditional gender roles and inequality has a direct bearing on women’s citizenship.

340

The Impact of the State on Civic Life In light of link between the welfare state and citizenship status, and the relationship between citizenship status and participation, it follows that the state must also have an important effect on associational life, for instance, by developing policies which strive to reduce inequalities. The notion that the state shapes civic life is also supported by cross- national studies revealing great variation in levels of networking between countries. Some scholars have argued that cross-national civic engagement levels might best be explained by looking at the religious legacy of a regime, suggesting that Catholic countries tend to have lower levels of civic activity than Protestant ones. Protestantism is said to encourage voluntarism and philanthropy compared to the hierarchy and elitism characteristic of

Catholicism. However, while this study found some support for this theory, it nevertheless does not adequately explain cases such as corporate-conservative Germany or liberal Canada - with an important mix of both Catholics and Protestants, but with very different civic engagement outcomes. Nor does it make clear how a country that systematically worked to quash the Catholic Church’s power so long ago – such as France – would still be behaving like a Catholic country today. It is, moreover, unclear whether religion is the cause of disengagement, or whether countries that are not particularly active happen to also be historically Catholic. Finally, the prevalence of secularization, weakening the church’s influence over society, and shifting many social welfare responsibilities from the Church to the secular state, also casts doubt on the explanatory power of a theory which rests on the influence of confessional forces.

This study did not find any support for any of the prevailing country-level theories, tying cross-national civic engagement levels to the ethnic diversity, family ties and religious denomination. According to the ethnic diversity thesis, the more heterogeneous a society, the

341 lower the level of civic engagement; it proposes that racial and ethnic diversity breeds social isolation and distrust. This explanation may adequately describe levels of organizational involvement in highly active, homogeneous, social democratic societies, such as Sweden.

However, it is less compelling when turning to the case of corporate-conservative regimes such as France, consistently surpassed in civic engagement by the historically more heterogeneous societies of both the Canadian and American liberal regimes. Ultimately, this study finds no evidence in support of this theory.

This study also find no evidence for the theory linking family ties to cross-national civic engagement patterns. According to this perspective, civic engagement pattern may be explicated by looking at family ties, proposing that countries that have traditionally close kin networks, will have little need to participate in the civil society since they will do most of their networking and meet many of their needs through the family. However, the findings from a statistical analysis show that family ties have no effect on participation. Additionally, this explanation does not adequately account for the apparently wide variation across corporate-conservative regimes. Cases such as Germany and even France, for instance, are said to have relatively weak family ties when compared to corporate-conservative counterparts such as Italy, yet all have relatively low levels of memberships in formal networks. These prevailing accounts on the impact of religious legacy, the degree of heterogeneity in civil society and the strength of family ties on participation fail to account for the effect of power structures on associational life both at the individual-level, through factors such as gender, race and class, as well as at the country-level through the role of the state. For instance, the religious legacy theory overlooks how the Church’s influence in the area of welfare and the voluntary sector shifted to the secular state through the rise and development

342 of the welfare state, and how this reduces levels of disparity, raises living standards for a greater number of people, and developed alongside a burgeoning voluntary sector, as Skocpol noted in America and Worms in France. The diversity hypothesis fails to account for the way states structure associational life in their approach to inequality as well as racial and ethnic tensions through immigration and social welfare policies. The family ties theory does not account for the gender implications of belonging to tight-knit family ties for welfare and support. It also disregards the state’s effect on the family-state-market nexus by establishing social and family policies which either working to eliminate traditional gender roles as is done in social democratic regimes, or implicitly encouraging them by failing to address them as is done in liberal regimes, or explicitly reinforcing them as is done in corporate-conservative regimes.

Two different approaches theorizing the effect of the state on civic engagement have been reviewed. The first, which is found in the dominant civic engagement literature, argues that large interventionist states crowd-out the civil society and thus view the state as a hindrance to associational life. The second argues that depending on the state structure can work to encourage networking and trust, and thus view the state as a catalyst for civic engagement. Ultimately, however, the collection of studies on the role of the state in civic engagement are too limited to be considered a literature proper, and with scholars disagreeing on which aspect of the state to focus on, how it should be measured, or how it actually affects civic engagement, they are too varied to be considered a theory. However, based on this literature, three theories about the role of the state on civic engagement were developed, each focusing on different dimensions of the state, and each predicting different regime-based civic engagement outcomes. According to the first theory, civic engagement is strongest when

343 states refrain from intervening in the market and family. This view contends that strong, large, interventionist states crowd-out civil society by meeting the needs of citizens that would otherwise be met by combining with others. Conversely, the second theory suggests that the vitality of associational life does not depend on the degree of interventionism, but rather on the extent to which states strive to promote equality and reduce stratification between citizens.

States that fail to address inequalities will create an environment in which it is difficult to build the trust necessary to facilitate collective action. Lastly, another theory suggests that variation in levels of civic engagement is dependent upon the way political authority is structured. States which locate authority in society, legitimized by the people, will have more opportunities for citizens to engage and fewer boundaries between the state and society, leading to more trusting, active citizenry.

An analysis of the World Value Survey indicates that cross-national civic engagement variation undeniably does exist, and does so in a manner that suggests the answer remains rooted in welfare regime legacies. Social democratic regimes are highly engaged, leading in levels of trust, as well as in various forms of participation, including informal networking, membership levels, and political action. Liberal regimes are also very involved, often rivalling social democratic countries across several civic engagement indicators, and surpassing them in the area of civic activity. However, corporate-conservative countries are found to lag across all civic engagement indicators.

These results do not lend credence to the theory that the level of state intervention explains cross-national civic engagement levels because such logic fails to explain how the highly interventionist corporate-conservative and social democratic regimes lead to two radically different civic engagement outcomes. Cross-national civic engagement findings also

344 fail to provide support for the egalitarian-based theory of the role of the state because it cannot explain how the relatively egalitarian corporate-conservative regimes consistently score lower across civic engagement indicators than the highly selective, meagre and stratifying liberal welfare regimes. In fact, some corporate-conservative countries, such as

France, pose a particular dilemma for the egalitarian theory, since many of their social and family policies are deemed by some feminist scholars to be as generous and progressive as social democratic policies; yet they fail to exhibit any of the civic norms and behaviours associated with such entitlements, such as high levels of generalized trust and an actively engaged civil society. The need to better distinguish between corporate-conservative and liberal regime ‘selectivity’ and their ensuing associational outcomes is particularly intriguing when considering welfare debates and reforms in France and Canada. The former has witnessed a trend from contributions-based insurance schemes towards greater

‘universalization,’ while the latter has experienced retrenchment and devolution.1 In some ways, such reforms might be viewed as a continuation of original regime-based legacies, with corporate-conservative étatism working to increase the state’s control over social security2 while, conversely, liberal reforms increasingly shrinking the government’s role in social welfare. Ultimately however, arguments centered on the selectivity of welfare institutions alone are unable to explain the pattern of cross-national associational involvement uncovered here.

1 French scholars have documented this shift in the financing of social security, reflected in the implementation of no less than nine basic income schemes established since the 1970s, as well as the levy of income tax schemes (quite apart from, and in addition to, social insurance contributions) such as the Contribution sociale généralisée (CSG) of 1990, and the Contribution pour le remboursement de la dette sociale (CRDS) of 1996 (Friot, 1999; Letard, 2005; Centre d'analyse stratégique, 2006; Monnier, 2006; Dider and Martinez, 2006). 2 This, albeit while drifting away from insurance-based schemes, a hallmark of the corporate-conservative regime-type.

345

Clearly, what is required is an explanation for comparative patterns of associational involvement that cuts across overly-simplistic state size and strength arguments, is not founded on single-factor answers such as religious denomination or ethnic diversity, and looks beyond specific policy areas or the selectivity of programs and benefits. Based on the results from descriptive and inferential statistical analyses, the political authority model best explains how the highly interventionist and egalitarian social democratic regime-type, and the polar opposite least interventionist and egalitarian liberal regime-type, could both be vibrant, active societies. According to this approach, patterns of civic engagement can be understood as the result of the regime-based political and cultural legacies shaping civic norms and behaviours. Even though his analysis is conducted at the sub-national level, Putnam’s Italian analysis supports this hypothesis. When looking at the political-sociological traditions of the civically active North and the disengaged South, he discovers two very distinctive patterns of governance, linked to divergent social and cultural regional features, dating back to the fourteenth century:

… [the] Norman feudal autocracy of the South and the fertile communal republicanism of the North… In the North, feudal bonds of personal dependence were weakened; in the South, they were strengthened. In the North, the people were citizens; in the south, they were subjects… Legitimate authority in the North was ‘only delegated [by the community] to public officials who remain responsible to those with whose affairs they are entrusted.’ Legitimate authority in the South was monopolized by the king, who… was responsible only to God (Putnam, 1993, p.130).

According to the statist theory developed in this study, which draws on the work of

Jepperson, a distinction can be made at the level of state organization between statist and non- statist systems of authority. In statist countries, which generally include corporate- conservative regimes, authority is concentrated within a highly unified state structure, and voluntary activity encouraged by the state and which has often historically been constrained.

346

The state is understood as having the duty and ability to direct the public realm and societal activity – in the name of promoting a well-defined national interest. Power is hierarchically structured in statist countries, and there are clear demarcations between state as decision- making, transformative and administrative centre and the masses. Conversely, in non-statist countries, including liberal and social democratic regimes, authority and the legitimacy of state power rests in the civil society. Since it is understood to be an instrument of society, the state is guided by pressures from various interests in society who work to shape policy, whether via pluralism in liberal regimes, or the bargaining process of corporatism in social democratic countries. Additionally, in non-statist countries, boundaries between the state and society are less defined, and “[s]tate structures are envisioned as coordinating and administering, rather than directing” (Jepperson, 2002, p.74). Based on this model, it is possible to link low levels of statism with greater voluntary activity. Thus, trends in associational activity seem to indicate that states and societies are the product of the socio- political legacies of regimes, and embedded in cultural and institutional logics. Such an approach lends itself well to regime-theory since it too is premised on the notion that values and behaviours are the outcome of the complex historical trajectories specific to certain nations.3

Ultimately, civic engagement requires participation in dense, horizontal, formal and informal networks that trust and reciprocity are produced. These components, working in concert, promise to provide active communities with positive returns, including enhanced

3 For instance, Esping-Andersen states: “We know already that a variety of historical forces has played a decisive part in the formation of welfare states; this is almost certainly why they differ, yet cluster as they do. The extraordinary power of the Church, the aristocracy, and an authoritarian state in early continental European capitalism is certainly a key to the explanation for this regime-type; in converse, the weakness of absolutism and the dominance of laissez-faire-inspired in the Anglo-Saxon nations goes a long way in accounting for the ‘liberal’ regime. In both cases, social policy was never neutral, but part of a more general campaign to weaken, or absorb, the socialist impulse and to secure a lasting institutionalization of politically preferable principles of social organization” (Esping-Andersen, 1990, p.110).

347 democracy, reflected in institutions that are accountable, effective and responsive to citizens’ interests and needs. As such, civic engagement can be said to move from the bottom-up – from the citizen to the state, as the dominant literature suggests. However, this only describes part of the whole story, for what is just as necessary is the presence of a political structure in which the state – whether strong or weak, large or small – has historically drawn its authority and legitimacy from the civil society. As the preceding chapter highlighted, civic engagement will be thwarted in a political system where power has largely been confined to insular groups of experts or elite-driven institutions, as is the case in many corporate-conservative regimes.

Rather, as the statist theory suggests, it thrives where there are weak boundaries between the state and society, providing citizens willing to become actively engaged with an opportunity for exchanges with one another and with the state to influence decision-making, as exemplified by the liberal and social democratic regimes (Janoski, 1998, p.17-24, p.108-112;

Jepperson, 2002, Op.cit).

Yet, the way political authority is structured in society alone does not fully explain how associational life is shaped. Social and economic factors matter, this is evident not only in the individual-level analyses in Part I, but also in the results of the hierarchical linear model in Chapter 6, controlling for country-level factors. Indeed, when including both inequality and regime-type indicators in the analysis concurrently, we find that egalitarian Nordic countries are so effective at reducing inequalities, that the effects of social and economic conditions, which typically pose a challenge to civic activity, are significantly tempered in these countries. This suggests that it is not only important to pay attention to how power is structured and distributed in a political system, and the need to create opportunities for

348 citizens to participate, but that creating policies which strive to reduce inequalities and social stratification, as is the case in social democratic countries, is beneficial.

Since the political authority theory places such great importance on the historical legacies of regimes, it is important to explore the statist/societal implications of a country such as Canada, with a heritage that incorporates a mix of both the British liberal, non-statist tradition, as well as French statism.4 Moreover, a compelling question for future cross- national research revolves around the extent to which regimes and their civic engagement outcomes are the result of self-reinforcing, or perhaps even path dependent processes

(Pierson, 2000; Sabetti, 1996). This is exemplified in Putnam’s conceptualization of societies as either being locked into virtuous or vicious circles, as well as his tracing the civic life of

Northern and Southern Italian regions to the their historical legacies – all the way back to the twelfth century (Putnam, 1993, p.121-162). Such tendencies also prevail in regime-theory and

Jepperson’s approach, both resting their explanations on long-term and enduring political and cultural trajectories (Esping-Andersen, 1990, p.110).

Yet, if this is the case, is there any hope for a corporate-conservative country to ever reverse its vicious cycle by increasing its level of civic activity to the extent that it join the ranks of another regime? It seems just as unrealistic to expect the French state to relinquish authority and seek legitimacy from the civil society, as it is to expect the French civil society to adopt the networking behaviour and norms of trust of Swedes or Canadians. However, because explanations grounded in historical legacy, such as path dependency, hold a static view of the world, or preclude the possibility of change, they can hardly be expected to adequately forecast the future (Pierson, 2000, p.265). As such, perhaps the greatest challenge

4 Jepperson suggests that Canada can be seen as a mixed model – a combination of the liberal polity in English Canada, and statism in French Canada (Jepperson, 2000, p.28).

349 facing future cross-national civic engagement research is not to chart continuity by detailing the distinctive features of regimes and their governing structures, but to uncover what conditions give rise to changes within and between them.

Overall, these findings have implications for how states might work to foster civic engagement through institutional design and policy practices. The impact of social and economic context and resources on participation uncovered at the individual-level, coupled with some important cross-national level findings regarding the effect of egalitarian welfare regimes, suggest that social policies which promote greater equality between citizens are key to an active civil society. The case of social democratic countries can be seen as an example of success in this respect, where the state has actively worked to weaken traditional gender roles and other forms of social stratification through the rights and entitlements of the welfare state; in so doing, they have built more egalitarian, socially cohesive, vibrant civil societies.

Designing and implementing policies geared at reducing inequalities can only help citizens gain the capacity and ability to become engaged. However, promoting equality is not the only ingredient required to foster civic engagement. Indeed, merely having an egalitarian welfare state, with social provisions geared to support women balancing labour market activity and family life, does not on its own guarantee a highly active and trusting civil society, as the case of France demonstrates. Despite its wide range of generous benefits, which led some feminist scholars to equate it with social democratic countries, France continues to exhibit low levels of trust and networking. What the findings of this study suggest is that, in addition to promoting equality, the way political authority has historically been structured has a deep and lasting impact on the civic profile of societies. Despite having comparatively generous

350 benefits, statists countries have traditionally located and legitimized political authority in the central state, apart from society, leading to apathy and distrust among the population. The lesson which can be learned from this is that in addition to developing egalitarian policies, it is imperative that states create openings and opportunities for citizens to participate - to inspire collective action.

351

References

Adam Frane and Borut Rončević, 2003. “Social Capital: Recent Debates and Research Trends,” in Social Science Information, Vol.42, No.2.

Adema Willem, 2006. Social Assistance Policy Development and the Provision of a Decent Level of Income in Selected OECD Countries, OECD Directorate for Employment, Labour and Social Affairs, Social, Employment and Migration Working Papers #38, Paris, August.

Adema Willem, 2000. Revisiting Real Social Spending Across Countries: A Brief Note, OECD Economic Study No.30, OECD Publishing.

Adema Willem and Maxime Ladaique, 2009. How Expensive is the Welfare State? Gross and Net Indicators in the OECD Social Expenditure Database (SOCX), OECD Social, Employment and Migration Working Papers No.92, OECD Publishing.

Adkins Lisa, 2005. “Social Capital: The Anatomy of a Troubled Concept,” in Feminist Theory, Vol.6, No.2.

Aizlewood Amanda and Ravi Pendakur, 2005. “Ethnicity and Social Capital in Canada,” in Canadian Ethnic Studies Journal, Vol.32, No.2.

Alesina Alberto, Eliana La Ferrera, 2002. “Who Trusts Others?” in Journal of Public Economics, Vol.85.

Alesina Alberto and Paola Giuliano, 2009. Family Ties and Political Participation, Institute for the Study of Labor, Discussion Paper No.4150, Bonn, April.

Alesina Alberto and Paola Giuliano, 2007. Power of the Family, Institute for the Study of Labor, Discussion Paper No.2750, Bonn, April.

Alesina Alberto, Reza Baqir and William Easterly, 1997. Public Goods and Ethnic Divisions, Working Paper 6009, National Bureau of Economic Research.

Allen Ann Taylor, 2005. “From Motherhood to Sex Roles: The Postwar Era, 1945-1970,” in and Motherhood in Western Europe, 1890-1970: The Maternal Dilemma, Palgrave MacMillan, New York.

Almond Gabriel A. and Sidney Verba, 1965. The Civic Culture: Political Attitudes and Democracy in Five Nations, Little Brown and Company, Boston.

Allport Gordon, 1954. The Nature of Prejudice, Addison-Wesley, Cambridge.

Alvestegui Gonzalo and Jean-Michel Charbonnel, 1999. “Mise en oeuvre du RMI et reformulation de la question sociale,” in Lien Social et Politiques – Revue international d’action communautaire (RIAC), Vol.42, Fall/

352

Arneil Barbara, 2006. Diverse Communities: The Problem With Social Capital, Cambridge University Press, Cambridge.

Amott Teresa L., 1990. “Black Women and AFDC: Making Entitlement Out of Necessity,” in Women, the State and Welfare, Linda Gordon (Ed.), University of Wisconsin Press, Madison.

Anderson Christopher J. and Aida Paskeviciute, 2006. “How Ethnic and Linguistic Heterogeneity Influence the Prospects for Civil Society: A Comparative Study of Citizenship Behaviour,” in Journal of Politics, Vol.68, No.4, November.

Anxo Dominique, 2002. “Time Allocation and the Gender Division of Labour in France and Sweden,” in The Future of Work, Employment and Social Protect: The Dynamics of Change and the protection of Workers, P. Auer, B. Gazier eds., ILO, Geneva.

Andersen G., 1993. “Support and Relief: The Swedish Contact Person and Contact Family Program,” in International Journal of Social Welfare, Vol.2, Issue 1, April.

Archambault Édith, 2001. “Historical Roots of the Nonprofit Sector in France,” in Nonprofit and Voluntary Sector Quarterly, Vol.30, No.2, June.

Archambault Édith, 1996. Le secteur sans but lucrative en France et dans le monde, Economica, Paris.

Arts Wil and John Gelissen, 2002. “Three Worlds of Welfare Capitalism or More? A State-of- the-art Report,” in Journal of European Social Policy, Vol.12, No.2.

Aubert Fabrice and Anne-Marie Fribourg, 2005. Le finacement du logement, Republique Française, Direction générale de l’Urbanism de l’Habitat et de la Construction, Paris.

Aucoin Kathy, 2005. Canadian Centre for Justice Statistics, Family Violence in Canada: A Statistical Profile 2005, Statistic Canada, Catalogue No.85-224-XIE, Ottawa.

Avdeyeva Olga, 2006. “In Support of Mothers' Employment: Limits to Policy Convergence in the EU?” in International Journal of Social Welfare, Vol.14.

Bakan Abigail B. and Audrey Kobayashi 2000. Employment Equity Policy in Canada: An Interprovincial Comparison, Status of Women Canada, Ottawa, March.

Bambra Clare, 2004. “The Worlds of Welfare: Illusory and Gender Blind?” in Social Policy and Society, Vol.3, No.3.

Banfield Edward C., 1958. The Moral Basis of a Backward Society, The Free Press, Glencoe.

Bang Henrik P., 2005. “Among Everyday Makers and Expert Citizens,” in Remaking Governance: Peoples, Politics, and the Public Sphere, Janet Newman.

353

Bang Henrik P., Eva Sørensen, 1998. The Everyday Maker: A New Challenge to Democratic Governance, ECPR Paper prepared for the 26th Joint Sessions of Workshops, University of Warwick, March.

Banti Alberto, 2000. “Public Opinion and Associations in Nineteeth-Century Itlay,” in Civil Society Before Democracy: Lessons From Nineteenth-Century Europe, Nancy Gina Bermeo and Philip G. Nord (Eds.), Rowman & Littlefield Publishers Inc.

Banting Keith G., 2005. “The Multicultural Welfare State: International Experience and North American Narratives,” in Social Policy and Administration, Vol.39, No.2, April.

Banting Keith G.,, Richard Johnston, Will Kymlicka and Stuart Soroka, 2006. “Do multiculturalism Policies Erode the Welfare State? An Empirical Analysis,” in Multiculturalism and the Welfare State: Recognition and Redistribution in Contemporary Democracies, Keith Banting and Will Kymlicka (Eds.), Oxford University Press, Oxford.

Barbetta Gian Paolo, 1997. The Nonprofit Sector in Italy, John Hopkins Comparative Nonprofit Sector Project, Manchester University Press, Manchester.

Beauvais Caroline and Jane Jenson, 2001. Two Policy Paradigms: Family Responsibility and Investing in Children, Canadian Policy Research Networks, Discussion Paper #F12, Ottawa, February.

Béland Daniel and Randall Hansen, 2000. “Reforming the French Welfare State: Solidarity, Social Exclusion and the Three Crises of Citizenship,” in West European Politics, Vol.23, No.1, January.

Bengtsson Bo, 2001. “Housing as a Social Right: Implications for Welfare State Theory,” in Scandinavian Political Studies, Vol.24, No.4.

Bergh Andreas, 2006. “Is the Swedish Welfare State a Free Lunch?” in Econ Journal Watch, Vol.3, No.2, May.

Bermeo Nancy and Philip Nord, 2000. Civil Society Before Democracy: Lessons From Nineteenth-Century Europe, Rowman and Littlefield Publishers Inc., New York.

Beveridge William, 1942. Report of the Inter-Departmental Committee on Social Insurance and Allied Services (The Beveridge Report), Presented to Parliament in November, H.M. Stationary Office, London.

Bezanson Kate, 2006. “Gender and the Limits of Social Capital,” in The Canadian Review of Sociology and Anthropology, Vol.43, No.4.

Birnbaum Pierre, Marie-Hélène Adler, and Franklin Hugh Adler, 1982. “The State Versus Corporatism,” in Politics & Society, Vol.11, No.4.

354

Blalock Jr. Hubert M., 1967. Towards a Theory of Minority-group Relations, John Wiley and Sons, New York.

Blanc Maurice, 2004. “The Changing Role of the State in French Housing Policies: A Roll- out Without Roll-back?” in European Journal of Housing Policy, Vol.4, No.3, December.

Bock Gisela, 2002. Women in European History, Blackwell Publishers Ltd., Oxford.

Blom J.C.H. and Emiel Lamberts, 2006. “Unity and Diversity in the Low Countries,” in History of the Low Countries, J. C. H. Blom & E. Lamberts (Eds.), Berghahn Books, New York.

Blumer Herbert, 1958. “Race Prejudice as a Sense of Group Position,” in Pacific Sociological Review, Vol.1, No.1, Spring.

Bobo Lawrence D., 1999. “Prejudice as Group Position: Microfoundations of a Sociological Approach to Racism and Race Relations,” in Journal of Social Issues, Vol.55, No.3.

Bobo Lawrence D. and Mia Tuan, 2006. Prejudice in Politics: Group Position, Public Opinion and the Wisconsin Treaty Rights Dispute, Cambridge, Harvard University Press.

Bock Gisela and Pat Thane, 1994. Maternity and Gender Policies: Women and the Rise of the European Welfare States 1880s-1950s, Routledge, New York.

Boryczka Jocelyn M., 2009. “The Separate Spheres Paradox: Habitual Inattention and Democratic Citizenship,” in Feminist Interpretations of (Re-Reading the Canon), Jill Locke, Eileen Hunt Botting (Eds.), Pennsylvania State University, University Park.

Bourdieu Pierre, 2001. Masculine Domination, Stanford University Press, Stanford.

Bourdieu Pierre, 1986. “The Forms of Capital,” in Handbook of Theory and Research in Sociology of Education, John G Richardson eds., Greenwald Press, New York.

Bourdieu Pierre, 1984. Distinction: A Social Critique of the Judgement of Taste, Harvard University Press, Cambridge.

Boychuck Gerard William, 1998. Patchworks of Purpose: The Development of Provincial Social Assistance Regimes in Canada, McGill-Queen's Press, Montreal.

Brehm John and Wendy Rahn, 1997. “Individual-Level Evidence for the Causes and Consequences of Social Capital,” in American Journal of Political Science, Vol.21, No.3, July.

Brodie Janine and Isabella Bakker, 2007. Canada's Social Policy Regime and Women: An Assessment of the Last Decade, Status of Women Canada, Canadian Cataloguing in Publication Data, Ottawa, March.

355

Bruegel Irene, 2005. “Social Capital and Feminist Critique,” in Women and Social Capital, Jane Franklin ed., Families and Social Capital ESRC Research Group, London South Bank University Press, London, April.

Brugi Noëlle, 2006. “The Laws of an Unravelled Social Order: France's Undeserving Poor,” in Le Monde Diplomatique, November.

Brush Lisa D., 2002. “Changing the Subject: Gender and Welfare Regime Studies,” in Social Politics, Summer.

Burns Nancy, Kay Lehman Scholzman and Sidney Verba, 2001. The Private Roots: Gender, Equality, and Political Participation, Harvard University Press, Cambridge.

Burri Susanne, 2003. L’expérience des Pays-Bas ; du congé parental à l’adaptation du temps de travail, Sythesis of presentation for the Conference: Emploi des femmes et charges familiales: Repenser le congé parental en France à la lumière des expériences étrangères, Centre de recherche en économie de Sciences Politiques, December 3rd.

Bussemaker Jet and Rian Voet, 1998. “Citizenship and Gender: Theoretical Approaches and Historical Legacies,” in Critical Social Policy, Vol.18, No.56.

Caiazza Amy and Barbara Gault, 2006. “Acting From the Heart: Values, Social Capital and Women’s Involvement in Interaith and Environmental Organizations,” in Gender and Social Capital, Brenda O’Neil and Elisabeth Gidengil (Eds.), Routledge, New York.

Caiazza Amy and Robert Putnam, 2002. Women’s Status and Social Capital Across the States, Institute for Women’s Policy Research, Briefing Paper #I911, July.

Caisse Nationale des Allocations Familialies (CNAF),2008. L’E-ssentiel: Publication de la CNAF, No.69, January.

Canadian Centre for Justice Statistics, 2003. Transitions Homes in Canada: National, Provincial and Territorial Fact Sheets 2001-2002, Statistics Canada, Catalogue No.85-404- XIE, Ottawa, June.

Canada Revenue Agency, 2009. Child Care Expenses, T778 E (09) Form, Government of Canada.

Castles Francis G., 1975. “Swedish Social Democracy: The Conditions of Success,” in The Political Quarterly, Vol.46, No.2, April.

Centre d'analyse stratégique, 2006. La réforme du financement de la protection sociale, Avis de synthèse, No.1, Septembre.

Christopher Karen, 2002. “Welfare States Regimes ad Mother’s Poverty,” in Social Politics, Vol.9, No.1, Spring.

356

Christopher Karen, 2001. Single Motherhood, Employment, or Social Assistance: Why are U.S. Women Poorer than Women in Other Affluent Nations?, Luxembourg Income Study, Working Paper No.285, November.

Christopher Karen, Paula England, Timonty M. Smeeding and Katherin Ross Phillips, 2002. “The Gender Gap in Poverty in Modern Nationa: Single Motherhood, the Market, and the State,” in Sociological Perspectives, Vol.45, No.3.

Christopherson Susan, 2001. “Women in the Restructuring of Care Work: Crossnational Variations and Trends in Ten OECD Countries,” in Women and Welfare: Theory and Practice in the United States and Europe, Nancy J. Hirschmann and Ulrike Liebert (Eds.), Rutgers University Press, New Jersey.

Clearinghouse on International Developments in Child, 2005. Youth and Family Policies, Columbia University, http://www.childpolicyintl.org/, (last viewed October 2010).

CMK foundation of childcare and the Ministry of Social Affairs and Employment in the Netherlands, 2004. Childcare in a Changing World, Country Files: France, 21-23 October, Groningen, the Netherlands.

Coalition of Child Care Advocates of BC, 2004. Time Line of Childcare in BC, http://www.cccabc.bc.ca/res/timeline.html, last visited in October 2010.

Collins Jock, Greg Noble, Scott Poynting and Paul Tabar, 2000. Kebabs, Kids, Cops and Crime: Youth, Ethnicity and Crime, Pluto Press.

Conseil Superieur de L'Emploi Des Reveus et des Couts, 1997. Minima Sociaux: Entre Protection et Insertion, La Documentation Francaise, Paris.

Conway Margaret M., 1991. Political Participation in the United States, 3rd edition, CQ Press, Washington.

Coleman James S., 1998. “Social Capital in the Creation of Human Capital,” in American Journal of Sociology, Vol. 94, Supplement: Organizations and Institutions.

Coleman James S., 1994. Foundations of Social Theory, Harvard University Press, Cambridge.

Cool Julie, 2004. Child Care in Canada: Regulated, Unregulated, Private or Public, Government of Canada, Political and Social Affairs Division, June 3rd.

Cordazzo Phillip, 2005. “Minimum Income Allocation System (RMI): A Longitudinal View,” in Evaluation Review, Vol.29, No.5.

Cortina Jose M., 1993. “What is Coefficient Alpha? An Examination of Theory and Application,” in Journal of Applied Psychology, Vol.78, No.1.

357

Costa Dora L. and Matthew E. Kahn, 2003. “Civic Engagement and Community Heterogeneity: An Economist’s Perspective, in Perspectives on Politics, Vol.1,No.1.

Crompton Rosemary and Nicky Le Feuvre, 2000. “Gender, Family and Employment in Comparative Perspective: The Realities and Representations of Equal Opportunities in Britain and France,” in Journal of European Social Policy, Vol.10, No. 4.

Cruishank Barbara, 1999. The Will to Empower: Democratic Citizens and Other Subjects, Cornell University Press, Ithica.

Curtis James E., Douglas E. Baer and Edward G. Grabb, 2001. “Nations of Joiners: Explaining Voluntary Association Membership in Democratic Societies,” in American Sociological Review, Vol.66, No.6, December.

Curtis James E., Edward G. Grabb and Douglas E. Baer, 1992. “Voluntary Association Membership in Fifteen Countries: A Comparative Analysis,” in American Sociological Review, Vol.57, No.2, April.

De Léséleuc Sylvain and Rebecca Kong, 2004. Canadian Centre for Justice Statistics, Victim Services in Canada: National, Provincial and Territorial Fact Sheets 2002-2003, Statistics Canada, Catalogue No.85-003-XIE, Ottawa, December.

DeFilippis James, 2002. “Symposium on Social Capital: An Introduction,” in Antipode Vol.34, No.4, September.

DeFilippis, James, 2001. “The Myth of Social Capital in Community Development,” in Housing Policy Debate, Vol.12, No.4.

De Souza Briggs Xavier, 2001. Social Capital and Segregation : Race, Connections, and inequality in America, paper presented at The Association of Collegiate Schools of Planning annual meeting, Cleveland, Ohio, November.

Daly Mary, 1994. “Comparing Welfare States: Towards a Gender Friendly Approach,” in Gendering Welfare States, Diane Sainsbury, Sage, London.

Dang Ai-Thu and Jean-Marie Monnier, 2004. “The French System of Direct and Indirect Family Transfers and the Principles of Justice,” in Journal of Economic Issues, Vol. 38, No. 3, September.

Das Raju J., 2004. “Social Capital and Poverty of the Wage-Labour Class: Problems with the Social Capital Theory,” in Transactions of the Institute of British Geographers, Volume 29, Issue 1, March.

Delli Carpini Michael X. and Keeter Steve, 1996. What Americans Know About Politics and Why it Matters, New Haven, Yale University Press.

358

Demailly Dominique, Jacques Bouchoux and Jean-Luc Outin, 2002. “Les contrats d’insertion du RMI: pratiques des institutions et perceptions des beneficiaries,” in Etudes et Resultats, DREES, No.193, September.

Department of Trade and Industry, 2006. UK, International Review of Leave Policies and Related Research 2006, Employment Relations Research Series, No.57.

Dider Michel, Michel Martinez, 2006. Contributions au Conseil d’Analyse Economique sur le basculement des cotisations patronales, Centre de Recherches pour l’Expansion de l’Economie et le Développement des Entreprises, Paris, juin.

Ditch John, Alan Lewis and Steve Wilcox, 2001. Social Housing, Tenure and Housing Allowance: An International Review, Government of the United Kingdom: Department for Work and Pensions, In-House Report #83, Her Majesty’s Stationary Office, London.

Doherty Gillian, Martha Friendly and Jane Beach, 2003. OECD Thematic Review of Early Childhood Education and Care in Canada, Government of Canada.

Dolan Kathleen, 2011. “Do Women and Men Know Different Things? Measuring Gender Differences in Political Knowledge,” in Journal of Politics, Vol. 73, No.1, January.

Edgar Bill and Joe Doherty, 2001. Women and Homelessness in Europe: Pathways, Services and Experiences, Bill Edgar, Joe Doherty ed., Policy Press, University of Bristol, Bristol.

Edwards Rosalind, Jane Franklin and Janet Holland, 2003. Families and Social Capital: Exploring the Issues, Families and Social Capital ESRC Research Group, South Bank University, London, February.

Enjolras Bernard, Jean Louis Laville, Laurent Fraisse and Heather Trickey, 2001. “Between Subsidiarity and Social Assistance: The French Republican Route to Activation,” in An Offer You Can’t Refuse: Workfare in International Perspective, Policy Press, University of Bristol, Bristol.

Erickson Bonnie, 2004. “Distribution of Gendered Social Capital in Canada,” in Creation and Returns of Social Capital: A New Research Program, Henk Flap and Beate Volker (Eds.), Routledge, New York.

Erickson Bonnie and T.A. Nosanchuck, 1990. “How Apolitical Associations Politicize,” in Canadian Review of Sociology and Anthropology, Vol.27.

Ertman Thomas, 2000. “Liberalization, Democratization, and the Origins of a ‘Pillarized’ Civil Society in Nineteenth-Century Belgium and the Netherlands,” in Civil Society Before Democracy: Lessons From Nineteenth-Century Europe, N. Bermeo & P. Nord (Eds.), Rowan and Littlefield Publishers, Lanham.

Esping-Andersen Gøsta, 1999 Social Foundations of Postindustrial Economies, Oxford University Press, Oxford.

359

Esping-Andersen Gøsta, 1990. The Three Worlds of Welfare Capitalism, Princeton University Press, Princeton.

European Foundation for the Improvement of Living and Working Conditions, 2002. Gender Pay Equity in Europe, European Industrial Relations Observatory On-Line, http://www.eurofound.europa.eu/eiro/2002/01/study/tn0201101s.htm (last viewed October 2010).

Eyerman Ron and Andrew Jamison, 1991. Social Movements: A Cognitive Approach, Pennsylvania State University Press, University Park.

Fabre Virginie and Olivie Sautory, 2009. Enquête sur les experimentations du RSA: Premier Resultats, Directionde la recherché, des etudes, de l’évaluation et des statistiques (DREES), Ministère du Travail, des Relations Sociales, de la Famille, de la Solidarité et de la ville, Ministère de la Santé et des Sports, et le Ministère du Budget, des Comptes Publics et de la Fonction Publique, Serie Études et Recherches, Document de Travail No.87, April.

Fagnani Jeanne, 2005. Family Policy in France: Old Challenges, New Tensions, Report No.2, CESifo - Center for Economic Studies (CES), the Ifo Institute for Economic Research, Munich.

Faulks Keith, 2000. Citizenship, Routledge, New York.

Faulks Keith, 2000. Political Sociology: A Critical Introduction, New York University Press, New York.

Field Andy, 2009. Discovering Statistics Using SPSS, 3rd edition, Sage, London.

Field John, 2008. Social Capital: Key Ideas, Routledge, New York.

Field John, 2005. Social Capital and Lifelong Learning, Policy Press, University of Bristol, Bristol.

Fieldhouse Edward and David Cutts, 2010. “Does Diversity Damage Social Capital? A Comparative Study of Neighbourhood Diversity and Social Capital in the US and Britain,” in Canadian Journal of Political Science, Vol.43, No.2, June.

Fine Ben, 2006. The New Development Economics: After the Washington Consensus, Jomo K.S. and Ben Fine (Eds.), Palgrave, New York.

Fine Ben, 2001. Social Capital Versus Social Theory: Political Economy and Social Science at the Turn of the Millennium, Routledge, New York.

Fine Ben, 1999. “The Developmental State is Dead – Long Live Social Capita?” in Development and Change, Vol.30, No.1.

360

Fitzpatrick Suzanne and Mark Stephens, 2007. An International Review of Homelessness and Social Housing Policy, Centre for Housing Policy, University of York, Department for Communities and Local Government, Queen’s Printer and Controller of Her Majesty’s Stationary Office, London, November.

Foley Michael W., and Bob Edwards, 1999. “Is It Time to Disinvest in Social Capital?” in Journal of Public Policy, Vol.19, No.2.

Forbes Hugh Donald, 1997. Ethnic Conflict, Culture and the Contact Hypothesis, Yale University Press, London.

Fraisse Genevieve, 1995. Muse de la Raison: La Democratie exclusive et la difference des sexes, Paris, Gallimard.

Fraser Nancy, 1994. “After the Family Wage: Gender Equity and the Welfare State,” in Political Theory, Vol.22, No.4.

Friendly Martha, Jane Beach and Michelle Turiano, 2002.“Early Childhood Education and Care in Canada 2001,” Childcare Resource and Research Unit, University of Toronto, December.

Friot Bernard, 1999. “Ciblage, universalité, discrimination: la fiscalité redistributive contre le salaire,” in Liens social et Politiques, Vol.42, Fall.

Fuchs Dieter, 2007. “The Political Culture Paradigm,” in The Oxford Handbook of Political Behaviour, Russell J. Dalton, Hans-Dieter Klingemann (Eds.), Oxford University Press, Oxford.

Fukuyama Francis, 2000. Social Capital and Civil Society, IMF Working Paper WP/00/74, IMF Institute, April.

Fukuyama Francis, 1999. The Great Disruption: Human Nature and the Reconstitution of Social Order, The Free Press, New York.

Fukuyama Francis, 1995. Trust: The Social Virtues and the Creation of Prosperity, Free Press Paperbacks, New York.

García Bedolla, Lisa,. and Becki Scola, 2006. “Finding Intersection: Race, Class, and Gender in the California Recall Vote,” in Politics and Gender, Vol. 2, No.1.

Gauthier Anne H., 2002. “Family Policies in Industrialized Countries: Is There Convergence?” in Population, Vol.57, No.3, May-June 2002, p.453. Karen Christopher, “Welfare States Regimes ad Mother’s Poverty,” in Social Politics, Vol.9, No.1, Spring.

Gibson James L., 2001. “Social Networks, Civil Society, and the Prospects for Consolidating Russia’s Democratic Transition,” in American Journal of Political Science, Vol.45, No.1, January.

361

Gidengil Elizabeth, Elizabeth Goodyear-Grant, Neil Nevitte, André Blais and Richard Nadeau, 2006. “Gender, Knowledge and Social Capital” in Gender and Social Capital, Brenda O’Neill and Elisabeth Gidengil, Routledge, New York.

Giles Michael W. and Arthur Evans, 1986. “The Power Approach to Intergroup Hostility,” in Journal of Conflict Resolution, Vol.30, No.3, September.

Ginsburg Norman, 2002. Sweden: The Social Democratic Case,” in Comparing Welfare States, Allan Chcrane, John Clarke and Sharon Gewirtz (Eds.), Sage, London.

Girard Philip, 1994 “Why Canada has No Family Policy: Lessons from France and Italy,” in Osgoode Hall Law Journal, Vol.32, No.3.

Glaeser Edward L., David L. Laibson, José A. Scheinkman and Christine L. Soutter, 2000. “Measuring Trust,” in The Quarterly Journal of Economics, Vol.115, No.3, August.

Goetz, Anne Marie, 2008. Who Answers to Women? Gender and Accountability, United Nations Fund for Women, Progress of the World’s Women 2008/2009.

Gorlick Carolyne and Guy Brethour, 1998. Welfare to Work Programs in Canada: A National Inventory, Canadian Council on Social Development, Ottawa, Ontario..

Gornick Janet C. and Marcia K, Meyer, 2005. Families That Work: Policies for Reconciling Parenthood and Employment, Russell Sage, New York.

Gornick Janet C. and Markus Jäntti, 2010. Women, Poverty and Social Policy Regimes: A Cross-National Analysis, Luxembourg Income Study, Working Paper No.534, April.

Government of Canada, Women’s Right to Vote in Canada, http://www2.parl.gc.ca/Parlinfo/compilations/ProvinceTerritory/ProvincialWomenRightToVo te.aspx?Menu=ElectionsRidings-Election (last visited October 2010).

Government of Finland, 2006. Ministry of the Environment, Right to Housing in Europe – Need for a Comprehensive Strategy, prepared for the Housing Rights in Europe Conference in Helsinki, Peter Fredriksson, Juho Pätäri eds., Helsinki, September.

Government of Nova Scotia, 2003. Final Report of the Ad-hoc Federal-Provincial- Territorial Working Group Reviewing Spousal Abuse Policies and Legislation, Department of Justice Canada, Nova Scotia Department of Justice co-chairs.

Government of Quebec, 2002. An Act to Facilitate the Payment of Support, R.S.Q. P-2.2.

Graham John R. and Alean Al-Krenawi 2001. “Canadian Approaches to Income Security,” in Canadian Social Welfare, Joanne C. Turner, Francis J. Turner eds., Pearson Education Canada Inc., Toronto.

362

Granovetter Mark S., 1973. “The Strength of Weak Ties,” in The American Journal of Sociology, Vol.78, No.6, May.

Green G.P., Tigges, L.M. and Brown I., 1995. “Social Resources, Job Search and Poverty in Atlanta,” Research in Community Sociology, Vol.5.

Guest Dennis, 1999. The Emergence of Social Security in Canada, UBC Press, Vancouver.

Gunnarsson Lars, Barbara Martin Korpi and Ulla Nordenstam, 1999. Early Childhood Education and Care Policy in Sweden, Ministry of Education and Science, Regeringskansliet, Government of Sweden, Background Report for the OECD’s Thematic Review of Early Childhood Education and Care, Stockholm, December.

Gustafsson Björn, 2004. “Social Assistance in Sweden,” in International Experience with Social Assistance Schemes: Five Country Case Studies, Internastional Labour Organization, ILO Cataloguing in Publication Data, Moscow.

Gurstein Penny and Silvia Vilches, 2009. “Re-visioning the Environment of Support for Lone Mothers in Extreme Poverty,” in Public Policy for Women: The State, Income Security and Labour Market Issues, University of Toronto Press, Toronto.

Haas Linda, 1996. “Family Policy in Sweden,” in Journal of Family and Economic Issues, Vol. 17, No.1, Spring.

Haas Linda and C. Philip Hwang, 2008. “The Impact of Taking Parental Leave of Father’s Participation in Childcare and Relationships with Children: Lessons From Sweden,” in Community, Work and Family, Vol.11, No.1, February.

Hagemann-White Carol, Judith Katenbrink and Heike Rabe, 2006. Combating Violence Against Women: Stocktaking Study on the Measures and Actions Taken in Council of Europe Member States, Council of Europe, Directorate General of Human Rights, Strasbourg.

Hagemann Karen, Sonya Michel and Gunilla-Friederike Budde, 2008. Civil Society and Gender Justice: Historical and Comparative Perspectives, Berghahn Books, New York.

Hall, Michael H. Cathy W. Barr, M. Easwaramoorthy, S. Wojciech Sokolowki, and Lester M. Salamon, 2005. The Canadian Nonprofit Sector in Comparative Perspective, Imagine Canada, Toronto.

Hall Peter A., 2002. “Great Britain: The Role of Government and the Distribution of Social Capital,” in Democracies in Flux: The Evolution of Social Capital in Contemporary Society, Putnam ed., Oxford University Press, Oxford.

Hall Peter A., 1999. “Social Capital in Britain,” in British Journal of Political Science, Vol.29, No.3, June.

363

Halleröd Björn, 2009. Minimum Income Schemes: Sweden, Peer Review in Social Protection and Social Inclusion and Assessment in Social Inclusion, The European Commission, April.

Halpern David, 2005. Social Capital, Polity Press, Cambridge. Handy Charles, 1992. “Balancing Corporate Power: A New Federalist Paper,” in Harvard Business Review, Vol.70, No.6, November-December.

Hardin Russell, 2002. Trust and Trustworthiness, Russell Sage Foundation, New York.

Harriss John, 2002. Depoliticizing Development: The World Bank and Social Capital, Anthem Press, London.

Haynes Bruce D. and Hernandez Jesus, 2008. “Place, Space and Race: Monopolistic Group Closure and the Dark Side of Social Capital,” in Networked Urbanism: Social Capital in the City, Talja Blokland and Mike Savage (Eds.), Ashgate Publishing Ltd., Burlington (VT).

Heater Derek, 2004. A Brief History of Citizenship, New York University Press, New York.

Heitlinger Alena, 1991. “Pronatalism and Women’s Equality Policies,” in European Journal of Population, Vol.7, No.4, December.

Herd Pamela and Madonna Harrington Meyer, 2002. “Care Work: Invisible Civic Engagement,” in Gender and Society, Vol.16, No.5, October.

Herda-Rapp Ann, 2000. “Gender Identity Expansion and Negotiation in the Toxic Waste Movement,” in The Sociological Quarterly, Vol.41, No.3, Summer.

Hero Rodney E., 2007. Racial Diversity and Social Capital: Equality and Community in America, Cambridge University Press, Cambridge.

Hero Rodney E., and Robert R. Preuhs, 2006. “Multiculturalism and Welfare Policies in the U.S.A: A State-Level Comparative Analysis,” in Multiculturalism and the Welfare State: Recognition and Redistribution in Contemporary Democracies, Keith Banting and Will Kymlicka (Eds.), Oxford University Press, Oxford.

Herreros Francisco, 2000. “Social Capital, Associations and Civic Republicanism,” in Democratic Innovation: Deliberation, Representation and Association,” Michael Saward, Routledge, New York.

Hicks Alexander and Lane Kenworthy, 2003. “Varieties of Welfare Capitalism,” in Socio- Economic Review, Vol.1, No.1.

Hirst Paul Q. and Veit M. Bader, 2001. Associative Democracy: The Real Third Way, Second edition. Routledge, New York.

364

Hobson Barbara, 1994. “Solo Mothers, Social Policy Regimes, and the Logics of Gender,” in Gendering Welfare States, Diane Sainsbury, Sage, London.

Hobson Barbara and Ruther Lister, 2002. “Citizenship,” in Contested Concepts in Gender and Social Politics, Hobson, Lewis, Siim (Eds.), Edward Elgar Publishing Limited, Northhampton.

Holmwood John, 2000. “Thee Pillars of Welfare State Theory: T.H. Marshall, Karl Polanyi, and Alva Myrdal in Defence of the National Welfare State,” in European Journal of Social Theory, Vol.3, No.1.

Hooghe Marc and Dietlind Stolle, 2003. Generating Social Capital: Civil Society and Institutions in Comparative Perspective,” Marc Hooghe and Deitlind Stolle eds., Palgrave Macmillan, New York.

Hooghe Marc, Tim Reeskens, Dietlind Stolle, and Ann Trappers, 2009. “Ethnic Diversity and Generalized Trust in Europe: A Cross-National Multilevel Study,” in Comparative Political Studies, Vol42, No.2, February.

Huber Evelyne and John D. Stephens, 2001. Development and Crisis of the Welfare State: Parties and Politics in Global Markets, University of Chicago Press, Chicago.

Hughes Christina and Loraine Blaxter, 2007. “Feminist Appropriations of Bourdieu: The Case of Social Capital,” in (Mis)recognition, Social Inequality and Social Justice: Nancy Fraser and Pierre Bourdieu, Terry Lovell (Ed.), Routledge, New York.

Hughes Melanie M. and Pamela Paxton, 2008. “Continuous Change, Episodes, and Critical Periods: A Framework for Understanding Women’s Political Representation Over Time,” in Politics and Gender, Vol.4.

Hwang, Monica, Robert Andersen and Edward G. Grabb. 2007. “Voluntary Association Activity in Quebec and English Canada: Assessing the Effects of Region and Language.” Canadian Journal of Political Science, Vol.40, No.1, March.

Immervoll Herwig and David Barber, 2005. Can Parents Afford to Work? Childcare Costs, Tax-Benefit Politics and Work Incentives, OECD, Directorate for Employment, Labour and Social Affairs, Social, Employment and Migration Working Papers Series, Paris, December.

Inglehart Ronald and Pippa Norris, 2003. Rising Tide: Gender Equality and Cultural Change Around the World, Cambridge University Press, Cambridge.

Iyer Nitya, 2002. Working Through the Wage Gap, Report of the Task Force on Pay Equity, February.

Jackman, Robert W. and Ross Alan Miller. 2005. Before Norms: Institutions and Civic Culture, Ann Arbor: University of Michigan Press.

365

Janoski Thomas, 1998. Citizenship and Civil Society: A Framework of Rights and Obligations in Liberal, Traditional, and Social Democratic Regimes, Cambridge University Press, Cambridge.

Janoski Thomas and Brian Gran, 2002. “Political Citizenship: Foundations of Rights,” in Handbook of Citizenship Studies, Engin F. Insin, Bryan S. Turner eds., Sage Publications, London.

Jaumotte Florence, 2004. Labour Force Participation of Women: Empirical Evidence on the Role of Policy and Other Determinants in OECD Countries, OECD Economic Studies, No.37, 2003/2.

Jenson Jane, 2007. “The European Union’s Citizenship Regime. Creating Norms and Building Practices,” in Comparative European Politics, Vol.5, No.1.

Jenson Jane and Mariette Sineau, 2001. “France: Reconciling Republican Equality with ‘Freedom of Choice,’” in Who Cares? Women’s Work, Childcare and Welfare State Redsign, Jane Jenson, Mariette Sineau eds., University of Toronto Press.

Jenson Jane and Sherry Thompson, 1999. Comparative Family Policy: Six Provincial Stories, Canadian Policy Research Networks, Renouf Publichsing Co. Ltd., Ottawa, Study No. F08.

Jepperson Ronald L., 2002. “Political Modernities: Disentangling Two “Underlying Dimensions of Institutional Differentiation,” Sociological Theory, Vol.20, No.1, March.

Jepperson Ronald L., 2000. Institutional Logics: On the Constitutive Dimensions of the Modern Nation-State Polities, European Forum Series, European University Institute, Robert Schuman Centre for Advanced Studies, RSC No. 2000/36.

Johansson Håkan and Bjørn Hvinden, 2007. Citizenship in Nordic Welfare States: Dynamics of Choice, Duties and Participation in a Changing Europe, Håkan Johansson and Bjørn Hvinden (Eds.), Routledge, New York.

Johnston Richard and Stuart N. Soroka, 2001.“Social Capital in a Multicultural Society: The Case of Canada,” in Social Capital and Participation in Everyday Life, Paul Dekker and Eric M. Uslaner (Eds.), Routledge, New York.

Junn Jane, 2007. “Square Pegs and Round Holes: Challenges of Fitting Individual-Level Analysis to a Theory of Politicized Context of Gender,” in Politics & Gender, Vol.3, No.1.

Jusko Karen Long and Shively W. Phillips, 2005. “Applying a Two-Step Strategy to the Analysis of Cross-National Public Opinion Data,” in Policy Analysis, Vol.13, No.4, Autumn.

Kääriäinen Juha and Heikki Lehtonen, 2006. “The Variety of Social Capital in Welfare State Regimes - A Comparative Study of 21 Countries” in European Societies, Vol.8, No.1, March.

366

Kaase Max, 1999. “Interpersonal Trust, Political Trust and Non-institutionalised Political Participation,” in West European Politics, No. 22, Vol.3.

Karp Jeffrey A. and Banducci Susan A., 2008. "When Politics is Not Just a Man's Game: Women's Representation and Political Engagement." Electoral Studies, Vol. 27, No.1.

Kemeny Jim,1981. The Myth of Home-Ownership: Private Versus Public Choices in Housing Tenure, Routledge, London.

Kenworthy Lane and Melissa Malami, 1999. “Gender Inequality and Political Representation: A Worldwide Comparative Analysis,” in Social Forces, Vol.78, No.1, September.

Kershaw Paul, 2004. “‘Choice’ Discourse in BC Child Care: Distancing Policy From Research, in Canadian Journal of Political Science, December, Vol. 37, No.4.

Kesler Christel and Irene Bloemraad, 2010. “Does Immgiration Erode Social Capital? The Conditional Effects of Immigration-Generated Diversity on Trust, Membership and Participation Across 19 Countries, 1981-2000,” in Canadian Journal of Political Science, Vol.43, No.2, June.

Kidd Alan, 2002. “Civil Society or the State: Recent Approaches to the History of Voluntary Welfare,” in Journal of Historical Sociology, Vol.15, No.3, September.

Kilkey Majella and Jonathan Bradshaw, 1999. “Lone Mothers, Economic Well-Being, and Policies,” in Gender and Welfare State Regimes, Diane Sainsbury ed., Oxford University Press, Oxford.

King Desmond S. and Jeremy Waldron, 1988. “Citizenship, Social Citizenship and the Defence of Welfare Provision,” in British Journal of Political Science, Vol.18, No.4, October.

Korpi Walter, 2000. “Faces of Inequality: Gender, Class and Patterns of Inequalities in Different Types of Welfare States,” in Social Politics, Vol.7, No.2.

Korpi Walter, 1980. “Social Policy and Distributional Conflict in the Capitalist Democracies: A Preliminary Comparative Framework,” in Western European Politics, Vol.3, No.3.

Korpi Walter, 1975. “Poverty, Social Assistance and Social Policy in Post-War Sweden,” Acta Sociologica, Vol. 18, No. 2-3.

Korpi Walter and Joakim Palme, 1998. “The Paradox of Redistribution and Strategies of Equality: Welfare State Institutions, Inequality and Poverty in the Western Countries,” in American Sociological Review, Vol.63, No.5, October.

Koven Seth and Sonya Michel, 1993. Mothers of a New World: Maternalist Politics and the Origins of Welfare States, Routledge, New York.

367

Koven Seth and Sonya Michel, 1990. “Womanly Duties: Maternalist Politics and the Origins of Welfare States in France, Germany, Great Britain and the United States, 1880-1920,” in The American Historical Review, October,, Vol.95, No.4.

Krasner Stephen D., 1984. “Approaches to the State: Alternative Conceptions and Historical Dynamics,” in Comparative Politics, Vol.16, No.2, January.

Kumlin Staffan and Bo Rothstein, 2005. “Making and Breaking Social Capital: The Impact of Welfare State Institutions,” in Comparative Political Studies, Vol.38, No.4, May.

La Due Lake Ronald and Robert Huckfeldt, 1998. “Social Capital, Social Networks, and Political Participation,” in Political Psychology, Vol.19, No.3, September.

Le Hir Marie-Pierre, 2000. “Cultural Studies Bourdieu’s Way: Women, Leadership, and Feminist Theory,” in Pierre Bourdieu: Field Work in Culture, Nicholas Brown, Imre Szeman (Eds.), Rowman and Littlefield Publishers Inc., Oxford.

Laborde Cécile, 2004. “Republican Citizenship and the Crisis of Integration in France,” in Lineages of European Citizenship: Rights, Belonging and Participation in Eleven Nation- States, Richard Bellamy, Dario Castiglione, Emilio Santoro (Eds.), Palgrave Macmillan, New York.

Laitkin David D., 1995. “The Civic Culture At 30,” in The American Political Science Review, Vol.89, No.1, March.

Lam Pui-Yan, 2006. “Religion and Civic Culture: A Cross-National Study of Voluntary Association Membership,” in Journal for the Scientific Study of Religion, Vol. 45.

Lancee Bram and Jaap Dronkers, 2008. “Ethnic Diversity in Neighbourhoods and Individual Trust of Immigrants and Natives: A Replication of Putnam in a West-European Country,” Paper presented at the International Conference on Theoretical Perspectives on Social Cohesion, Royal Flemish Academy of Belgium for Science and the Arts, Brussels.

Larsen Christian Albrekt, 2008. “The Institutional Logic of Welfare Attitudes: How Welfare Regimes Influence Public Support,” in Comparative Political Studies, Vol.41, No.2, February.

Larsen Christian Albrekt, 2006. Social Capital and Welfare Regimes: The Impact of Institution Dependent Living Conditions and Perceptions of Poor and Unemployed, Centre for Comparative Welfare Studies, Working Paper No.2006-38, Aalborg University, Denmark.

Laurence James and Anthony Heath, 2008. Predictors of Community Cohesion: Multi-level Modeling of the 2005 Citizenship Survey, Department of Communities and Local Government, London.

Lehning Percy B., 1998. “Towards a Multicultural Civil Society: The Role of Social Capital and Democratic Citizenship” in Government and Opposition, Vol.33, No.2.

368

Leigh Andrew, 2006. “Trust, Inequality and Ethnic Heterogeneity,” in The Economist Record, Vol.82.

Leitner Sigrid, 2003. “Varieties of Familialism: The Caring Function of the Family in Comparative Perspective,” European Societies, Vol.5, No.4.

Letard Valerie, 2005. Rapport D'Information No.334 Fait au nom de la commission des Affaires sociales sur les minima sociaux, Comission des Affaires Sociales sur les Minima Sociaux, Paris, Presented Mai 11th.

Letki Natalia, 2008. “Does Diversity Erode Social Cohesion? Social capital and Race in British Neighbourhoods,” in Political Studies, Vol.52.

Levi Margaret, 2000. “Political Trust and Trustworthiness,” in Annual Review of Political Science, Vol.3.

Lewis Jane, 1992. “Gender and the Development of Welfare Regimes,” in Journal of European Social Policy, Vol.2, No.3.

Lister Ruth, Fiona Williams, Anneli Anttonen, Jet Bussemaker, Ute Gerhard, Jacqueline Heinen, Stina Johansson, Arnlaug Leira, Birte Siim and Constanza Tobio with Anna Gavanas, 2007. Gender Citizenship in Western Europe: New Challenges for Citizenship Research in a Cross-National Context, The Policy Press, University of Bristol, Bristol.

Lister Ruth, 2005. “Feminist Citizenship Theory: An Alternative Perspective on Understanding Women’s Social and Political Lives,” in Women and Social Capital, Jane Franklin (Ed.), Families and Social Capital ESRC Research Group, London South Bank University, London, April.

Lister Ruth, 2003. Citizenship: Feminist Perspectives, New York University Press, New York.

Lister Ruth, 1997a. “Dialectics of Citizenship,” in Hypatia, Vol.12, No.4, Fall.

Lister Ruth, 1997b. “Citizenship: Towards a Feminist Synthesis,” in Feminist Review, Vol.57, No.1.

Loury Glenn C., 2002. The Anatomy of Racial Inequality, Harvard University Press.

Loury Glenn C.,, 1977. “A Dynamic Theory of Racial Income Differences,” in Women, Minorities and Employment Discrimination, Phillis Ann Wallace, Annette M. Lamond (Eds.), Lexington Books, Lexington.

Lovell Terry, 2000. “Thinking Feminism With and Against Bourdieu,” in Reading Bourdieu on Society and Culture, Bridget Fowler (Ed.), Blackwell Publishers, Oxford.

369

Lowndes Vivien, 2006. “It’s Not What You’ve Got, But What You Do With It: Women, Social Capital and Political Participation,” in Gender and Social Capital, Brenda O’Neill and Elisabeth Gidengil (Eds.), Routlege.

Lowndes Vivien, 2004. “Getting on or Getting By: Social Capital and Political Participation,” in British Journal of Politics & International Relations, Vol. 6, No.1, February.

Lowndes Vivien, 2000. “Women and Social Capital: A Comment on Hall’s ‘Social Capital in Britain,’” in British Journal of Political Science, Vol.30, No.3, July.

Luke Douglas A., 2004. Multilevel Modeling, Sage University Paper #07-143, Quantitative Applications in the Social Sciences Series, London.

Lyttelton Adrian, 2000. “Liberalism and Civil Society in Itlay: From Hegemony to Mediation,” in Civil Society Before Democracy: Lessons From Nineteenth-Century Europe, Nancy Gina Bermeo and Philip G. Nord (Eds.), Rowman & Littlefield Publishers Inc.

McCarty Therese A., 1993. “Demographic Diversity and the Size of the Public Sector,” in Kyklos, Vol.46.

McClain Linda, 2001. “The Domain of Civic Virtue in a Good Society: Families, Schools, and Sex Equality,” in Fordham Law Review, Vol.69, No.5, April.

McClurg Scott D., 2003. ‘Social Networks and Political Participation: The Role of Social Interaction in Explaining Political Participation,” in Political Research Quarterly, Vol.56, No.4, December.

McDonagh Eileen, 2002. “Political Citizenship and Democraticzation: The Gender Paradox,” in American Political Science Review, Vol.96, No.3.

McPherson Miller, Lynn Smith-Lovin and James M. Cook, 2001. “Birds of a Feather: Homophily in Social Networks,” in Annual Review of Sociology, Vol.27.

Mabille François and Corinne Valasik, 2004. “Welfare, Church and Gender in France,” in Welfare, Church and Gender in Eight European Countries, Ninna Edgardh Beckman ed., Welfare and Religion in a European Perspective: Working Paper 1, Uppsala University, Uppsala.

Madsen Per Kongshøj, 2009. “Denmark,” in The Labour Market Triangle: Employment Protestion, Unemployment Compensation and Activation in Europe, Paul de Beer and Trudie Schilis (Ed.), Edward Elgar Publishing Ltd., Northamptom.

Mandel Hadas and Michael Shalev, 2009. “How Welfare States Shape the Gender Pay Gap: A Theoretical and Comparative Analysis,” in Social Forces, Vol.87, No.4, June.

370

Mandel Hadas and Moshe Semyonov, 2003. Welfare Family Policies and Gender Earnings Inequality: A Cross-National Comparative Analysis, Luxemburg Income Study, Working Paper Series No.364, December.

Mann Michael, 1987. “Ruling Class Strategies and Citizensihp,” in Sociology, Vol.21, No.3, August.

Mariën Sofie, 2008. Trends and Gender Differences in Political Participation and Political Trust: A Comparative Analysis, PartiRep Working Paper, PartiRep Research Network, Brussels/Leuven, January.

Marshall T.H., 1992. “Citizsenship and Social Class (1950),” in Citizenship and Social Class, T.H. Marshall, and Tom Bottomore, Pluto Press, London.

Mayer Margit, 2003. “The Onward Sweep of Social Capital: Causes and Consequences for Understanding Cities, Communities and Urban Movements,” in International Journal of Urban and Regional Research, Vol.27, No.1, March.

Meagher Sharon M. and Patrice DiQuinzio, 2005. Women and Children First: Feminism, Rhetoric and Public Policy, State University of New York Press, Albany.

Melby Kari, Anna-Birte Ravn and Christina Wetterberg, 2009. “A Nordic Model of Gender Equality?” in Gender Equality and Welfare Politics in Scandinavia: The Limits of Political Ambition? Kari Melby, Anna-Birte Ravn and Christina Wetterberg (Eds.), The Policy Press, Bristol.

Mettler Susan, 2005. Soliders to Citizens: The G.I. Bill and the Making of the Greatest Generation, Oxford University Press, Oxford.

Millar Jane, 1996. “Mothers, Workers, and Wives” in Good Enough Mothering: Feminist Perspectives on Lone Mothering, Elizabeth Bortolaia Silva (Ed.), Routledge, London.

Mink Gwendolyn, 2002. “Violating Women: Rights Abuses in the American Welfare Police State,” in Women’s Work is Never Done: Comparative Studies in Care-Giving Employment, and Social Policy Reform, Sylvia Bashevkin ed., Routledge, New York.

Mink Gwendolyn, 1990. “The Lady and the Tramp: Gender, Race, and the Origins of the American Welfare State,” in Women, the State and Welfare, Linda Gordon (Ed.), University of Wisconsin Press, Madison.

Mishler William and Richard Rose, 2001. “What Are the Origins of Political Trust?” in Comparative Political Studies, Vol.34, No.1.

Moller Stephanie and Joya Misra, 2005. Familialism and Welfare Regimes: Poverty, Employment and Family Policies, Luxembourg Income Study, Working Paper No.399, January.

371

Moller Stephanie, Joya Misra and Michelle Budig, 2007. “Reconciliation Policies and the Effects of Motherhood on Employment, Earnings, and Poverty,” in Journal of Comparative Policy Analysis, Vol.9 No.2.

Mondak Jefferey J. and Mary R. Andersen, 2004. “The Knowledge Gap: A Reexamination of Gender-Based Differences in Political Knowledge,” in Journal of Politics, Vol.66, No.2, May.

Monnier Jean-Marie, 2006. Fiscalisation, politique fiscale globale et recentralisation de la sphère financière publique et sociale, Centre d’économie, équippe MATISSE, L’Université Paris-Sorbonne, Préparé pour: Colloque International: État et Regulation sociale, septembre.

Moore Gwen, 1990. “Structural determinants of Men and Women’s Personal Networks,” in American Sociological Review, Vol.55, No.5, October.

Morel Sylvie, 2002. The Insertion Model or the Workfare Model? Transformations of Social Assistance within Quebec and Canada, Status of Women Canada, Canadian Cataloguing in Publication Data, September.

Morgan Kimberly J., 2006. Working Mothers and the Welfare State: Religion and the Politics of Work-Family Policies in Western Europe and the United States, Stanford University Press, Stanford.

Morgan Kimberly J., 2003. “The Politics of Mother's Employment: France in Comparative Perspective,” in World Politics, Vol.55, January.

Morgan Kimberly J., and Kathrin Zippel, 2003. “Paid to Care: The Origins and Effects of Care Leave Policies in Western Europe,” in Social Politics, Vol.10, No.1, Spring.

Munch Allison, J. Miller McPherson and Lynn Smith-Lovin, 1997. “Gender, Children and Social Contact: The Effects of Childrearing for Men and Women,” in American Sociologist Review, Vol.62, No.4, August.

Murard Numa, 2002. “Mating Welfare and Workfare: Scenes of a Public Policy in a French Province,” in Ethnography, Vol. 13, No.3.

National Council of Welfare, 2006. Welfare Incomes 2005, Ottawa, Summer.

National Council of Welfare, 2004. Poverty Profile 2001, National Council of Welfare Reports, Government of Canada, Vol.122, Ottawa, Autumn.

National Council of Welfare, 2002. Welfare Incomes 2000-2001, Vol.116, Spring.

Neubeck Kenneth J. and Noel A. Cazenave, 2001. Welfare Racism: Playing the Race Card Against America’s Poor, Routledge, New York.

372

Newton Kenneth, 2001. “Trust, Social Capital, Civil Society and Democracy,” in International Political Science Review, Vol.22, No.2.

Nord Philip, 1994. “The Welfare State in France, 1870-1914,” in French Historical Studies, Vol.18, No.3, Spring.

Norris Pippa, 2002 . “Women’s Power at the Ballot Box,” in Voter Turnout Since 1945: A Global Report on Political Participation, Rafael Lopez Pintor and Maria Gratschew (Eds.), International Institute for Democracy and Electoral Assistance, Stockholm.

Norris Pippa and Ronald Inglehart, 2006. “Gender and Social Capital: Bowling in Women’s Leagues?” in Gender and Social Capital, Brenda O'Neill (Ed.), Routledge, New York.

Norris Pippa and Ronald Inglehart, 2004.“Religious Organizations and Social Capital,” in The International Journal for Not-for-Profit Law, Vol.6, Np.4 September.

Norton Andrew, 1998. “The Welfare State: Depreciating Australia’s Social Capital,” in Policy: A Review of Public Policy and Ideas, March-April.

Novak William J. , 2008. “The Myth of the ‘Weak’ American State,” in The American Historical Review, Vol.113, No.3, June.

O’Connor Julia S., 1998. “Gender, Class and Citizenship in the Comparative Analysis of Welfare State Regimes: Theoretical and Methodological Issues,” in Power Resources Theory and the Welfare State: A Critical Approach, University of Toronto Press, Toronto.

O’Ferrall Fergus, 2001. “Civic Republican Citizenship and Voluntary Action,” in The Republic: A Journal of Contemporary and Historical Debate, Vol. ,No.2, Spring/Summer.

O'Hara Kathy, 1998. Comparative Family Policy: Eight Countries' Stories, Canadian Policy Research Networks Inc., CPRN Study No.F 04, Ottawa.

O’Neill Brenda, 2006 “Canadian Women’s Religious Volunteerism: Compassion, Connectsion and Comparisons,” in Gender and Social Capital, Brenda O’Neil and Elisabeth Gidengil (Eds.), Routledge, New York.

O’Neill Brenda and Elisabeth Gidengil, 2005. Gender and Social Capital; Lisa Adkins, “Social Capital: Anatomy of a Troubled Concept,” in Feminist Theory, Vol.6, No.2.

OECD, 2008. Growing Unequal? Income Distribution and Poverty in OECD Countries, Paris.

OECD, 2007a. The Social Expenditure Database: An Interpretive Guide.

OECD, 2007b. PF13: Typology of Childcare and Early Education Services, Family Database, Social Policy Division, Directorate of Employment, Labour and Social Affairs.

373

OECD, 2006, Starting Strong II: Early Childhood Education and Care, Annex E: Canada, Directorate for Education.

OECD, 2006b. Starting Strong II: Early Childhood Education and Care, Annex E: France, Directorate for Education.

OECD, 2006c. Starting Strong II: Early Childhood Education and Care, Annex E: Sweden, Directorate for Education.

OECD, 2006d. Society at a Glance: OECD Social Indicators.

OECD, 2005a. Society at a Glance: OECD Social Indicators.

OECD, 2005b. Usual Working Hours per Week by Gender (Aged 15-64), OECD Family Database.

OECD, 2004. Early Childhood Education and Care Policy in France, Country Note, Directorate for Education, February.

OECD, 2003. Early Childhood Education and Care Policy: Canada Country Note, OECD: Directorate for Education, Second Round of Reviews.

OECD, 2002a. OECD Employment Outlook 2002: Surveying the Job Horizon.

OECD, 2002b. Employment Outlook 2002: Women at Work: Who Are They and How Are They Faring? Paris. Oldfield Adrian, 2003. “Citizenship: An Unnatural Practice?” in Citizenship: Critical Concepts Vol.1, Bryan S. Turner, Peter Hamilton, Routledge, New York.

Orloff Ann Shola, 2006.“From Maternalism to ‘Employment for All:’ State Politicies to Promote Women’s Employment Across the Affluent Democracies,” in The State After Statism: New State Activities in the Age of Liberalization, Jonah D. Levy (Ed.), Harvard University Press, Cambridge.

Orloff Ann Shola, 2002. Women’s Employment and Welfare Regimes: Globalization, Export Orientation and Social Policy in Europe and North America, United Nations Research Institute for Social Development, Social Policy and Development Programme, Paper No.12, June.

Orloff Ann Shola, 1993. “Gender and the Social Rights of Citizenship: The Comparative Analysis of Gender Relations and Welfare States,” in American Sociological Review, Vol. 58, No. 3, June.

Osburn H.G., 2000. “Coefficient Alpha and Related Internal Consistency Reliability Coefficients,” in Psychological Methods, Vol.5, No.3, September.

374

Ostner Ilona, 1996. Individualization, Breadwinner, Norms, and Family Obligations. Gender Sensitive Concept in Comparative Welfare source, Feminist Research Centre in Aalbord, Working Paper No.38, Department of Development and Planning, Aalborg, December.

Palley Howard A. and Elizabeth K. Bowman, 2002. “A Comparison of National Family Policies: France and Sweden,” in Children and Youth Services Review, Vol.24, No.5.

Pateman Carole, 1992. “Equality, Difference, Subordination: The Politics of Motherhood and Women’s Citizenship,” in Beyond Equality and Difference, Gisela Bock and Susan James (Eds.), Routledge, London.

Pateman Carole, 1989. “The Civic Culture: A Philosophical Critique,” in The Disorder of Women: Democracy, Feminism, and Political Theory, Stanford University Press, Stanford.

Pateman Carole, 1988. The Sexual Contract, Stanford University Press, Stanford.

Paxton Pamela, 1999. “Is Social Capital Declining in the United States? A Multiple Indicator Assessment,” in The American Journal of Sociology, Vol.105, No.1, July.

Pedersen Susan, 1995. Family Dependence and the Origins of the Welfare State: Britain and France 1914-45, Cambridge University Press, Cambridge.

Pennant R., 2005. Diversity, Trust and Community Participation in England, Home Office, Research, Development and Statistics Directorate, London.

Périvier Hélène, 2003. Rappels des effets de l’APE rang 2 sur l’activité des femmes, Synthesis of presentation for the Conference: Emploi des femmes et charges familiales: Repenser le congé parental en France à la lumière des expériences étrangères, Centre de recherche en économie de Sciences Politiques, December 3rd.

Pestoff Victor A., 2009. A Democratic Architecture for the Welfare State, Routledge, New York.

Pichler Florian and Claire Wallace, 2007. “Patterns of Formal and Informal Social Capital in Europe,” in European Sociological Review, Vol.23, No.4.

Pierson Paul, 2000. “Increasing Returns, Path Dependence, and the Study of Politics,” in The American Political Science Review, Vol.94, No.2, June.

Popielarz Pamela, 1999. “(In)Voluntary Assocation: A Multilevel Analysis of Gender Segregation ins Voluntary Organizations,” in Gender and Society, Vol.13, No.2, April.

Portes Alejandro, 1998. “Social Capital: Its Origins and Applications in Modern Sociology,” in Annual Review of Sociology, Vol.24.

375

Portes Alejandro and Patricia Landolt, 1996. “Unsolved Mysteries – The Tocqueville Files II: The Down Side of Social Capital,” in The American Prospect, Vol.7, Issue 26, May-June.

Priemus Hugo and Frans Dieleman, 2002. “Social Housing Policy in the European Union: Past, Present and Perspectives,” in Urban Studies, Vol. 39, No.2.

Prince Michael J., 1998. “Holes in the Safety Net, Leaks in the Roof: Changes in Canadian Welfare Policy and Their Implications for Social Housing Programs,” in Housing Policy Debate, Vol.9, No.4.

Prince Michael K., 1995. “The Canadian Housing Policy Context,” in Housing Policy Debate, Vol.6, No.3.

Quadagno Jill , 1994. The Color of Welfare: How Racism Undermined the War on Poverty, Oxford University Press, Oxford.

Robert D. Putnam, 2007. “E Pluribus Unum: Diversity and Community in the Twentiy-first Century (The 2006 Johan Skytte Prize Lecture),” in Scandinavian Political Studies, Vol.30, No.2.

Putnam Robert D., 2002., Democracies in Flux: The Evolution of Social Capital in Contemporary Society, Oxford University Press, Oxford.

Putnam Robert D., 2000. Bowling Alone: The Collapse and Revival of American Community, Simon & Schuster, New York.

Putnam Robert D., 1996. “The Strange Disappearance of Civic America,” in The American Prospect, Vol.7, Issue 24, December.

Putnam Robert D., 1995. “Bowling Alone: America’s Declining Social Capital,” in Journal of Democracy, Vol.6, No.1.

Putnam Robert D., 1993. Making Democracy Work: Civic Traditions in Modern Italy, with Robert Leonardi, Raffaella Y. Nanetti, Princeton University Press, Princeton.

Raissiguier Catherine, 2003. “Troubling Mothers: Immigrant Women From Africa in France,” in Jenda: A Journal of Culture and African Women Studies, No.4.

Ramirez Francisco O., Yasemin Soysal and Suzanne Shanahan, 1997. “The Changing Logic of Political Citizenship: Cross-National Acquisition of Women’s Suffrage Rights, 1890- 1990,” in American Sociological Review, Vol.62, No.5, October.

Raudenbush Stephen W. and Anthony S.Bryk, 2002. Hierarchical Linear Models: Applications and Data Analysis Methods, Second Edition, Sage Publications, London.

376

Ray Rebecca, Janet C. Gornick and John Schmitt, 2009. Parental Leave Policies in 21 Countries: Assessing Generosity and Gender Equality, Center for Economic and Policy Research, Washington.

Reeskens Tim and Mark Hooghe, 2008. “Cross-cultural Measurement Equivalence of Generalized Trust. Evidence from the European Social Survey (2002 and 2004),” in Social Indicators Research, Vol.85, No.3.

Reher David Sven, 1998. “Family Ties in Western Europe: Persistent Contrasts,” in Population and Development Review, Vol. 24, No. 2, June.

Reitsma-Street Marge, Josie Schofield, Brishkai Lund and Colleen Kasting, 2001. Housing Policy Options for Women Living in Urban Poverty: An Action Research Project in Three Canadian Cities, Status of Women Canada, Ottawa, February.

Revillard Anne, 2006. “Work/Family Policy in France: From State Familialism to State Feminism?,” in International Journal of Law, Policy and the Family, Vol.20, No.2.

Risman Barbara J. and Myra Marx Ferree, 1995. “Making Gender Visible: Comment on Coleman,” in American Sociological Review, Vol.60, No.5, October.

Rojas Mauricio, 2005. Sweden After the Swedish Model: From Tutorial State to Enabling State, Timbro publishers, Stockholm, May.

Rothstein Bo, 2005. Social Traps and the Problem of Trust: Theories of Institutional Design, Cambridge University Press, Cambridge.

Rothstein Bo, 2002. “Sweden: Social Capital in the Social Democratic State,” in Democracies in Flux: The Evolution of Social Capital in Contemporary Society, Putnam ed., Oxford University Press, Oxford.

Rothstein Bo, 1998. Just Institutions Matter: The Moral and Political Logic of the Universal Welfare State, Cambridge University Press, Cambridge.

Rothstein Bo, 1991. “State Structure and Variation in Corporatism: The Swedish Case,” in Scandinavian Political Studies, Vol.14, No.2.

Rothstein Bo, Dietlind Stolle, 2008. “The State and Social Capital: An Institutional Theory of Generalized Trust,” in Comparative Politics, Vol.40, No4., July.

Rothstein Bo, Dietlind Stolle, 2003. “Social Capital, Impartiality and the Welfare State: An Institutional Approach,” in Generating Social Capital: Civil Society and Institutions in Comparative Perspective, Palgrave MacMillan.

Rothstein Bo and Eric M. Ulsaner, 2005. “All for All: Equality, Corruption, and Social Trust,” in World Politics, No. 58, October.

377

Rule Wilma, 1994. “Women’s Representation and Electoral Systems,” in Political Science and Politics, Vol.27, No.4, December.

Ryan Barbara, 1992. Feminism and the Women’s Movement: The Dynamics of Change in Social Movement Ideology and Activism, Routledge, New York.

Sabetti Flilippo, 2007. “Democracy and Civic Culture,” in Oxford Handbook of Comparative Politics, Carles Boix, Susan Carol Stokes (Eds.), Oxford University Press, Oxford.

Sabetti Filippo, 1996. “Path Dependency and Civic Culture: Some Lessons From Italy About Interpreting Social Experiments,” in Politics & Society, Vol.24, No.1, March.

Sainsbury Diane, 1993. “Dual Welfare and Sex Segregation of Access to Social Benefits: Income Maintenance Policies in the UK, the US, the Netherlands and Sweden,” in Journal of Social Policy, Vol.22, No.1.

Sainsbury Diane, 1991. “Analysing Welfare State Variation: The Merits and Limitations of Models Based on the Resodual-Institutional Distinction,” in Scandinavian Political Studies, Vol.14, No.1.

Saint-Jacques G., 2009. Individualization of Taxes and Transfers and the Labor Supply Decision of Women: Simulating the Current French System and Possible Reforms, Paris School of Economics, Paris, September.

Salamon Lester M., 1996. Defining the Nonprofit Sector: The United States, Working Papers of the John Hopkins Comapartive Nonprofit Sector Project, No.18, Lester M. Salamon and Helmut K. Anheier, John Hopkins Institute for Policy Studies, Baltimore.

Salamon Lester M. and Helmut K. Anheirer, 1998. “Social Origins of Civil Society: Explaining the Nonprofit Sector Cross-Nationally,” in Voluntas: International Journal of Voluntary and Nonprofit Organizations, Vol.9, No.3.

Salamon Lester M. and Helmut K. Anheirer, 1997. Defining the Nonprofit Sector: A Cross- national Analysis, Johns Hopkins Comparative Nonprofit Sector Project, Manchester University Press.

Salamon Lester M. and Wojciech Sokolwski, 2001. Volunteering in Cross-National Perspective: Evidence From 24 Nations,” Working Papers of the John Hopkins Comparative Nonprofit Sector Project, No.40, Baltimore, The John Hopkins Centre for Civil Society Studies.

Salamon Lester M., Wojciech Sokolowski, and Regina List, 2003. Global Civil Society: An Overview, The John Hopkins Comparative Nonprofit Sector Project, Baltimore.

Saraceno Chiara, 1994. “The Ambivalent Familialism of the Italian Welfare State,” in Social Politics, Spring.

378

Sarvasy Wendy, 1997. “Social Citizenship From a Feminist Perspective,” in Hypatia, Fall.

Schlesinger Arthur M., 1944. “Biography of a Nation of Joiners,” in The American Historical Review, Vol.50, No.1, October.

Schmitter Philippe C., 1974. “Still the Century of Corporatism?” in The Review of Politics, Vol. 36, No.1, January.

Schnabel Claus and Joachim Wagner, 2005. “Determinants of Union Membership in 18 EU Countries: Evidence from Micro Data, 2002/2003,” in Institute for the Study of Labor, Discussion Paper No.1464, Bonn.

Schlozman Kay, 2000. “Did Working women Kill the PTA?” in The American Prospect, Vol.11, No.20, September.

Schlozman Kay Lehman, Henry E. Brady and Sidney Verba, 1997. “The Big Tilt: Participatory Inequality in America,” in The American Prospect Online, Vol.8, Issue 32, May- June.

Schmitt Neal, 1996. “Uses and Abuses of Coefficient Alpha,” in Psychological Assessment, Vol.8, No.4.

Schofer Evan and Marion Fourcade-Gourinchas, 2001. “The Structural Contexts of Civic Engagement: Voluntary Association Membership in Comparative Perspective,” in American Sociological Review, Vol.66, No.6, December.

Schuurman Frans J., 2003. “Social Capital: The Politico-Emancipatory Potential of a Disputed Concept,” in Third World Quarterly, Vol.24, No.6, December.

Seibel Wolfgang, 1990. “Government/Third-Sector Relationship in a Comparative Perspective: The Cases of France and West Germany,” in Voluntas, Vol.1, No.1, May.

Siim Birte, 2004. “Globalization, Democracy and Participation – The Dilemmas of the Danish Citizenship Model,” in The Politics of Inclusion and Empowerment: Gender, Class and Citizenship, Jogn Andersen, Birte Siim.

Siim Birte, 2000. Gender and Citizenship: Politics and Agency in France, Britain and Denmark, Cambridge University Press, Cambridge.

Smidt Corwin, 1999. “Religion and Civic Engagement: A Comparative Analysis,” in The ANNALS of the American Academy of Political and Social Science, Vol. 565, No. 1, September.

379

Skocpol Theda, 1999 “Advocates Without Members: The Recent Transformation of American Civil Life,” in Civic Engagement in American Democracy, Theda Skocpol and Morris P. Fiorina (Eds.), Brookings Institute Press, Washington.

Skocpol Theda, 1997. “The Tocqueville Problem: Civic Engagement in American Democracy,” in Social Science History, Vol.21, No.4, Winter.

Skocpol Theda, 1996. “Unsolved Mysteries: The Tocqueville Files: Unravelling From Above” in The American Prospect, Vol.7, Issue 25, March-April.

Skocpol Theda, 1995. Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States, Harvard University Press, Cambridge.

Skocpol Theda and Gretchen Ritter, 1991. "Gender and the Origins of Modern Welfare Policies", in Studies in American Political Development, No.5, Spring.

Skocpol Theda, Marshall Ganz and Ziad Munson, 2000. “A Nation of Organizers: The Institutional Origins of Civic Voluntarism in the United States,” in The American Political Science Review, Vol.94, No.3, September.

Skcopol Theda and Morris Fiorina, 1999. “Making Sense of the Civic Engagement Debate,” in Civic Engagement in American Democracy, Theda Skcopol and Morris Fiorina (Eds.), Russell Sage, Washington.

Snyder R. Claire, 2004. “Machiavelli and the Citizenship of Civic Practices,” in Feminist Interpretations of Niccolo Machiavelli, Maria J. Falco (Ed.), The Pennsylvanian State University Press, University Park.

Snyder R. Claire, 1999. Citizen-Soldiers and Manly Warriors: Military Service and Gender in the Civic Republican Tradition, Rowman and Littlefield Publishers Inc., Lanham.

Spencer Jennifer W., Thomas P. Murtha and Stefanie A. Lenway, 2003. How Governments Matter to New Industry Creation, The GW Center for the Study of Globalization, Occasional Paper Series, Washington, February.

Statistics Canada, 2006. General Social Survey: Paid and Unpaid Work.

Statistics Canada, 2001a. Census of Canada.

Statistics Canada, 2001b. Profile of Canadian families and Households: Diversification Continues.

Statistics Canada, 2003. 2001 Census: Earnings of Canadians – Making a Living in the New Economy, Ottawa, March.

380

Stolle Dietlind and Marc Hooge, 2005. “Inaccurate, Exceptional, One-Sided or Irrelevant? The Debate about the Alleged Decline of Social Capital and Civic Engagement in Western Societies,” in British Journal of Political Science, Vol.35, No.1.

Stolle Dietlind and Michele Micheletti, 2006. “The Gender Gap Reversed: Political Consumerism as a Woman-Friendly Form of Civic and Political Engagement,” in Gender and Social Capital, Brenda O’Neill and Elisabeth Gidengil, Routledge, New York.

Strong-Boag Veronica, 1995. “Wages for Housework: Mothers’ Allowances and the Beginnings of Social Security in Canada,” in Social Welfare Policy in Canada: Historical Readings, Raymond B.Blake, Jedd Keshen, eds., Copp Clark Ltd., Toronto.

Studlar Donaley T. and Richard E. Matland, 1996. “The Dynamics of Women’s Representation in the Canadian Provinces: 1975-1994,” in Canadian Journal of Political Science, Vol29, No.2, June.

Swedish Ministry of Health and Social Affairs, 2005. Swedish Family Policy Fact Sheet, Government of Sweden, Regeringskanskliet, Fact Sheet No.11, April.

Taylor Marylee C., 1998. “How White Attitudes Vary with the Racial Composition of Local Populations: Numbers Count,” in American Sociological Review, Vol.63, No.4, August.

Taylor-Gooby Peter, 2005. “Is the Future America? Or, Can Left Politics Preserve European Welfare States from Erosion Through Growing ‘Racial’ Diversity?” in Journal of Social Policy, Vol.34, No.4.

The Swedish Institute, 1996. Housing and Housing Policy in Sweden, April.

Titmuss Richard, 1969. “The Social Division of Welfare,” in Essays in Welfare, Beacon Press, Boston.

Topper Keith, 2005. The Disorder of Political Inquiry, Harvard University Press, Cambridge.

Trägårdh Lars, 2007. “Democratic Governance and the Creation of Social Capital in Sweden: The Discreet Charm of Governmental Commissions,” in State and Civil Society in Northern Europe: The Swedish Model Reconsidered, Lars Trägårdh, ed., Berghahn Books, New York.

Turner Bengt and Chrstine M. E. Whitehead, 2002. “Reducing Housing Subsidy: Urban Studies, Vol.39., No.2, February.

Turner Bryan S., 1990. “Outline of a Theory of Citizenship,” in Sociology, Vol.24, No.2.

United Nations, 2009. Trends in International Migrant Stock: The 2008 Revision, Department of Economic and Social Affairs, Population Division.

Uslaner Eric M., 2002a. The Moral Foundations of Trust, Cambridge University Press, New York.

381

Uslaner Eric M., 2002b. “Religion and Civic Engagement in Canada and the United States,” in Journal of the Scientific Study of Religion, No.41, Vol.2.

Uslaner Eric M., 1999. “Democracy and Social Capital,” in Democracy and Trust, Mark Warren ed., Cambridge University Press, Cambridge.

Uslaner Eric M. and Gabriel Badescu, 2004. “Honest, Trust and Legal Norms in the Transition to Democracy: Why Bo Rothstein is Better Able to Explain Sweden than Romania,” in Creating Social Trust in Post-Socialist Transition, János Kronai, Bo Rothstein and Susan Rose-Ackerman (Eds.), Palgrave MacMillan, New York.

Ullman Claire Frances, 1998. The Welfare State’s Other Crisis: Explaining the New Partnership Between Nonprofit Organizations and the State in France, Indiana University Press, Bloomington. van der Meer Thomas, Peer Scheepers and Manfred te Grotenhuis, 2009. “States As Molders of Informal Relations?” in European Societies, Vol.11, No.2. van Kerbergen Kees and Monique Kremer, 2008. “ and the Welfare State: Intervening to Preserve,” in Culture and Welfare State: Values and Social Policy in Comparative Perspective, Wim van Oorschot, Michael Opielka and Birgit Pfau-Effinger (Eds.), Edward Elgar Publishing, Cheltenham. van Oorschot Wim and Ellen Finsveen, 2009. “The Welfare State and Social Capital Inequality: An Exploration Using Longitudinal Europeean/World Values Study data From 13 Western Welfare States,” in European Societies, Vol.11, No.2. van Oorschot Wim and Wilfred Uunk, 2007. “Multi-level Determinants of the Public’s Informal Solidarity Towards Immigrants in European Welfare States,” in Social Justice, Legitimacy and Welfare State, Ashgate, Hampshire. van Oorschot Wim and Wils Arts, 2005. “The Social Capital of European Welfare States: The Crowding-Out Hypothesis Revisted,” in Journal of European Social Policy, Vol.15, No.1.

Valelly Rick, 1996. “Unsolved Mysteries: The Tocqueville Files: Couch-Potato Democracy?” in The American Prospect, Vol.7, Issue 25, March-April.

Vassallo Francesca, 2004. “Political Activism and Association Involvement: The French Choice in Comparative Perspective,” in French Politics, Vol.2, No.3, December.

Verba Sidney, Kay Lehman Scholzman and Henry E. Brady, 1995. Voice and Equality: Civic Voluntarism in American Politics, Harvard University Press, Cambridge.

Visser Jelle, 2006. “Union Membership Statistics in 24 Countries,” in Monthly Labor Review, Vol. 129, No.1.

382

Vosko Leah, 2002. “Mandatory ‘Marriage’ or Obligatory Waged Work: Social Assistance and Single Mothers in Wisconsin and Ontario,” in Women’s Work is Never Done, Sylvia Bashevkin ed., Routledge, New York.

Wadensjö Eskil, 2009. “Sweden,” in The Labour Market Triangle: Employment Protestion, Unemployment Compensation and Activation in Europe, Paul de Beer and Trudie Schilis (Ed.), Edward Elgar Publishing Ltd., Northamptom.

Wall Ellen, Gabriele Ferrazzi, and Frans Schryer, 1998. “Getting the Goods on Social Capital,” in Rural Sociology, Vol.63, No.2, June.

Warin Philippe, 2002. “The Role of Nonprofit Associations in Combating Social Exclusion in France,” in Public Administration and Development, Vol.22, No.1.

Wekerle Gerda R., 1997. “The Shift to the Market: Gender and Housing Disadvantage,” in Women and the Canadian Welfare State: Challenges and Change, Patricia Evans, Gerda R. Wekerle eds., University of Toronto Press, Toronto.

Whittington Keith E., 1998. “Revisiting Tocqueville’s America: Society, Politics and Association in the Nineteenth Century,” in American Behavioral Scientist, Vol. 42, No.1, September.

Wikander Ulla, Alice Kessler-Harris and Jane Lewis, 1995. Protecting Women’s Labour Legislation in Europe, the United States, and Australia, 1880-1920, University of Illinois Press, Chicago.

Wills Garry, 2002. “Putnam’s America,” in The American Prospect, November.

Wilson John and Thomas Janoski, 1995 . “The Contribution of Religion to Volunteer Work,” in Sociology of Religion, Vol.56, No.2.

Winter Ian, 2000. “Towards a Theorised Understanding of Family Life and Social Capital,” Australian Institute of Family Studies, Working Paper No.21, April.

Woolcock Michael, 1998. “Social Capital and Economic Development: Toward a Theoretical Synthesis and Policy Framework,” in Theory and Society, Vol.27, No.2, April.

Worms Jean-Pierre, 2002. “Old and New Civic and Social Ties in France,” in Democracies in Flux: The Evolution of Social Capital in Contemporary Society, Putnam ed., Oxford University Press, Oxford.

Wuthnow Robert, 2002. “The United States: Bridging the Privileged and the Marginalized?” in Democracies in Flux: The Evolution of Social Capital in Contemporary Society, Putnam ed., Oxford University Press, Oxford.

383

Wuthnow Robert, 1991. Acts of Compassion: Caring for Others and Helping Ourselves, Princeton University Press, Princeton.

Young Iris Marion, 2000. Inclusion and Democracy, Oxford University Press, Oxford.

Young Iris Marion, 1997. “Polity and Group Difference: A Critique of the Ideal of Universal Citizenship,” in Contemporary : An Anthology, Robert E. Goodin and Philip Pettit (Eds.), Blackwell Publishers Ltd., Oxford.

Zukin Cliff, Scott Keeter, Molly Andolina, Krista Jenkins and Michael X. Delli Carpini, 2006. A New Engagement? Political Participation, Civic Life, and the Changing American Citizen, Oxford University Press, Oxford.