Dynamic Institutional Intersections

Figure 11.1 diagrams the “nested intersections” of institutions, family processes, and child and adult outcomes. The first box in the diagram indicates the socioeconomic structures that affect family formation and dissolution noted in the second box, which create the group differences in inequalities in individual outcomes outlined in the lower box. The structural effects occur at the intersections of family, market, and state institutions in a dynamic interplay that varies across and within national contexts over time.

Figure 11.1 Nested intersections of institutions, family processes, and outcomes

Intersectionality is not commonly used to refer to institutional effects. Conventionally, intersectionality is a feminist paradigm emphasizing that no single master social category such as gender or class depicts the different experiences and social locations of all group members (McCall Reference McCall2005). Consequently, intersectionality demands that we consider identities and experiences at the intersection of an individual’s group memberships (Collins Reference Collins2002; McCall Reference McCall2005). Hence, Figure 11.1 includes a shaded circle spanning institutions, family processes, and outcomes to indicate that effects vary at the intersection of an individual’s gender, class, race–ethnicity, immigrant status, sexual orientation, and the like. Due to space constraints, discussion in this chapter is primarily limited to gender and class differences, with education used as the main proxy for class.

Similarly, no master institution accounts for institutional effects on people’s lives. Instead, individual and group differences are nested in the intersections of family, market, and state institutions, and supra-national institutions such as the European Union and World Bank. This may seem like common sense, but bears emphasizing when comparing possible causes and consequences of family inequalities across Europe and the Americas as done in this volume. For example, the institution of family includes norms about what or who constitutes a family, along with the expected behavior of family members. During early industrialization, the patriarchal heterosexual, nuclear, male breadwinner/female carer model of family became hegemonic in many, but not all Western economies.

Men’s economic dominance under this model was theoretically enshrined in US economist Gary Becker’s (Reference Becker1981) specialization theory of family. Becker argued that families in industrial societies optimize household production and reproduction when one partner specializes in paid work and the other in unpaid family work such as housework and child care. The math behind the theory is gender-neutral in that either partner could specialize in paid or unpaid work depending on their individual aptitudes and preferences. However, reflecting the patriarchal world in which he was raised, Becker (Reference Becker1981) ultimately concluded women’s childbearing gives them a comparative advantage in family work, whereas the gender wage gap indicates men’s comparative advantage in employment. Governments in most, but not all, countries reinforced the patriarchal order in the post-World War II expansion of employment-based welfare provisions payable to the primary breadwinner (Cooke Reference Cooke2011).

Perpetuation of the male breadwinner model of family, though, requires a labor market that enables all members of a family to survive on a single income. This became possible for the newly-created European and North American middle classes beginning in the nineteenth century (Cooke Reference Cooke2011). The possibility extended to the working classes as well in the brief postwar period when workers enjoyed the fruits of their growing productivity (Cherlin Reference Cherlin2014; Gottschalk and Smeeding Reference Gottschalk and Smeeding1997). Since then, the evolution from industrial to postindustrial labor markets made the patriarchal male breadwinner model of family increasingly unsustainable.

Deindustrialization and deunionization made less-skilled men the first to lose their ability to support a family. High-wage manufacturing jobs disappeared, replaced by growing employment in low-wage service sector jobs (Carlson, Chapter 1; Cherlin Reference Cherlin2014). Neoliberal policies exacerbated less-skilled workers’ labor market losses. For instance, in a bid to enhance employer flexibility in competitive global markets, some governments eased employment protections and allowed the real value of minimum wages to fall (Immervoll Reference Immervoll2007). Wage inequalities widened further as returns to a university degree sharply increased beginning in the 1980s in the United States and Great Britain, and in the 1990s in other Western countries (Gottschalk and Smeeding Reference Gottschalk and Smeeding1997; Machin Reference Machin, Gregg and Wadsworth2010).

Nonetheless, the middle classes are not safe either. Since the late 1980s, technological expansion has led to falling employment shares among middle-waged occupations in North America (Autor Reference Autor2010) and Europe (Goos, Manning, and Salomons Reference Goos, Manning and Salomons2009), but the specific pattern of polarization varies across Western labor markets, as indicated in Table 11.1. The loss of middle-waged occupations in Austria, France, and Italy was offset by strong growth in the highest-waged occupations. In contrast, the middle-wage job loss in Finland and Norway was offset by growth in only low-wage occupations. In the United Kingdom as in the United States (Autor Reference Autor2010), shrinking middle-wage employment was offset by approximately equal growth in both low- and high-wage jobs.

Consequently, to varying degrees, postindustrial labor markets no longer support the patriarchal male breadwinner model as did the postwar labor markets. In almost every OECD country, the employment rate of prime-age (25 to 54) men decreased since the 1970s (OECD 2016b; see also Eberstadt, Chapter 5). The trajectories of younger adults making critical decisions about education, employment, and family are more precarious than at any other time during the past half-century (Eurofound 2016). The evolving labor markets, coupled with cultural and policy shifts, eroded men’s comparative advantage in employment. Simultaneously, gendered divisions of paid work narrowed.

Evolving Family Divisions of Labor

Any gender and other group hierarchies embedded in the institution of family are constantly contested (Connell and Messerschmidt Reference Connell and Messerschmidt2005; Ferree Reference Ferree2010). In 1963, Friedan’s book on The Feminine Mystique struck a chord with Western housewives who felt trapped, alienated, and vulnerable inside their homes and economically dependent on husbands. Friedan (Reference Friedan1963) called for a revolution, for women to seize education and return to paid work. Friedan’s timing was perfect, coinciding with the introduction of the birth control pill, coupled with the easing of anti-contraception laws in many Western nations (Cooke Reference Cooke2011).

Women particularly embraced higher education. By the 1990s, women’s educational attainment in most Western countries had caught up with men’s. By 2011, college attainment among women aged 25–34 exceeded that of men in twenty-eight of thirty-four OECD countries (OECD 2013a). A consistent pattern across countries is that educated women are more likely to be employed than less-educated women (Cooke Reference Cooke2011; Harkness Reference Harkness, Gornick and Jantti2013; Pettit and Hook Reference Pettit and Hook2009). Nevertheless, growth of the service sector expanded job opportunities for less-skilled women, who take these jobs more often than similarly-skilled men (OECD 2016b). Consequently, and in contrast to men, women’s labor force participation rates steadily increased from the 1970s, as shown in Table 11.2.

The intersection of the state with family and market institutions is evident in the regional variation in these trends. As noted earlier, most postwar welfare state policies reinforced a male breadwinner model. The exception to this was the Nordic model offering extensive policy supports for maternal employment such as public provision of child care and paid parental leave (Cooke and Baxter Reference Cooke and Baxter2010). Consequently, Finnish and Swedish women’s labor force participation rates already exceeded 70% in 1975. The link between women’s education and employment is also weaker in these more egalitarian countries (Harkness Reference Harkness, Gornick and Jantti2013; Pettit and Hook Reference Pettit and Hook2009). At the other end of the policy spectrum, the very low 1975 labor force participation rates of Dutch, Italian, and Spanish women reflected the national policy reinforcement of the male breadwinner model at that time.

The power of policy to change behavior across diverse cultural and historical contexts is evident in the wake of the European Union’s 2000 Lisbon Treaty. The Lisbon Treaty contained an explicit goal of 60% female labor force participation in all member states by 2010, supported by expansion of public child care and other family policies found in the Nordic model.Footnote ³ The over-time trends presented in Table 11.2 indicate the Lisbon strategy had some success. By 2015, the labor force participation rate of women in almost all European countries exceeded that in the United States. Yet one downside of the employment growth in the service economy is the increase in part-time rather than full-time jobs. In all countries, women are more likely than men to work part-time. Still, the percentage of women’s total employment that is part-time varies from a low of 7.4% in the Czech Republic to a high of more than 60% among Dutch women (Table 11.2).

Persistent gender differences in hours employed contributes to the persistent gender wage gap in median earnings, indicated in the final column of Table 11.2. Even in the Nordic countries, the gender wage gap varies from a low of 6% in Norway, to a high of 20% in Finland. Finland’s gender wage gap is in fact larger than the gender wage gap in the less-regulated British and US economies. Nevertheless, women’s gains in employment and relative earnings over the past half-century are extraordinary, although gender employment equality remains elusive under even the most supportive policy framework to date.

One likely reason women made greater employment gains over the past few decades as compared with men is because of their greater predilection for education. Women might also navigate changing labor markets more successfully than men because of the growing employer demand for social in conjunction with cognitive skills (Deming Reference Deming2015). The gender response to US job polarization provides an example of women’s better adaptation. The decrease in middle-wage employment between 1979 and 2007 was more than twice as large for US women as men, 15.8% vs. 7%, respectively (Autor Reference Autor2010, p. 10). Nevertheless, female employment overwhelmingly moved up the occupational wage distribution as middle-wage employment fell. The decrease in US men’s middle-wage employment led to a more even split in employment growth in men’s low- and high-wage occupations (Autor Reference Autor2010). This highlights that the most-skilled men continue to make gains in the new economy, sustaining their advantage over high-skilled women. However, Autor (Reference Autor2010) found evidence of losses even among university-educated men, more of whom became employed in middle- and low-wage occupations. Future comparative work is needed to confirm whether these gender differences occurred in countries with varying patterns of polarization.

In any event, changing gender divisions of paid labor require some adaptation in how unpaid family work gets done. In this area as well, women have made greater behavioral changes than men. Unfortunately, men’s failure to become full partners in unpaid family work encourages greater class inequality among women within and across labor markets.

Institutional Effects on Class Inequalities in Family Forms

Becker (Reference Becker1981, Reference Becker1985) believed that the mutual dependence created by gender specialization in paid or unpaid work enhances marital stability and fertility. Certainly, the three-institutional reinforcement of Becker’s patriarchal model in the postwar decades reinforced a family formation sequence of marriage, childbearing, and children being raised by the two biological parents. This anomalous period in modern industrial history comprised the “golden age of marriage” in Western societies (Festy Reference Festy1980). Couples married earlier, leading to a spike in fertility in many countries (Van Bavel and Reher Reference Van Bavel and Reher2013). Whether the sequence reflected choice or constraint is debatable. Few women had the independent economic resources to remain single or to leave an unhappy or abusive marriage. Conception outside marriage was deeply stigmatizing for both the mother and the child, and most often resulted in either a “shot-gun” wedding or putting the child up for adoption.

Yet institutional intersections are dynamic. They vary across countries at any given point in time, as well as over time within countries. Esteve and Florez-Paredes (see Chapter 2) note that marriage was not historically preeminent in Central and Latin America. Instead, cohabitation and union instability were structural dimensions of family life when marriage reached its zenith in the West (Esteve and Florez-Paredes, Chapter 2). During that time in the West, pronatalist policies combined with supports for maternal employment in Nordic and socialist countries correlated with higher nonmarital fertility rates (Cooke Reference Cooke2011; Perelli-Harris, Chapter 4). Relatedly, cohabitation in social democratic Denmark and Sweden began to increase in the 1960s, a decade ahead of other countries in Europe (Hall and White Reference Hall and White2005, p. 30).

Over the past fifty years, more women throughout Europe and the Americas cohabit rather than marry and raise children outside marriage whether because of divorce or nonmarital childbearing (Carlson, Chapter 1; Esteve and Florez-Paredes, Chapter 2; Perelli-Harris, Chapter 4). What intrigues or worries many social scientists and policymakers are the educational and/or racial–ethnic disparities in these family patterns that reflect and/or magnify inequalities among families. For most Whites in the United States and non-Nordic European countries, avant-garde family arrangements in the 1960s such as cohabitation or divorce were the purview of a small elite. As alternative family forms became more culturally acceptable and legally possible, Goode (Reference Goode1970) anticipated they would become more prevalent among the less- rather than highly educated. His prediction has largely been borne out, although educational gradients in marriage, cohabitation, nonmarital childbearing, and divorce vary in their institutional context (Carlson, Chapter 1).

Within a patriarchal structure, men’s economic capacity predicted by education still plays an important role in women’s family choices as Cherlin argues here and elsewhere (Cherlin Reference Cherlin2014). For example, Cherlin (see Chapter 3) attributes changes in the educational gradient in US single motherhood to the disparate gender earnings effects of job polarization. His prima facie case seems convincing. Recall that moderately educated women improved their earnings position in response to polarization while that of moderately educated men deteriorated (Autor Reference Autor2010). The percentage of US children living with moderately educated single mothers increased in each decade since the 1980s, whereas that for both the least- and most-educated women remained fairly stable since the 1990s (Cherlin, Chapter 3).

However, education predicts much more than economic outcomes – even in the United States – that also have a bearing on family commitment and stability. These associations should not be given short shrift in discussions of educational gradients in family formation because they offer much deeper insights as to the causes as well as consequences of observed family patterns. To date, Kalmijn (Reference Kalmijn2013) is the only demographer to explore the interplays between the institutional context and the possible meanings of education behind the gradients.

Institutional Intersections and Group Differences in Family Outcomes

Senator Moynihan’s Reference Moynihan1965 report for the US Department of Labor brought discussion of child outcomes associated with single-mother households into the public debate. In that report, he noted the very high divorce and nonmarital birth rates among African–American families as compared with Whites, and the strong correlations between father absence and children’s low intelligence scores, school truancy, crime, drug addiction, etc. (Moynihan Reference Moynihan1965). There were very strong educational gradients in effects that he attributed to the deep-seated US racism undermining African–American men’s access to education and employment and, in turn, their relative employment advantage over African–American women. Moynihan (Reference Moynihan1965, p. 29) concluded the poor intergenerational outcomes were indicative of the “pathology of matriarchy” within a society that presumed and rewarded male leadership in public and private life. Moynihan did not consider patriarchy superior or inevitable, just normative at the time. One of his key insights was that it was the mismatch between individual and current normative circumstances that accounted for the ill effects, not a pathology inherent to matriarchy.

In this chapter, I have detailed the similar dismantling of economic rewards since Moynihan wrote his report for a growing proportion of men based on education, along with the sizeable increase in women’s public participation and private power. Yet the assumed pathology of matriarchy persists in much of the US literature even as accumulating evidence indicates norms are giving way to more gender-egalitarian family arrangements. McLanahan (Reference McLanahan2004) gave it a less provocative name of “diverging destinies,” with the likelihood of father absence predicated on mothers’ education rather than race–ethnicity.Footnote ⁷

To be sure, studies from a range of Western countries confirm that parental separation and subsequent family transitions predict some risk of negative effects on children’s psychological well-being, behavior, grades, test scores, educational attainment, own early onset of sexual activity, early childbearing, and risk of divorce in adulthood (Garriga and Berta, Chapter 6; Härkönen, Bernardi, and Boertien Reference Härkönen, Bernardi and Boertien2017; McLanahan, Tach, and Schneider Reference McLanahan, Tach and Schneider2013; Perelli-Harris, Chapter 4). McLanahan (Reference McLanahan2004) contends the negative outcomes derive from the lower resources of single-parent households, in terms of both money and parental attention. As less-educated women are more likely to be single parents, their children face a larger resource deficit.

Confirming the causal direction of effects, however, is tricky (Autor et al. Reference Autor, Figlio, Karbownik, Roth and Wasserman2016; Perelli-Harris, Chapter 4). Lower socioeconomic status already predicts worse outcomes for children whether in two- or single-parent households. Another possibility is that some other characteristic might account for both family instability and children’s outcomes (Autor et al. Reference Autor, Figlio, Karbownik, Roth and Wasserman2016; Perelli-Harris, Chapter 4). Analyses controlling for children’s stable unobserved characteristics, prior behavior, or school performance indeed find smaller effects than when comparing across children (Härkönen, Bernardi, and Boertien Reference Härkönen, Bernardi and Boertien2017; McLanahan, Tach, and Schneider Reference McLanahan, Tach and Schneider2013). In other words, these children would have done worse regardless of family structure.

Furthermore, like patriarchy, the negative effects associated with spending time in a single-mother household are not inevitable. Most studies report “average” effects. In all countries, a sizeable minority if not majority of children experiencing family instability do just fine or perhaps better than had their fathers been present (Härkönen, Bernardi, and Boertien Reference Härkönen, Bernardi and Boertien2017). Furthermore, more redistributive welfare states and greater policy support for maternal employment minimize the negative intergenerational effects because they increase available resources (Härkönen, Bernardi, and Boertien Reference Härkönen, Bernardi and Boertien2017; Nieuwenhuis and Maldonado Reference Nieuwenhuis and Maldonado2018). In Europe, this realization resulted in the adoption of a policy discourse around social investment rather than social protection. Whereas traditional welfare policies aimed to reduce current family poverty, social investment policies aim to break the intergenerational cycle of poverty (Jenson Reference Jenson2009). The goal is to ensure current and future employment growth within a (skilled) knowledge economy (Bonoli Reference Bonoli2005). Patterned on the Nordic model, social investment policies stress greater education and skills for the next generation, simultaneous with current high levels of female labor force participation facilitated by more policy supports for employed parents and carers (Bonoli Reference Bonoli2005; Jenson Reference Jenson2009).

The impact of public investment on families’ economic resources is evident in the final three columns of Table 11.3. The third column indicates the percentage of GDP spent on family transfers in each of the countries (information not available for Latin America). Note the particularly low level of US public investment (0.7% of GDP in 2010) as compared with Europe and Canada. With this low level of public investment comes a high level of poverty even among US two-parent households – commensurate with the far less affluent Mediterranean countries struggling under austerity measures. The poverty rate for US single-mother households was a staggering 45.9% in 2010, exceeding that of any country in the table except Peru. In contrast, poverty rates of single-mother households in the redistributive Nordic countries were on average similar to the US poverty rate for two-parent households.

These aggregate differences manifest at the individual level. McLanahan, Tach, and Schneider’s (Reference McLanahan, Tach and Schneider2013) review consistently found significant negative effects of father absence on US children’s educational and mental health outcomes, but effects were often weaker or entirely absent in other countries. For example, the educational penalty for father absence is twice as large in the United States as in Germany or the United Kingdom (Bernardi and Boertien Reference Bernardi and Boertien2017a). What research is just beginning to untangle is how child outcomes differ systematically at the intersection of gender and class (and race–ethnicity).

Family Futures: Making Gender Equality a “Complete” Institution

In this chapter, I have highlighted how group inequalities configured at the intersections of family, market, and state institutions vary across place and evolve over time. My argument is that the institutional arrangements supporting patriarchy in the postwar decades have been crumbling for quite some time. A high-skill, technologically-driven global economy requires brains rather than brawn, and adaptable, socially engaged service providers. In Western societies, many of the requisite traits are traditionally feminine. Indeed, in the past half-century we have seen a remarkable ascendancy of women in the economic, social, and political order.

As the new institutional order unfolds, the value of education continues to increase (Autor Reference Autor2010; Gottschalk and Smeeding Reference Gottschalk and Smeeding1997; Machin Reference Machin, Gregg and Wadsworth2010). Education not only imparts skills, but encourages more egalitarian attitudes and predicts more positive social behaviors as well. Yet men’s educational attainment has not kept pace with women’s over the past few decades. Consequently a sizable proportion of men struggle to adapt to the new socioeconomic demands in and outside the home.

The gender revolution is far from complete, but many women perceive themselves as sufficiently independent to go it alone at some stage of raising their children when their partners fall short of economic or behavioral expectations. Although much more comparative research is needed, available evidence finds that boys’ essentialist behavioral problems magnify when fathers are absent from the household. These behavioral problems eclipse boys’ educational development, which blunts the possibility that a larger proportion of the next generation can enjoy the greater family stability associated with greater educational attainment.

This vicious circle of intergenerational inequality down the male line does not indicate a pathology of matriarchy as initially suggested by Moynihan in the mid-1960s. It instead point to the growing pathology of patriarchy in postindustrial economies, because better institutional supports for gender along with class equality yield the best intergenerational outcomes. The reason we have not yet eradicated the risks is because gender equality remains an “incomplete institution” even in the most progressive contexts. I borrow this term from Cherlin’s (Reference Cherlin1978) seminal article on remarriage after divorce. In that article, Cherlin argued that remarriages were less stable than first marriages because they lacked the institutional support in language, law, and custom that benefited first marriages. Similarly, I hold that the detrimental behaviors among boys and men will be staunched only once gender equality has become fully institutionalized in the family, market, and state. Institutionalizing gender equality eases the pressure on men to dominate paid work, allowing more adaptive masculinities to develop in which men’s equal contributions to both paid and unpaid work support family stability. I conclude with some brief thoughts on the major market and policy challenges to achieving this.

The first challenge is to enhance children’s and particularly boys’ educational engagement from preschool that will carry them through to complete higher levels of education. This perspective is core to the EU’s social investment strategy, but as someone who worked on educational reform in her pre-academic career, I can attest that the challenge is not the what, but the how. Most compulsory educational systems developed with the assumption of an at-home mother (Cooke Reference Cooke2011), and that children would adapt to the school structures and processes. Both of these assumptions undermine academic achievement.

Instead, educational processes need to adapt to children’s and families’ needs. This includes additional public funding for more aides of both genders in the classroom, innovative approaches to curriculum delivery, high-quality care and learning opportunities before and after standard school days, further supports for children with any type of special need (including behavioral), and coordinated extracurricular activities that do not require parents to shuttle children to and from venues. These supports should extend through adolescence, during which young persons are at greatest risk of becoming NEET – not in education, employment, or training (Eurofound 2016).

The second challenge is that both parents need more workplace flexibility to ensure they can be actively involved in their children’s daily lives. At present, organizations still reinforce patriarchal expectations of an ideal worker without competing family demands (Acker Reference Acker2006). These expectations manifest in disparate gendered penalties when employees seek workplace flexibility. For example, one US study found that male employees who experienced a family conflict received lower performance ratings and lower reward recommendations, whereas ratings of women were unaffected by family conflicts (Butler and Skattebo Reference Butler and Skattebo2004).

There is also a strong class dimension to organizational gendered expectations. Glass (Reference Glass2004) found that mothers in professional or managerial occupations incurred slightly larger wage penalties when they worked reduced hours or worked from home, as compared with mothers in other occupations who took up similar workplace options. Similarly, Brescoll, Glass, and Sedlovskaya (Reference Brescoll, Glass and Sedlovskaya2013) found that employers were more likely to grant low- than high-status men’s requests for flexible work schedules for family reasons. High-status men were more likely to be granted leave for career development (Brescoll, Glass, and Sedlovskaya Reference Brescoll, Glass and Sedlovskaya2013).

This nascent literature supports Goldscheider and Sassler’s optimism (see Chapter 9) that the second half of the gender revolution is now unfolding among less-educated couples, which should ultimately reduce educational gradients in marriage and divorce. At the same time, workplaces are stymying further gender equality progress among the more highly educated, beyond providing market alternatives for domestic production that widen class gaps among women (Gupta et al. Reference Gupta, Evertsson, Grunow, Nermo, Sayer, Treas and Drobnič2010). Persistent patriarchal advantage at the top of the class hierarchy blocks the thorough institutionalization of gender equality. One way to quell this is with more aggressive redistributive tax policies, the proceeds of which could be fed into the educational system.

Also needed are more aggressive positive discrimination policies, but targeted at the top of the occupational structure. Affirmative action entered US equality legislation with Johnson’s 1965 Executive Order 11246, although it has subsequently come under fire as discriminating against unprotected groups. Positive discrimination is also allowed under the European Commission’s 2006/54/EC directive on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation. Again, however, many countries have argued against positive discrimination because in principle it violates men’s rights to equality (Cooke Reference Cooke2011).

Gender equality at the executive level remains most elusive. The European Union is ahead of the Americas in tackling this with specific targets of increasing the percentage of women in key decision-making positions (European Commission 2016). In 2003, Norway mandated that 40% of nonexecutive board positions be filled by women. Although that controversial law is now considered a success, its implementation has not trickled down to increase the percentage of women in executive positions (Bertrand et al. Reference Bertrand, Black, Jensen and Lleras-Muney2015). Women’s ability to succeed when appointed to executive positions is contingent on eliminating the patriarchal organizational norms for executives as noted above.

The resistance to positive discrimination highlights the cultural resistance to fully disavowing patriarchy, a cultural resistance that has proven slow to change. Goldscheider and Sassler (see Chapter 9) discuss the gender essentialism behind such resistance, so I will not repeat the arguments here. However, as they also argue and as indicated throughout this chapter, both markets and policies can drive us further along the path to gender equality and, in turn, better family outcomes. Only when gender equality is fully institutionalized in markets and policies can the family, however configured, and all of its members, thrive.

The View from 10,000 Feet

Efforts to describe the family in comparative terms are a fraught enterprise, as they must account for cultural, economic, and legal changes in differing countries with diverse heritages. It is understandable therefore that the papers do not agree on an overall framework as to what exactly has caused changing family structures. Indeed, to the extent they agree on anything, it is most likely to be certain basic facts, and identification of the theories whose predictions cannot be validated. We therefore start with the factual assertions on which there is at least some agreement, then move on to the claims that do not stand up to examination, and close with the identification of the missing parts of a full analysis.

To the extent that there is a shared set of assumptions for this volume, they are basic. First, the family has changed. Between 1980 and 2000, fertility declined substantially across most of the developed world, though with greater variation after 2000. Nonmarital cohabitation and childbearing increased during the same period in every developed country. The patterns, however, are not uniform. Sweden and Iceland, which had much higher rates of nonmarital childbearing compared to other European countries in 1980, did not experience a sharp increase. The growth in nonmarital births is leveling off in both Sweden and Iceland (Perelli-Harris, Chapter 4, Figure 4.1).

Second, all developed countries experienced similar economic changes with a reduction in middle-wage jobs. The countries varied, however, in the degree to which they experienced a corresponding increase in higher or lower wage occupations (Cooke, Chapter 11, Table 11.1). France and Denmark, for example, experienced large increases in high-paying occupations, the Republic of Ireland and Finland saw their low-paying jobs expand most; while both high-paying and low paying jobs grew in the United Kingdom and the Netherlands.

Third, women’s roles have changed, with women increasing their workforce participation throughout Europe and North America since 1960, although those rates have plateaued in many countries (Cooke, Chapter 11, Table 11.2). The impact of these changes on family, however, varies across countries and regions. In countries with patriarchal gender attitudes, for example, the highest-earning women are less likely to marry than women with less education while in countries with more egalitarian gender attitudes, the highest-earning women have become more likely to marry than other women (Cooke, Chapter 11). In addition, access to employment and contraception has given women greater independence, but the form that independence takes varies considerably. Comparatively, more patriarchal countries such as Italy and Greece may have relatively low levels of nonmarital childbearing, for example, but also have substantially lower fertility rates.

Fourth, the increase in family instability has not affected the well off to the same extent as the middle-class and low-income families, particularly in the countries that have experienced the sharpest increases in nonmarital births and cohabitation. While the class-based divergences are not universal (Garriga and Berta, Chapter 6), such differences are widespread and there is a shared concern that family change may exacerbate economic inequality (Wilcox and Price, Chapter 8).

Despite agreement on the basic facts, there is little agreement on cause or effect. Indeed, the greater theoretical agreement may be on the failure of existing theories to account for what has occurred. Initial studies of the changing family have treated it as a process of cultural change, with elite women among the first to question patriarchal marriage and to embrace a redefinition of intimate relationships. In economics, Gary Becker predicted that low-income couples would experience the greatest increases in relationship instability as they forewent the benefits of “specialization” in the respective spheres of home and market.Footnote ¹ (Becker Reference Becker1981) Demographers described the changes in terms of a shift in values toward greater individualization and search for self-fulfillment, with higher educated people leading the way (Lesthaeghe Reference Lesthaeghe2010; Van de Kaa Reference Van de Kaa1987). The problem with these theories, however, is that they do not fit the evidence. In the United States in particular, elite families have seen little increase in nonmarital child-rearing and their families have experienced comparative stability (Lundberg, Pollack, and Stearns Reference Lundberg, Pollak and Stearns2016) at the same time working-class groups with more traditional gender attitudes have experienced the most dramatic increases in relationship instability (Cherlin, Chapter 3). Thus, the authors in this volume share skepticism about the existing theories, but none offers a single, comprehensive, consensus-based alternative account.

Developing such a theory is particularly challenging because, as we have argued elsewhere, it involves integrating the economic and normative changes. The interaction between the two is a dynamic process in which causality likely flows in multiple directions. For example, industrialization made education the new pathway into the middle class (economic); and the family and women’s roles within it changed to facilitate greater investment in children (normative) (Carbone and Cahn Reference Carbone and Cahn2014; Lesthaeghe Reference Lesthaeghe2010). The hallmark of the middle class thus became women’s ability to stay out of the paid labor market; a luxury beyond the reach of most of the working class until well into the twentieth century, except perhaps during the short-lived postwar economy of the 1950s.

Yet, the embrace of women’s distinctive role in overseeing the moral training and development of the young occurred readily in only some places. In the United States for example, farmwives embraced the new gender model before the urban working class; viewing it as an elevation in women’s status as they became the moral arbiters of family life (Carbone and Cahn Reference Carbone and Cahn2014). European scholars associate similar increases in parental investment in children with greater parental affection (Lesthaeghe Reference Lesthaeghe2010), and view rural families as lagging behind urban ones in embracing the new values (Scott and Tilly Reference Scott and Tilly1975). On both continents, the shifts in “sentiment” associated with expressive individualism, the move away from arranged marriages, greater female status, and greater investment in children unfolded over centuries (Stone Reference Stone1977, p. 198). In Europe, for example, women’s ability to devote themselves to the home came earlier in Britain, with only 9% of married women in the labor market at the turn of the twentieth century, in contrast to France, where 38% of married women remained in the labor market during the same time period (Scott and Tilly Reference Scott and Tilly1975). Cherlin observes in this volume that the American working class acquired the ability to keep wives and children out of the workplace and to invest more heavily in children’s education only after World War II (Cherlin, Chapter 3).

Today’s information economy has created a similar long-term transformation in the relationship between home and family. The new, postindustrial economy has generated greater demand for women’s market labor, making two-income families more important to middle-class status, and rewarding even greater investment in girls and boys. This has also required a reorganization of the family. To realize two incomes, college graduates embrace contraception and delay marriage and childbearing. When they do form families, they engage in a greater degree of assortative mating, with spouses choosing mates with similar interests and socioeconomic status. In managing children, the spouses trade off work force participation and child care, which requires a greater degree of trust and flexibility in managing relationships. We have previously termed this new system of family patterns as “blue” (Cahn and Carbone Reference Cahn and Carbone2010).

Each of these new systems – one developing during the rise of industrialization and the other coinciding with the information economy – combine adaptation to new economies with new moral understandings. Yet, the process of universalizing these new systems can be slower if the new values are contested, or if the society is unwilling to support economic policies that allow transmission of the new model to those left behind. Unemployed men do not necessarily contribute to the creation of more egalitarian parenting relationships, even if they adore their children. Moreover, community health mediates the impact of growing economic inequality on family stability, with the result that close-knit communities built around shared religious or cultural values and communities with more robust safety nets may not see as much of an increase in family dysfunction. The result – at least in the short term – is greater income and educational inequality.

We have written about these shifts over the last twenty years in the United States (Cahn and Carbone Reference Cahn and Carbone2010; Carbone Reference Carbone2000; Carbone and Cahn Reference Carbone and Cahn2014), and we have tried to capture a 10,000-foot view of the nature of the changes. Our story involves the intersection of economic change with family organization and increased class divisions. We have argued that the changes that took place with the rise of industrialization involved the same issues that occur today – greater economic inequality as opportunities increase for some while remaining beyond the reach of others, changing women’s roles that reallocate power in intimate relationships, and changing norms that become a source of tension in the recreation of culture. Cooke, with more of a European focus, describes the changes in similar terms (Cooke, Chapter 11). Underlying these changes has been the recreation of class advantage as the middle classes reorganize the family in order to realize the new opportunities – men’s entry into the management positions and professions of the industrial economy, and women’s similar expansion into the paid labor market today – while securing the investment in children necessary to realize these advantages (Lesthaeghe Reference Lesthaeghe2010). In the eighteenth and nineteenth centuries, this involved women’s increased status within the home and greater emphasis on child-rearing. In the twenty-first century, it involves shared parenting and greater reliance on paid child care (Cooke, Chapter 11; Goldscheider and Sassler, Chapter 9). Seeing the changes as part of a long-term, unevenly disseminating process provides the basis for more in-depth explorations of the integration of economics and normative change.

Today’s shifts – which involve investing in women as well as men’s income opportunities, embracing the birth control pill and postponing childbearing, high investment in children based on intensive male and female parenting – also face a difficult challenge to greater acceptance. The reasons, however, lie not with the overall ideal – more egalitarian family practices appear to have won the day in principle throughout the developed world – but rather with the difficulties of implementation. These challenges involve what we label the “nitty-gritty,” and they do help explain why analyses of the family remain so divisive.

The “Nitty–Gritty”

What we refer to here as “the nitty-gritty” involves the factors that explain how new normative systems spread, such as acceptance of new gendered roles. These factors are context-dependent – potentially varying substantially from rural Calabria to urban Stockholm for example. As explained below, the factors interact with each other in an iterative fashion. In short, they require a dynamic systems analysis, not just the isolation of individual causal agents. In this section, we identify a number of issues that complicate the analyses in this book and serve as important factors that explain differences in perspectives and outcomes. The factors involve men’s status, women’s sources of support for child-rearing, the legal treatment of intimate and child-rearing relationships, and the interaction between employment and migration in determining the composition of various communities.

“Why Can’t We All Get Along?”

The predominant view of the contributions to this volume is that we are in the midst of a profound family change. The family instability we are witnessing could simply be an issue of transition. For example, comparative studies show that the higher the amount of cohabitation in a society, the more cohabitants resemble married couples (Soons and Kalmijn Reference Soons and Kalmijn2009). A more universal embrace of the new system could therefore improve stability, as couples once again internalize similar expectations about their relationships (Cooke, Chapter 11; Goldscheider and Sassler, Chapter 9). Alternatively, a move toward more egalitarian gender relationships may also provide single parents greater freedom to raise children on their own and more incentive for the state to support them. Greater variety may therefore become a more permanent feature of family life, but variety does not have to be associated with instability. In the meantime, we should do everything to mitigate the negative consequences of the transition; and to cushion the negative impact of greater economic inequality on children’s life chances – an effort that other countries are undertaking (Boertien, Bernardi, and Härkönen, Chapter 7).

Three substantial obstacles complicate this process. First, there are cultural differences rooted in religion (Perelli-Harris, Chapter 4). The new egalitarian system rests on providing women substantial control of reproduction, partly through the availability of contraception and abortion. This postponement of family formation, and state support for the universalization of the means to do so, offends many religious teachings (Cahn and Carbone Reference Cahn and Carbone2010). Where religious opposition becomes entrenched, as which occurred in the United States to systematic sex education and provision of contraception, or in the Republic of Ireland to divorce and abortion, the result tends to be the exacerbation of class and regional differences. In the case of abortion for example, the Irish elite evade the religiously based restrictions through travel abroad while the poor are subject to them (Aiken, Gomperts, and Trussell Reference Aiken, Gomperts and Trussell2016). In the United States, the class-based differences in unintended pregnancies grew substantially between the early 1990s and 2009 at the same time that abortion rates fell for all groups (Carbone and Cahn Reference Carbone and Cahn2014). The changes in unintended pregnancy rates correlated with the increasing class divergence in family formation practices, particularly among Whites and Latinos. Thus, while Reeves’ proposals in this volume (see Chapter 10) for more universal access to contraception make eminent sense and have already produced impressive results in Colorado and many European countries in preventing early births, we believe universal adoption of such policies will be gradual. Indeed, the 2017 Trump Administration proposals would further undermine access to contraception, not just abortion, in the name of religious liberty.

The second obstacle increasing inequality pertains to the same practices lacking the same meanings throughout society. College graduate women, for example, have embraced an ethic that they should not have children with a partner they do not trust. In the United States, higher income women have relatively low levels of unintended pregnancies, and abort a higher percentage of unintended pregnancies than any other group (Guttmacher 2016). Their willingness to marry may in turn reflect acceptance of the egalitarian custody norms marriage now imposes, which include shared parenting and equal custody rights. In contrast, working-class women are more likely to have a child as a result of an unintended pregnancy or in circumstances where they view few available men as worthy of trust (Barber et al. Reference Barber, Kusunoki, Gatny and Melendez2017; Burton et al. Reference Burton, Cherlin, Winn, Estacion and Taylor2009). In these circumstances, the egalitarian custody norms applicable at divorce may be inappropriate (Carbone and Cahn Reference Carbone and Cahn2013). For example, an American study of divorcing couples found that the award of increased rates of shared parenting time correlated with increased domestic violence complaints, holding the other factors as constant (Brinig Reference Brinig2017). The author speculates “should the same logic hold true for unmarried parents, who as noted experience more domestic violence in their relationships, the concerns about insisting upon parenting orders for them at the time support is established would be justified” (Brinig Reference Brinig2017).

The third factor involves the lack of shared ways of discussing the cultural changes. Most studies find that greater family stability of all kinds, including more stable employment and income, fewer residential moves, and more stable household composition, benefits children. Further, while groups differ on the degree to which economic vs. cultural factors influence such stability, they agree that individual decisions play a role in family outcomes. Yet, different cultural groups fundamentally differ not just on the content, but the sources of individual responsibility and moral values. Modernist societies (which we have labeled “blue”) differ from traditionalist ones (red) in the way they allocate discussions of morality to the public and private spheres. In modernist societies, the public virtues are equality and tolerance; notions of family form, consensual sexual behavior, and appropriate child-rearing practices are matters of private choice. More traditionalist societies insist on the importance of upholding shared values (such as childbearing within marriage) in the public square in order to reinforce the right values at home (Carbone 2017). Blue societies thus distrust a public emphasis on marriage per se, either because they associate with it with an older form of hierarchical gender roles or because they see it as a substitute for a process of individual selection of the right partner and the right values. Instead, they see family stability as coming from relationships premised on flexibility and trust. This in turn place much more emphasis on parental guidance in raising children who develop the individual moral codes, which they internalize as central to their personal integrity and which in turn make them trustworthy. These individual codes can vary, but responsible adulthood and personal self-respect depend on having one that orders a person’s adult commitments and life choices, and that informs selection of an appropriate partner. Central to this process is avoiding childbearing until one is fully capable of assuming the responsibilities that comes with it in part because the system’s success depends on a substantial degree of parental investment in children. Within this system, fully mature adults with well-developed personal codes do not need marriage; they would largely behave the same way without it. Also, without development of the underlying individual codes, marriage in an era in which it rests on flexibility and trust is unlikely to succeed. Indeed, within the United States, red states tend to have higher teen birth and divorce rates (Cahn and Carbone Reference Cahn and Carbone2010; Glass and Levchak Reference Glass and Levchak2014).

Traditionalist societies often criticize the modernist approach as license. They associate tolerance in the public square with irresponsibility within the family. They prefer systems that seek to instill universal values that come from religious traditions or shared cultural norms. Historically, this system sought to insure family stability through marriage soon after completion of one’s education, and socialization into adulthood through the assumption of gendered family roles (Cherlin Reference Cherlin2005). Yet, this system also produced stability in part because of women’s dependence. Today, such an approach works best with two parents who share the same traditionalist commitments, especially when the couple resides in a community that reinforces these values.

These two systems talk past each other, about both the source of moral values and the way to promote them. Traditionalists see moral failings as a product of insufficient public affirmation and private acceptance of responsibility; modernists see them as a product of the failure to create the conditions that promote individual flourishing. With greater inequality undermining both public support for family well-being and individual ability to live up to traditional precepts, the cultural clash between these views intensifies.

As a result, there are no easy policy prescriptions for increasing family stability because the interaction between economic changes and cultural norms is multi-causal, dynamic, and interactive (Cooke, Chapter 11), and there is no shared set of assumption for discussing the issues – other than an agreement that family stability is good for children. The evidence indicates that a narrow focus on family form, while it may produce benefits for relatively homogenous groups, does not work as a more universal public policy prescription. For example, in the United States, studies show that where both spouses are religious and attend the same church, divorce rates are low even if the spouses marry young. Other studies indicate, however, that in more religious communities in the United States, those who attend church less often have higher divorce rates than comparable couples in less religious communities. This is true in part because the religious practice of younger marriage in these communities tends to lower the average age of marriage for everyone, and marriage at younger ages carries with it a higher risk of divorce (Glass and Levchak Reference Glass and Levchak2014). While the selection effects make the picture more complicated, encouraging a return to marriage as the sole locale for child-rearing, for example, does not solve the problem of economic instability, which tends to weaken the resilience of most communities. In the United States, race tempers the financial benefits of marriage; as the St. Louis Federal Reserve reported, “when we focus on family-structure differences within racial or ethnic groups, rather than between groups, there is essentially no relationship at all” between family structure and wealth (Emmons & Rickett Reference Emmons and Ricketts2017). A cross-cultural study similarly found that differences in family structure have virtually no impact on children’s educational attainment, after controlling for other factors (Bernardi and Boertien 2015). As Boertien, Bernardi, and Härkönen ask, in Chapter 7: “Does the result that family structure can explain little of socioeconomic background differences in educational attainment imply that family structure does not matter for socioeconomic inequality of opportunity in general?” They conclude that we simply do not know. The existing evidence on the relationship between family structure and educational attainment is simply too limited, with a small set of countries and time periods (Boertien, Bernardi, and Härkönen, Chapter 7).

Even if marriage were at least a partial solution, marriage promotion programs do not “affect marriage or poverty rates” for low-income couples (Randles Reference Randles2017, p. 14), particularly in the absence of societal efforts to address the economic instability that correlates with unstable relationships (Cherlin, Chapter 3). Moreover, growing up in a single-parent family home has different effects on children’s outcomes, depending on the country (Garriga and Berta, Chapter 6), or even the number of single-parent families in the community (Soons and Kalmijn Reference Soons and Kalmijn2009). Interventions strengthening children’s well-being may be more effective than marriage promotion per se (Reeves, Chapter 10), and policies that support reproductive rights, greater access to health care, and improved workers’ rights (such as a higher minimum wage) are also associated with family stability (Robbins and Fremstad Reference 316Robbins and Fremstad2016).

By presenting the complexity of variations between and within countries, this volume shows that generalizations about the impact of family structure do not work. Nonetheless, the weight of the evidence in this volume suggests that economic change is producing both family change and greater inequality, at different rates, in different forms across different societies. Cultural factors in these societies may, in turn aggravate or ameliorate the effects on family well-being; so too may societal interventions that cushion or worsen the consequences of greater economic inequality. All of these societies are moving toward at least slightly more egalitarian gender relationships as women’s workforce participation has increased. Moreover, all societies are finding that it is difficult to maintain traditional understandings as a universal basis for family relationships.  We conclude that, while there is much we still do not know about the interactions between changing families and changing economies, future family stability and the marshaling of resources necessary for children’s well-being will require acceptance of at least a degree of family change and also a deeper integration of egalitarian relationships into our understanding of family function.