The Growth of Female Labor Force Participation

There is a substantial literature on factors favoring the growth of female labor force participation. In addition to the overwhelming effect of industrialization (see the review in Pampel and Tanaka (Reference Pampel and Tanaka1986), in which they show that the relationship is actually curvilinear, as early increases in energy use actually forced women out of the labor force), the major classic studies have emphasized the importance of the growth in the demand for “female labor” beginning in the post-World War II years and accelerating in the later decades of the twentieth century (Goldin Reference Goldin1990; Mincer Reference Mincer and Lewis1962; Oppenheimer Reference Oppenheimer1970). Continued industrialization resulted in the emergence of occupations quite different from agricultural and early industrial jobs, which required both substantial physical strength and long hours. More recent research, however, has also emphasized the importance of the long-term rise in female education (Walters Reference Walters1984), which began in the late nineteenth century, and led eventually to women surpassing men in the attainment of college degrees in the last few decades of the twentieth century (DiPrete and Buchman Reference DiPrete and Buchman2013). At least initially, in the late nineteenth and the first half of the twentieth century, this did not reflect reverse causality, in that the pursuit of educational attainment was justified as necessary for becoming a successful wife, mother, and homemaker (Walters Reference Walters1984). Not until the growth in female labor force participation in the second half of the twentieth century did female education increase in response to the growth of economic opportunity.

The demographic transition from high to low fertility and mortality was also important, as it made it no longer feasible to define all of women’s adult years as needed for childbearing and rearing (Stanfors and Goldscheider Reference Stanfors and Goldscheider2017; Watkins, Menken, and Bongaarts 1989). Increased earnings no doubt mattered as well, although analyses of the negative effects on earnings of women’s concentration in “pink-collar” occupations (Cotter et al. Reference Cotter, DeFiore, Hermsen, Kowalewski and Vanneman1997) suggest that women were eager to earn anything that could contribute more to their families than they were able to do with just their domestic skills and time. The crumbling of the “marriage bar” (Goldin Reference Goldin1990; Stanfors and Goldscheider Reference Stanfors and Goldscheider2017), which for so long prevented married women from taking many jobs, such as teaching, contributed as well.

The growth of female labor force participation emerged in response to these interrelated economic, social, and demographic changes primarily over the second half of the twentieth century. While women’s employment has not reached parity with the labor force participation of men, it has reached the point where most women expect to spend much of their adult lives in employment (and most men expect their partners to do so as well). This creates an additional incentive for women to work, and to obtain more education in order to prepare for better jobs (e.g., those with more pay and better working conditions). It has also altered men’s expectations of desirable attributes in a spouse; if their wives will be working, one who can earn more is preferred to one who is unlikely to earn much (Goldscheider and Kaufman Reference Goldscheider and Kaufman2006; Schwartz and Han Reference Schwartz and Han2014).

Clearly, these major changes in the structure of industrializing societies have been critical for increasing female labor force participation, and hence, are also likely to affect men’s increasing involvement in the family, about which much less is known. However, if, as Cherlin and others argue, supportive family policies and socioeconomic inequality are also critical for encouraging or delaying the second half of the gender revolution, and perhaps also the first half, it is important to assess these impacts. How might social policies and greater socioeconomic inequality affect women’s labor force participation, the first half of the gender revolution? We address these issues in the following sections, and then address the same issues for the second half of the gender revolution.

Family Policy and the First Half of the Gender Revolution

Until recently, few governments outside the (former) Soviet sphere of influence enacted policies explicitly to support women’s employment.Footnote ¹ Nevertheless, some policies developed for other purposes often affected women’s likelihood of being employed, both positively and negatively. The most common direct policy involves the development of quality, subsidized child care, allowing women to outsource some of their family responsibilities (without having to involve their male partners). The United States developed an extensive network of child-care facilities during World War II, as women were needed to replace the men who had been conscripted (Skocpol Reference Skocpol1995), although it was dismantled after the war ended.

Some other European countries encouraged female employment more permanently. The Scandinavian countries’ initial response to the increased demand for labor in the post-World War II period was to encourage employment for mothers, going beyond the provision of subsidized, high-quality child care to provide up to a year of paid family leave, which allowed new mothers to maintain their connection with their employers while being able to care for their tiny infants. These countries preferred to meet the need for a growing labor force by supporting female employment rather than encouraging the immigration of male workers (while making it more likely that women would stay home), which was the policy choice taken by Germany in this period (Kamerman and Kahn Reference Kamerman and Kahn1991).

Policies with indirect effects on female labor force participation have primarily been those designed to increase fertility, although there are others, such as retirement plans and family-level vs. individual-level taxation, that have strong effects as well. As early as the 1930s, some countries experiencing dramatic fertility decline linked with the Great Depression, such as France, introduced baby bonuses, hoping to encourage more couples to take on the costs of childbearing. Others instituted family allowances and paid maternity leave, so that rather than the money being tied to childbearing, it was tied to (women’s) child-rearing, often for an extended period, as in Germany.

In each case, the pro-fertility policies were designed to support a traditional division of labor in families, in which women stayed home, and could afford to stay home, to care for their families. Interestingly, there was little evidence that either of these policies was successful (Friedlander and Goldscheider Reference Friedlander and Goldscheider1979).Footnote ^2, Footnote ³ Research examining the relationship between length of parental leave and women’s subsequent career progression has generally found that more generous leave hampers longer-term earnings (e.g., Aisenbrey, Evertsson, and Grunow Reference Aisenbrey, Evertsson and Grunow2009).

Policies designed to reinforce mothers’ ability to combine work and family might have a stronger effect on female employment, particularly by introducing subsidized, high-quality child care and job-protected, paid maternity leave. Although these policies tend to reinforce women’s family roles by making the “second shift” less onerous, they also encourage the attainment of well-paid jobs, as the payments received during leave are normally tied to prebirth earnings. The need to attain a well-paid job before taking family leave also greatly reduces teen childbearing; job protection means a parent need not search for work after taking leave, which hastens the return to work even for the United States’ unpaid Family and Medical Leave Act (Klerman and Leibowitz Reference Klerman and Leibowitz1999).

Evidence that the provision of high-quality, subsidized child care supports female labor force participation also comes from a recent study comparing German-speaking residents in Belgium with those in an adjacent area in Germany (Klüsener, Neels, and Kreyenfeld Reference Klüsener, Neels and Kreyenfeld2013). These two regions’ family patterns closely resembled each other prior to Belgium’s introduction of family-friendly policies; thereafter, these two regions diverged in terms of both fertility (German fertility continued to fall while that of German-speaking Belgians did not because the German-speaking region of Belgium had much lower childlessness and higher completed family size than in Germany) and especially female labor force participation. By 2001, the proportion of mothers with children aged 0–2 in full- or part-time employment was only 30.1% in western Germany, compared with 60.5% of similar mothers in the German-speaking region of Belgium (Klüsener, Neels, and Kreyenfeld Reference Klüsener, Neels and Kreyenfeld2013). This is a clear case in which family-friendly policies not only support higher fertility but also, even more clearly, increase female labor force participation.

However, as we will discuss below in our consideration of the effects of policy on the second half of the gender revolution, policies that encourage men to share more of the family leave appear to have been even more successful in incorporating men into the care of their homes and families. Whether or not such policies are necessary for increasing men’s involvement in the care of their homes and families, as we will see, they clearly help.

Inequality and the First Half of the Gender Revolution

Although the relationship between family policy and female labor force participation is fairly clear, that between socioeconomic inequality and female labor force participation is much less so. The primary studies have focused either on macro-level analyses of countries’ levels of female labor force participation, mostly from the 1980s, or on more microstudies of how changes in inequality focused on how husbands affect married women’s response. More problematic (and interesting) is the fact that given a rapid change in behavior, there is also a substantial change in the relationship between women’s behaviors in the family and economic context.

The first major macrostudy (Semyonov Reference Semyonov1980) undertook a comparative analysis of the social context of women’s labor force participation. The study used data from sixty-one countries for the 1960s and 1970s and found that women were less likely to be economically active in societies where inequality was high. This was followed with an analysis of a much narrower set of countries (sixteen) but which included longitudinal data for five time points between 1955 and 1975 (Ward and Pampel Reference Ward and Pampel1985). In contrast to the broader analysis, however, this study found that increases in inequality increased female labor force participation. There were a number of other differences between the two studies, including the controls applied, however. It is not clear whether further studies of this question have been undertaken.

There are more, and more recent, analyses, primarily by economists, which examine how the labor force participation of married women responds to their husbands’ earnings, motivated by the rapid increase in inequality in men’s earnings and the initially high level of women’s responsiveness to their family’s economic well-being. Early studies of female labor force participation showed that family characteristics had a larger impact on married women’s employment than the women’s own characteristics. Small children were a major deterrent to married women’s employment, as was husband’s income; women were much more likely to work if they had no small children and their husbands’ incomes were low (Blau and Kahn Reference Blau and Kahn2007). Hence, increases in inequality of male incomes have contributed to the increase in female labor force participation, as the growth-in-inequality losers (the number of men with low incomes) is always far greater than the growth-in-inequality winners (the number of men at the top of the earnings pyramid). The relationship between inequality in men’s earnings and female labor force participation, however, is not strong, because while declines in male employment and earnings have been greatest for low-wage men, employment and earnings gains have been largest for wives of middle- and high-wage men (Juhn and Murphy Reference Juhn and Murphy1997). This is because over the last few decades of the twentieth century, American women’s labor force participation became much more responsive to their own characteristics (education, work experience) than they had been in the past (Leibowitz and Klerman Reference Leibowitz and Klerman1995), so that in this way they became more like men. Further, educational homogamy has strengthened, so that inequality at the family level increased, as highly educated men and women married and both earned, while couples with few resources struggled (Karoly and Burtless Reference Karoly and Burtless1995).

The Growth of Male Family Involvement

Any attempt to understand the growth of male involvement in the tasks of their families and homes must first recognize that not that long ago, men were highly involved in such tasks: Training their children and providing wood, water, and far more to their homes in the household-based agricultural economy. With industrialization came the construct of “separate spheres” as men left agriculture and took on the industrial and commercial jobs that were emerging in the new economy, leaving women behind in the home. It took more than a century until the social, economic, and demographic changes underway reached the point where women could join men in the public sphere as well. By then, the home had taken on its gendered nature, as a place for women and children (Stanfors and Goldscheider Reference Stanfors and Goldscheider2017), making it more challenging for men to include domestic tasks as part of their responsibilities.

Men resist not only the gendered tasks in the home, which causes a problem in their relationships, but also the gender-linked tasks in the labor market, which is a problem for their role as provider, given that, with the decline in agriculture, the decline in manufacturing has meant that the jobs “real men” used to take have been vanishing (see Chapter 3). We know much less about the factors that increase men’s involvement in the home, and hence the progress of the second half of the gender revolution, than we do about the growth of women’s labor force participation, but it is important to examine what we do know, particularly how it might be linked with family policy and socioeconomic inequality, as Cherlin worries (see Chapter 3).

Family Policy and the Second Half of the Gender Revolution

Many of the Scandinavian countries (and Quebec) have developed policies to address the second half of the gender revolution directly, by encouraging men to take a portion of what is now called “family” (rather than “maternity”) leave (Duvander and Johansson Reference Duvander and Johansson2015). Such policies require that, to gain, and not lose, the full benefit, each parent has to take some of the paid leave. If only one parent takes family leave, the duration is less than if each parent takes some, with lost leave ranging from one to three months. As a result, nearly all couples take at least the minimum required for the second parent (often called “daddy days,” given that women continue to take most family leave). In Norway, 89% of fathers take most or all of the daddy days, with similar levels in Iceland (88%) and Sweden (86%) (Haas and Rostgaard Reference Haas and Rostgaard2011). Evidently, even managers in private industry do not want their workers to actually lose a paid benefit.

A major problem with assessing the impact of policies on citizens’ behavior is whether the causal arrow is being correctly interpreted. Perhaps only more egalitarian countries enact egalitarian policies. Alternatively, men who take advantage of such policies may be selective, in that they are just more family-oriented than other men. In Sweden, for example, the fact that men who take more family leave with a first child are more likely to go on to have additional children than those who take less (Oláh Reference Oláh2003) might just reflect their greater family orientation. However, a recent study addressed this latter issue. Duvander (Reference Duvander2014) has shown that although Swedish men with more familistic attitudes do indeed take more family leave, this is also the case for men with more egalitarian attitudes (controlling for their attitudes toward families).

The question of path dependency – that more egalitarian countries are more likely to enact egalitarian policies – is challenged by a recent analysis of trends in men’s home involvement in the United States compared with the Scandinavian countries. Despite (or perhaps because of) the lack of policies reducing work–family conflict, the United States’ levels and trends in men’s share of housework and child-care time closely approximate those in Scandinavia (Stanfors and Goldscheider Reference Stanfors and Goldscheider2017). This is despite the fact that the ratio of female to male labor force participation is considerably lower in the United States than in the Scandinavian countries – about 85% in 2014 compared with ratios closer to 95% in Norway, Sweden, and Finland – reflecting the lack of public policies, such as child care, that support female labor force participation (Frejka, Goldscheider, and Lappegård Reference Frejka, Goldscheider and Lappegård2018). This suggests that more US families are forced to depend on male involvement (perhaps coupled with shift work) to care for children while both parents are employed than in countries with policies reducing work–family conflict.

Another series of studies that attempt to hold cultural values (and hence path dependency) constant but vary potential family policies are “experiments” done in the United States (e.g., Pedulla and Thébaud Reference Pedulla and Thébaud2015; Thébaud and Pedulla Reference Thébaud and Pedulla2016). Respondents were young, unmarried, and childless, and were asked about their future work–family arrangements (essentially, who should be the primary earner and who the primary homemaker). In one study (Pedulla and Thébaud Reference Pedulla and Thébaud2015), some respondents were randomly provided a scenario that asked them to imagine that there were policies in place that included paid family leave, subsidized child care, and the option to work flexible hours, and then asked about their preferred work–family balance. The results were dramatic, at least for women. More highly educated women (those with some college education or higher) who were told these policies would be virtually universal chose a work–family balance of equal sharing (94.5%) compared with similar women who did not receive this information (more than 30 percentage points less). Among less-educated women, the differences were also sharp, but smaller (82.2%, given the presence of these family-friendly policies; over 20 percentage points more than similar women who could not assume the presence of such policies).

Interestingly, men were much less responsive to the presence or absence of family-friendly policies (and differences were not significant). This finding suggests that men who had not confronted the possibility of work–family conflict had not really thought about it. Nevertheless, a substantial majority of men – both those with some college or more and those with high school or lower – gave egalitarian responses (about 70%).

A second study (Thébaud and Pedulla Reference Thébaud and Pedulla2016) explored this gender difference, and established that men’s preferred work–family arrangement was conditioned by their understanding of their peers’ attitudes, not policy. The results showed that in scenarios where supportive work–family policies were available, men who were asked to assume that most of their peers preferred gender-egalitarian relationships were significantly more likely to prefer such arrangements for themselves. The researchers did not test to see whether similar results would have obtained if men had partners who did or did not prefer egalitarian relationships, or whether the families they had grown up in featured father involvement in the domestic sphere, which also has a strong impact on men’s domestic participation (Lahne and Wenne Reference Lahne and Wenne2012).

Inequality and the Second Half of the Gender Revolution

There is no question that in many parts of the industrialized world there are sharp differences in domestic sharing by socioeconomic status (Evertsson et al. Reference Evertsson, England and Mooi-Reci2009), as there are in other family patterns. In the United States, socioeconomic differences in family patterns seem to reflect nearly different universes (Lundberg and Pollak Reference Lundberg and Pollak2015), as Cherlin has dissected in his most recent book (Cherlin Reference Cherlin2016), and as do Sassler and Miller (Reference Sassler and Miller2017). The college-educated are currently far more likely to get and remain married, have children within marriage, and yet experience relatively egalitarian work–family relationships than those with educational levels of high school or lower, shaping what Sara McLanahan (Reference McLanahan2004) has termed the “diverging destinies” of young Americans from more and less advantaged backgrounds. Cherlin (Reference Cherlin2014) therefore argues that the second half of the gender revolution may only be emerging among the more educated (and perhaps most rapidly in a context of strong state support for families).

The issues revealed in recent research suggest that the patterns are extremely complex, which makes tracking class differences particularly challenging. Most fundamentally, the “private sphere” of family tasks is heterogeneous (like the public sphere), with the biggest divides between child care and housework, and within those two categories, between routine and less routine activities (Craig and Mullan Reference Craig and Mullan2013) and between taking responsibility and implementation (Raley, Bianchi, and Wang Reference Raley, Bianchi and Wang2012). As we will show, men’s entry into this complexity has varied over time and by class.

For example, men’s time in child care and housework has been increasing more rapidly on weekends rather than weekdays (Neilsson and Stanfors Reference Neilsson and Stanfors2014), a pattern that favors men with regular hours. Co-resident men are more involved than those who do not live with their children, which also favors the more educated, who are less likely to separate from their children, in part because their behaviors often undermine relationship stability as well as consistent child involvement (Barber et al. Reference Barber, Kusunoki, Gatny and Melendez2017; Cherlin, Chapter 3; Edin and Nelson Reference Edin and Nelson2013). Nevertheless, many non-co-residential fathers remain involved (Tach, Mincy, and Edin Reference Tach, Mincy and Edin2010).

Further, context appears to matter as well. Far more research is needed to disentangle these issues, but here we will focus on the effects of class (normally indicated by her education, but sometimes by his) on housework and child care, the second half of the gender revolution. We focus primarily on class differences and changes in child care. Child care is more consequential than housework, as perhaps the most fundamental domestic investment activity. Housework is both less problematic to ignore (beyond the most basic level) and easier to outsource. There is also considerable variation in how housework is measured over time, and too few studies have empirically explored class differences in the ways that dual career couples juggle family responsibilities.

The general finding is that fathers with more education are more involved with child care than fathers with less education. In the United States, more educated fathers have particularly increased their involvement in basic caregiving and teaching time, and have increased their belief in the importance of fathers in children’s lives (Amato, Meyers, and Emery Reference Amato, Meyers and Emery2009; Hofferth et al. Reference Hofferth, Pleck, Goldscheider, Curtin, Hrapezynski, Cabrera and Tamis-LeMonda2014; Raley, Bianchi, and Wang Reference Raley, Bianchi and Wang2012; Sayer, Bianchi, and Robinson Reference Sayer, Bianchi and Robinson2004). These findings are similar to those from a cross-national study of four countries (Australia, Denmark, France, and Italy) (Craig and Mullen Reference Craig and Mullan2013), that found, as well, that it was fathers’ education rather than mothers’ that mattered.

Nevertheless, a growing body of both quantitative and qualitative research is documenting that less-educated men are also viewing child care involvement as a central role (Carlson, Hanson, and Fitzroy Reference Carlson, Hanson and Fitzroy2016), but may perform their engagement differently than do fathers who are professionals (Edin and Nelson Reference Edin and Nelson2013; Shows and Gerstel Reference Shows and Gerstel2009). Some research suggests that although men in the working/service classes express strongly gender-differentiated views, they often find themselves actually living quite egalitarian lives (Usdansky Reference Usdansky2011). This finding reinforces a concern that such men might understate their household contributions.

The “Family Gap”

Other contributors have detailed key trends in family life, especially the rise in nonmarital births, single parenthood, and increased relationship turnover. In the United States, four in ten children are born to unmarried parents; among women under 30, the number is closer to half. In most cases the mother is cohabiting; but only a minority of cohabitees are still with their partner by the time the child reaches five (McLanahan and Sawhill Reference McLanahan and Sawhill2015). Two thirds have split up before their child reaches age 12, compared with a quarter of married parents (Kennedy and Bumpass Reference Kennedy and Bumpass2008).

The “stability gap” between cohabiting and married couples can be seen in all countries and at every level of education (DeRose et al. Reference DeRose, Lyons-Amos, Wilcox and Huarcaya2017). In Norway, for example, children born to highly educated cohabiting mothers are twice as likely to see their parents’ relationship end before the age of 12 as those born to highly educated married mothers (17 percent vs. 8 percent).

While there is a stability gap between married and cohabiting couples, there is also stability gap by social background. Rates of lone parenthood vary widely by education, race, and income. College graduates are unlikely to become single parents, compared to those with less education (Reeves Reference Reeves2017). Six in ten black children are in a single-parent family, almost double the rate for white and Hispanic children (Reeves and Rodrigue Reference Reeves and Rodrigue2015a). The “retreat from marriage” has been much more rapid among less-educated, poorer Americans (Reeves Reference Reeves2014b).

Indeed, on every measure of family stability, including rates of unintended pregnancies and births, marriage, single parenthood, and divorce, there are wide, and in many cases, widening, gaps by education, income, and background. Trends in poverty, income inequality, or intergenerational mobility cannot be properly examined in isolation from trends in family life.

Supporting Family Planning

The vast majority of Americans become sexually active before marriage: To be precise, 95 percent compared to 80 percent in the 1970s. One in three has had at least six sexual partners before marrying (Daugherty and Copen Reference Daugherty and Copen2016). Social norms on this issue have evolved even more quickly than behavior. Most Americans aged between 15 and 35 (58 percent of women and 68 percent of men) agree with the following statement: “It is all right for unmarried 18 year olds to have sexual intercourse if they have strong affection for each other” (Daugherty and Copen Reference Daugherty and Copen2016).

Sex (between a man and woman) usually holds the promise – or threat – of resulting in pregnancy. Of course, effective contraception can prevent this, but not enough couples are being “planful,” to use Paula England’s phrase, about their fertility (England et al. Reference England, Caudillo, Littlejohn, Bass and Reed2016). There are now highly effective, convenient forms of contraception available, known in wonky circles as LARCs (long-acting reversible contraceptives), but only a minority use them. Over a five-year period, among those relying solely on condoms for contraception, 63 percent will get pregnant. For women using the best IUDs, the rate is just 1 percent (Reeves Reference Reeves2016), but progress toward greater promotion of LARCs among policymakers and health professionals has been slow. As a result, among sexually active women aged 15 to 24, just 10 percent use a LARC method of contraception. By comparison, around 20 percent of women in their early 20s report using an illegal drug in the previous month (US Department of Health and Human Services 2014). When there are twice as many young women using an illicit drug as using an effective contraceptive, we can be sure there is room for improvement.

We care because unintended childbearing is significantly associated with poorer child outcomes (Haskins, Sawhill, and McLanahan Reference Haskins, Sawhill and McLanahan2015; Sawhill Reference Sawhill2015). It is hard to read Robert Putnam’s book Our Kids and not notice the recurring pattern of chaotic starts to parenthood (Putnam Reference Putnam2015). Darleen got pregnant two months into a relationship with Joe, her boss at Pizza Hut. “It didn’t mean to happen,” she reports. “It just did. It was planned and kind of not planned.” David, an 18-year-old in Port Clinton, Ohio (Putnam’s home town), becomes a father. “It wasn’t planned,” David says. “It just kind of happened.”

Given the rapid liberalization of social norms regarding sex but slow take-up of effective contraception, it should be no surprise to learn that 50 percent of pregnancies are unintended (Coleman and Garratt Reference Coleman and Garratt2016), or that 60 percent of births to single women under age 30 are unplanned (Sawhill Reference Sawhill2014). The high rate of unintended pregnancies and births, especially among women in their 20s, has serious implications for poverty, inequality, public spending, housing, and health care provision. There is a strong class dimension to this story too; women from affluent families are more likely to use contraception, much more likely to use the most effective kinds, and very much less likely to have an unintended pregnancy or birth (Reeves and Venator Reference Reeves and Venator2015). One reason for this is that young women from affluent families are more likely to be in regular contact with health providers, who often prescribe contraception for other health issues such as menstrual cramps. Although hard to pinpoint empirically, there is no doubt that unintended pregnancy and childbearing – what Sawhill and Venator (Reference Sawhill and Venator2015) call “drifting” into parenthood – is a very big factor influencing family stability. When a couple become parents by mistake, which is true of most cohabiting couples, it is little surprise that their relationship is less likely to endure than those who plan, as seen in Figure 10.3.

Figure 10.3 Unintended pregnancy rates are much higher among unmarried couples

Ensuring access to affordable, effective contraception is the most powerful pro-family policy available. For a start, this means ensuring equal access to family planning services. Even if the Affordable Care Act (ACA) is repealed in the United States (which seems highly uncertain at the time of going to press) we must hope legislators see the value of this element of the plan. If not, geographic inequalities within the United States may deepen as many states are likely to bolster their own efforts in this area.

It is worth noting here that if all states had implemented Medicaid expansion under the ACA – at a cost to the federal government of around $952 billion over ten years – millions more low-income women would have been able to access family planning services more easily (Holahan et al. Reference Holahan, Buettgens, Carroll and Dorn2012; Ranji, Bair, and Salganicoff Reference Ranji, Bair and Salganicoff2016). (Vice President Mike Pence, as Governor of Indiana, was one of ten Republicans accepting Medicaid expansion under the ACA.)

There is also a need for greater awareness-raising and training among health professionals. Indeed, staff training alone seems to have a significant impact on the take-up of LARCs, according to a randomized control trial. The work of organizations like UpStream training providers in states including Ohio, New York, Texas, and Delaware is extremely promising.Footnote ¹ Other steps can be taken to broaden access, including ensuring sufficient supplies in health clinics, simplifying billing procedures, and providing same-day service.

What is most needed here is a cultural shift, specifically with regard to responsibility around childbearing. Premarital sex among adults is now the norm, both in terms of attitudes and behavior. Nine in ten US adults under 44 now have sex before they marry (NSFG 2016). What is needed is a cultural shift toward more planfulness in childbearing, as Sawhill (Reference Sawhill2014) argues:

I think this is exactly right. The key is to ensure that the liberalization of attitudes toward sex does not lead to a liberalization of attitudes toward the moral responsibility to plan when, how, and with whom we bring children into the world. Casual sex may be fine; casual childbearing is not.

Here the popular media is likely to be much more important than policymakers or pundits. As work by Kearney and Levine (Reference Kearney and Levine2015) shows, MTV’s 16 and Pregnant had a measurable impact on teen pregnancy rates.

There are, nonetheless, ways policymakers can nudge cultural change along toward more responsible parenting, and use of effective contraception. Sawhill and Venator (Reference Sawhill, Venator, Kearney, Harris and Anderson2014) propose social marketing campaigns to increase awareness of pregnancy risks and to inform individuals about the most effective forms of contraception, modeled on Iowa’s “Avoid the Stork,” Colorado’s “Prevention First,” and other similar initiatives. Specifically, they propose that $100 million a year of Title X money be invested through the Office of Population Affairs to state-led campaigns. On fairly conservative assumptions, they predict $5 of savings from each $1 spent on well-crafted campaigns.

Skills and Education: Not Just BAs

Family poverty and instability are strongly associated with employment status, security, and wages, which in turn are greatly, and increasingly, tied to levels of education and skill. The United States is particularly weak in the area of vocational training, compared for example to most European countries. Germany provides a particularly strong model. One important step forward would be to elevate the status of vocational postsecondary learning. The obsessive focus on four-year degrees is now starting to do some real harm, as inexperienced, unprepared students are taking on debt in order to attend low-quality, often profit-seeking institutions. Many drop out, meaning that they end up with the downsides of debt without the upsides of higher earnings potential (Akers and Chingos Reference Akers and Chingos2016).

Community colleges, which have so much potential as an engine of upward mobility, remain “America’s forgotten institutions,” in Darrell West’s (Reference West2010) phrase. “Two-year colleges are asked to educate those students with the greatest needs, using the least funds, and in increasingly separate and unequal institutions,” was the conclusion of an expert task force assembled by The Century Foundation (2013). Fewer than half of those enrolling in community colleges make it through their first year. Six in ten community college students need some extra developmental or remedial education when they arrive (Complete College America 2012).

Given the growing economic, racial, and social divide between two-year and four-year institutions in the United States, there is a strong case for some Title I-style federal investments in community colleges, increasing funds for those working with the most challenging students. Other important steps include simplifying and streamlining the pathway through community college, as Thomas Bailey, Shanna Jeggars, and Davis Jenkins (Reference Bailey, Jeggars and Jenkins2015) argue in their book, Redesigning America’s Community Colleges (Reeves and Rodrigue Reference Reeves and Rodrigue2015b); improving transfer options from two-year to four-year colleges (Reeves and Rodrigue Reference Reeves and Rodrigue2016); and providing more academic support (most community colleges are able to fund one academic adviser per 800 to 1,200 students). (Bailey, Jeggars, and Jenkins Reference Bailey, Jeggars and Jenkins2015).

Apprenticeships offer another promising alternative pathway to traditional college-based education, especially for young men. Improving earnings potential may increase the chances of family stability. Providing more vocationally oriented options at high school is a fruitful avenue too; it is striking that the only educational intervention shown to increase marriage rates among men is Career Academies (Kemple and Willner Reference Kemple and Willner2008). These typically work with around 200 students from grades nine to twelve, and combine academic and technical learning around a specific career theme. They also establish partnerships with local employers to provide work-based learning opportunities.

Family-Friendly Jobs

Many households now need two incomes; and huge changes in women’s education opportunities, aspirations, and labor market participation have rendered the old breadwinner-male plus homemaker-wife model almost obsolete. Most families need to juggle paid work and raising children, but the workplace is still often constructed around previous norms. As a result both men and women can become stretched. Very often policy is focused on extending the school day, or providing child care at unusual hours, essentially attempting to create job-friendly families, rather than family-friendly jobs.

Two potential areas for policy illustrate the potential for promoting more family-friendly jobs. Paid leave for both mothers and fathers would ease the pressure on families, especially when children are young. There is some bipartisan interest in this issue, and right now, the United States is the only advanced country without a national paid leave policy. President Trump’s proposals in this area are restricted to women. Senator Marco Rubio has proposed a 25 percent tax credit for companies that provide at least four weeks of paid leave to employees. A joint AEI-Brookings working group on this issue has recommended a new federally mandated right to paid leave. (AEI-Brookings 2017).

Another area of potential reform is in relation to the unpredictability of hours for paid work, especially in lower paid jobs. Four in ten hourly workers aged 26 to 32 learn their schedules less than a week in advance, according to a study by Lambert, Fugiel, and Henly (Reference Lambert, Fugiel and Henly2014) at the University of Chicago, drawing on the National Longitudinal Survey of Youth (NLSY). Half of them have children. Among the 74 percent of workers who report weekly fluctuations in hours, average hours varied by 49 percent of their “usual” weekly hours. A 10-hour week can follow a 30-hour week. Half have no say over their hours.

Policymakers have some ideas about stabilizing hourly employees’ schedules, some of which have been put into practice. San Francisco passed a bill requiring employers to set schedules at least two weeks in advance (Ludden Reference Ludden2014). Similar legislation has previously been introduced into the US Congress (Kasperkevic Reference Kasperkevic2014). Some business-owners oppose these efforts, and successful legislation will have to balance the interests of employees and firms. Policymakers who are serious about family stability, though, have to be serious about job stability too.

It is worth mentioning here one area of policy that I have not argued for: Direct marriage promotion. This is partly because of a liberal sensibility about governments lecturing adults on how to conduct their lives, but also partly because such efforts rarely work. In the case of marriage promotion, extensive evaluations of various projects funded under the “Building Strong Families” (BSF) initiative have shown little if any impact. The summary of the evaluation reads as follows: “After three years BSF had no effect on the quality of couples’ relationships and did not make couples more likely to stay together.” As Haskins (Reference Haskins2014) concludes:

Of course, in just the same way that public individuals and institutions can use their voices to promote contraceptive use and responsible childbearing, they could also send out a much stronger pro-marriage message, to “preach what they practice,” in Charles Murray’s (Reference Murray2012) phrase. This was in fact one of the recommendations of a AEI-Brookings working group on poverty and opportunity (AEI-Brookings Working Group 2015).

My own view is that the way to promote marriage is to promote the things that predict and bolster marriage, rather than as a head-on goal of policy. Delayed, responsible childbearing; greater economic security and resources through better education; and more family-friendly workplaces. These will likely lead to more marriage, and to more stable families, which is what we are really after.

Enhance Material Resources

One reason children in single-parent, or cohabiting, or unstable households do less well is simply that they are more likely to be poor. A straightforward solution then is to give these families more money, through transfer payments, tax credits, higher minimum wages, or some combination of all of these. Great care however needs to be taken to structure support in a way that does not reduce incentives for paid work or for family formation.

The big question is how far extra money will go in producing better outcomes for the children, especially given that any increases are likely to be fairly modest. The evidence here is that cash does make a difference. Studies of the Earned Income Tax Credit (EITC), for example, show that in families receiving higher payments, test scores go up (Duncan, Morris, and Rodrigues Reference Duncan, Morris and Rodrigues2011). Even college enrolment numbers seem to be positively affected (Manoli and Turner Reference Manoli and Turner2015). Russ Whitehurst summarizes a series of studies that link increased EITC payments to the performance in school of children in the home, and estimates that, for each $1,000 spent on a higher credit, bigger increases in test scores can be observed that in most studies of pre-K education or smaller class sizes (Whitehurst Reference Whitehurst2016), as seen in Figure 10.4.

Figure 10.4 Impact on test scores of low-income children during the school years: reports in effect size of $1,000 of public expenditure on family support, class size reduction, or school readiness

These effects are not trivial, but they are not large either. Given the huge differences in outcomes between children from different backgrounds, the best we can hope for is a modest lessening of the disadvantages faced by children in poorer homes. It also seems likely that there will be considerable heterogeneity in the relationship between additional income and child outcomes. EITC recipients, by definition, are active in the labor market, and while far from comfortable, financially speaking, they do not make up the poorest household, which typically lack wage income. There is however, some evidence that access to food stamps is associated with better health and, at least for women, better long-term economic outcomes (Hoynes, Schanzenbach, and Almond Reference Hoynes, Schanzenbach and Almond2016).

Improve Parenting

To say that children are deeply influenced by their parents, for good and for ill, is to state the obvious. In earlier work with Kimberly Howard on the parenting gap, I showed stark inequalities in parent quality across class and family structure. Children from under-resourced, less stable families are more likely to have parents who are struggling to parent well, which hinders their opportunities (Reeves and Howard Reference Reeves and Howard2013). Waldfogel and Washbrook (Reference Waldfogel and Washbrook2011) estimate that parenting behavior (including maternal sensitivity, reading to a child, out-of-home activities, parenting style, and expectations) explains about 40 percent of the income-related gaps in cognitive outcomes for children at age 4. There are large gaps on measures of parenting by income, education, race, and family structure (and of course these strongly overlap with each other). Focusing specifically on the last of these in Waldfogel and Washbrook’s data, family structure, shows that the odds of being raised by a strong or weak parent (in fact, in these data, mother) varies significantly across household types. Single or cohabiting mothers are significantly more likely to fall into the weakest parents category. However, controlling for other variables, including income and education, the specific effect of being unmarried drops sharply. This suggests that it is the circumstances and traits that are associated with being an unmarried mother that are important, not just the fact of being unmarried itself. Nonetheless, improving the skills of parents, if this is possible, would go a long way to mitigating some of effects of the family gap on child outcomes.

Is it possible, though? Perhaps. Some home visiting programs, aimed at strengthening parenting, have shown positive results (Reeves and Rodrigue Reference Reeves and Rodrigue2015c). Initiatives targeting reading, such as the Home Instruction for Parents of Preschool Youngsters (HIPPY) program, seem to improve literacy among children. For example, the DC Briya/Mary’s Center program may be effective in giving parents skills and building clusters of opportunity in the mold of the Harlem Children’s Zone (Butler, Grabinsky, and Masi Reference Butler, Grabinsky and Masi2015). This is a policy area where not only constant, high-quality evaluation is required but there also looks to be some real potential. Early indications are that the Trump administration is turning away from funding for home visiting. This would be premature and almost certainly counterproductive in terms of promoting family life.

Here too the cultural side is likely to have much stronger effects than policy. There are some signs that in some dimensions at least, such as reading to children, the parenting gap is beginning to close, as less-advantaged parents “catch up,” resulting in some narrowing of school-readiness gaps, too (Reardon and Portilla Reference Reardon and Portilla2016). That parenting styles diverge by class is hardly news. In recent decades, however, time investments in parenting toward the top of the distribution have increased, widening the gap. At the same time, the importance of parenting is now being appreciated more broadly, altering behavior more generally, and potentially narrowing the gap again (Ryan et al. Reference Ryan, Claessens and Markowitz2015).

Enhance Learning Opportunities for Children Outside the Home

If children are negatively affected by their home and family circumstances, the natural inclination for policymakers is to try and improve those circumstances, for instance, through greater redistribution of money or by upskilling parents, as described above. However, it may be that mitigation for family circumstances will lie outside the family, in the shape of better formal education, access to mentors, stronger and safer communities, and so on.

Here I restrict my attention to two potential areas for nonfamily interventions to help mitigate the family gap: More intensive schooling and better communities. Schools can extend the length of their day, and the length of their year. Many charter schools do so. In many cases here the goal is, either explicitly or implicitly, to keep the children away from their families and neighborhoods. Going further still, some schools catering to students from very disadvantaged backgrounds have become boarding schools: The SEED schools in Baltimore, Washington, DC, and Miami provide one example, though there are many others. These schools are expensive, for obvious reasons; and research suggests a higher rate of return from extended-day charters (Curto and Fryer Reference Curto and Fryer2014). The effects are not trivial, though: “We show that attending a SEED school increases achievement by 0.211 standard deviations in reading and.229 standard deviations in math, per year of attendance.” Boarding schools are likely to be most valuable to students from the most-disadvantaged, unstable backgrounds, but so far, little evaluation of impact for students of different types has been undertaken.

Another way to improve the nonfamily environment is through housing policy, and specifically through programs that allow and encourage low-income families to move to more affluent neighborhoods. Recent studies suggest that Moving to Opportunity (MTO), which provides vouchers that enable families to move, have positive long-term effects for children. Specifically, moving to a less poor neighborhood in childhood (i.e., before the age of 13) increased future annual income by the mid-twenties by roughly $3,500 (31 percent), enhanced marriage rates (by 2 percentage points), and raised both college attendance rates (by 2.5 percentage points) and quality of college attended. The age of the child moved was a critical factor: Moving to a less poor neighborhood in the teenage years had no significant impact on later earnings or other adult outcomes (Rothwell Reference Rothwell2015).

The causal factors of these improved outcomes are inevitably hard to tease out. However, the wealth of sociological research on the importance of place suggests that a mixture of school quality, peer effects, social capital, physical security, and safety all likely played a part. Children of single parents benefit from moving to an area where they are the exception, rather than the rule. One important point about the MTO is that the short-term impact on the adults looked to be virtually nonexistent; it was the next generation who really benefited from the move.