Search Type Title Author/Investigator Search Advanced Search

Table of Contents | Previous | Next

CHAPTER II

FAMILY FORMATION IN LOW-INCOME POPULATIONS

Interventions that seek to directly affect family relationships should be grounded in a thorough understanding of the context and circumstances under which parents decide to marry, cohabit, or live alone, and of how these choices impact their children. A large body of research shows family structure matters for children’s well-being and development (McLanahan and Sandefur 1994; Duncan and Brooks-Gunn 1997; Amato 2001; Amato and Rivera 1999). Even after taking low income into account, findings show single parenting still contributes to lower educational attainment and more behavioral and psychological problems in children (McLanahan 1997). This chapter discusses factors of particular relevance to the formation and positive development of low-income families and their children, with the aim of identifying the most promising areas for intervention.

Children develop within a complex “ecosystem” of direct and indirect influences (Bronfenbrenner 1986). Figure II.1 displays some of these influences, with an emphasis on family-level factors because they are the predictors of most interest in this conceptual framework. Child well-being is shown as directly influenced by both parenting behavior and key aspects of the parental relationship and family functioning, such as the level of parental conflict and cooperation. In turn, parenting and family functioning are related to the quality, stability, and structure of the mother-father relationship. These family-related factors are associated with each parent’s individual characteristics and resources. And all of these factors are embedded within, and often affected by, the broader culture, economic conditions, and the structure of tax and transfer policies.

 Figure II.1 Family Formation in Low-income Populations

This chapter begins by discussing aspects of the broader environment that affect family formation, such as cultural attitudes and expectations regarding marriage and childbearing, the economy, and the structure of public policies for low-income families (Section A). Section B explores key characteristics that are static but nevertheless relevant here because they may suggest how to target interventions and identify which services are needed for different populations. Section C describes dynamic characteristics that can be improved by interventions. The chapter ends with a brief description of three broad opportunities for interventions (Section D).

A. ENVIRONMENTAL FACTORS

As shown in Figure II.1, broader environmental factors that are thought to affect family formation fall into three main categories: (1) cultural influences, (2) economic conditions, and (3) policies affecting low-income families. Some factors may affect family formation decisions directly, such as societal attitudes or policy disincentives for marriage. Other factors affect family formation through individual-level resources, such as employment and the ability to provide for a family.

1. Cultural Factors

Four key cultural factors may have an important influence on family formation: acceptance of alternatives to marriage, gender role expectations, gender distrust, and women’s economic independence.

Acceptance of Alternatives to Marriage. In recent decades, Americans have dramatically changed their view of sexuality, marriage, and childbearing. Studies show men and women of different racial/ethnic groups and income levels see marriage as an ideal (Thornton 1989; Tucker and Mitchell-Kernan 1995; Tucker 2000; Oropesa and Gorman 2000; McLanahan et al. 2001). However, the “normative imperative” to marry has weakened since the 1950s, and there now is much greater acceptance of singlehood, cohabitation, and nonmarital childbearing (Thornton 1989; Thornton and Young-Demarco 2001). This mindset is reflected in changes in family formation over time. Fewer cohabitations result in marriage than in the past, more cohabiting couples are raising biological children together, and couples are increasingly likely to cohabit rather than marry in response to premarital pregnancy (Bumpass et al. 1991; Bumpass and Lu 2000).

Evidence suggests pro-marriage attitudes and acceptance of alternatives to marriage, while generally strong, may vary by racial and ethnic group. For example, nationally representative data indicate Mexican Americans are more supportive of marriage than are non-Latino whites or Puerto Ricans (Oropesa 1996). The meaning of cohabitation—whether a precursor or a substitute to marriage—also has been found to differ across racial/ethnic groups. Cohabitation appears to function primarily as a transition to marriage among whites but does not appear to be associated with later marriage among blacks (Manning and Landale 1996). Some scholars argue that the racial patterns in black marriage have their roots in long-standing cultural traditions that arose from slavery and Western African cultural traditions (Patterson 1998; Peterson 1991; Morgan 1993). Others make the case that cultural norms cannot fully explain the racial/ethnic disparities in marriage rates and that changes in the structural supports for marriage, such as the availability of employment for low-skilled men, are more to blame (Furstenberg 2001; Wilson 1996; Anderson 1999; Sassler and Schoen 1999).

Gender Role Expectations. Some suggest that people now have higher expectations of marriage, and that women in particular expect more from their spouses in terms of respect, intimacy, and communication (Ooms 2002; Furstenberg 1996; Edin 2000). Researchers theorize that these higher expectations may arise from an increasingly blurred division of labor and specialization of roles within marriage. For example, household labor, including childrearing, now is divided somewhat more equally between men and women than in the past (Shelton 2000).

Expectations appear to differ across racial/ethnic groups as well. Black women appear to be less likely than white or Hispanic women to marry someone who will not provide financial security (Sassler and Schoen 1999; Edin 2000; Edin 2001). This pattern is confirmed in new national data showing the unemployment rate for men is more important in predicting the transition from cohabitation to marriage among black women than it is among white women (Bramlett and Mosher 2002).

Gender Distrust. Ethnographic research on unmarried low-income women suggests issues of trust between men and women can act as a serious deterrent to marriage in this population (Edin 2001). Edin found that women in the study did not trust men’s ability to be sexually faithful or to be responsible with the family’s money and with their children. Although many of these women said marriage is an ideal they aspire to, they voiced concerns about getting trapped in marriages that do not offer the benefits they expect and that they believe would only complicate their lives.

Women’s Economic Independence. Scholars have theorized that the increasing economic opportunities open to women may reduce their incentive to marry by increasing their ability to support themselves financially outside of marriage. However, empirical research suggests earning more money may make women more, rather than less, likely to marry (Oppenheimer 2000; Lichter and Graefe 2001). One analysis using nationally representative data shows economic independence increases women’s likelihood of marrying, particularly during their late 20s and early 30s (Sassler and Schoen 1999).

Role of Religious Institutions. Churches and other religious institutions can provide a key source of cultural support for marriage; indeed the vast majority of weddings occur in a place of worship. Religious institutions can encourage beliefs and behaviors that are conducive to the success of marriage, such as sexual fidelity and an ethic of sacrifice and commitment to the relationship. Yet as some scholars point out, one paradox of religious behavior in America is that of all racial/ethnic groups, African Americans have the highest rates of religious observance and the lowest rate of marriage (Wilcox 2002). Some surmise that this is because many black churches have responded to the high rates of nonmarital childbearing in their congregations by downplaying pro-marriage norms (Anderson, Browning and Boyer 2002; Wilcox 2002).

2. Economic Conditions

Researchers suggest that falling or stagnant wages and rising unemployment among low-skilled men may be contributing to a shortage of marriageable men, especially in black communities, and that this has contributed to the decline in marriage rates (Wilson 1987). Employed adults are more likely to marry; in one study, single black men who were steadily employed were twice as likely to marry as single black men who were not in school or working (Testa and Krogh 1995). Moreover, the relationship between economic factors and marriage has been found to be stronger for blacks and Hispanics than for whites (Tucker and Mitchell-Kernan 1995), and the relationship between marriage and employment may be growing stronger over time (Testa and Krogh 1995).

Recently released data confirm that community-level economic conditions are linked with marital stability. An analysis of national longitudinal data using five different indicators of community-level socioeconomic status (male unemployment rate, median family income, percent of families below poverty, percent of households receiving public assistance, and percent of college-educated adults) consistently shows community affluence was strongly associated with the stability of marriages and cohabitations, while community impoverishment was not conducive to these outcomes (Bramlett and Mosher 2002).

3. Public Policies

Many tax and transfer policies that affect low-income families include disincentives to marriage. Most means-tested programs base benefits on combined family income. Hence, a woman who marries a man with earnings (whether or not the man is her child’s biological father) may suffer a reduction in a wide variety of benefits, including TANF, food stamps, Medicaid, housing assistance, and child care subsidies. A woman who does not marry but lives with the father of her child also is subject to these disincentives, but the disincentives to cohabitation are not as strong because it is easier not to report a cohabiting partner than to not report a spouse. Moreover, living with someone who is not the child’s father may not reduce public benefits at all. Disincentives to marriage can be substantial; in Oklahoma, parents who cohabit but do not report their cohabitation may have total income (earnings and benefits) of up to twice that of married couples with the same earnings (Hepner and Reed 2002).

B. PARENTAL BACKGROUND CHARACTERISTICS: STATIC FACTORS

A wide variety of personal-level factors are associated with family formation. It is useful to consider these as either static factors, those that are relatively difficult or impossible to change through a direct-service intervention, and dynamic factors, those that are more amenable to change. This general logic has been used by developers of a variety of programs for families (e.g., Stanley 2001) and is useful in both determining the target population and the types of services that would be most beneficial for different populations. Figure II.1 shows some background characteristics of the mother and father that strongly influence family formation and that are static. These factors include the parents’ age when their child was born, their racial/ethnic backgrounds, whether they have another child or children by different partners, the nature of their relationship and plans for marriage when the child is born, and experiences of domestic violence.

1. Parent Age at Child's Birth

Many nonmarital births occur to teenagers; in the Fragile Families study, 29 percent of unmarried parents were 19 or younger when their children were born (Carlson 2002). These parents were less likely than older parents to marry one another. Moreover, couples who first marry before age 20 are more likely to divorce than people who first marry between 20 to 23 (Johnson et al. 2002). Researchers surmise that because adolescents have not yet matured, they are less prepared to assume the responsibilities of marriage (Booth and Edwards 1985).

2. Racial/Ethnic Background

Racial/ethnic differences exist in marriage patterns, family structure, and family formation. For example, among welfare recipients, more than 82 percent of blacks have never married compared with 45 percent of whites and 58 percent of Hispanics (Jacobson 2002). However, research is inconclusive about whether these racial/ethnic differences reflect structural differences in factors correlated with race/ethnicity, such as the availability of jobs, rather than attitudinal or cultural differences. A recent analysis finds disparities in views about marriage account for only a small proportion of differences between the marriage rates of blacks and whites; economic factors play a much bigger role (Sassler and Schoen 1999).

3. Type and Quality of Relationship at Child's Birth

It was once thought that most children born out of wedlock were the product of casual sexual liaisons, characterized by less-than-meaningful relationships between the adults. The Fragile Families study shows this largely is not the case. As noted in Chapter I, more than 82 percent of low-income unmarried couples said they were romantically involved with each other when their children were born, many were cohabiting, and many expected to marry (Carlson 2002). It also is true that 10 percent of unmarried couples reported having little or no contact with each other and consequently were unlikely to develop an interest in marrying one another.

4. Multiple Partner Fertility

Parenting and couple relationships can be particularly complicated if one or both parents have children with other partners. According to a recent analysis of the Fragile Families data, 36 percent of both mothers and fathers have children with other partners (Mincy 2001). Of those mothers who have two or more children, more than half have a child by someone other than the father in the study. Multiple partner fertility is found to be more common among blacks (46 percent) than whites or Hispanics (22 and 29 percent, respectively). It is more common among mothers 25 years and older (who are more likely than younger women to have had two or more children) (Mincy 2001). Some evidence suggests multiple partner fertility discourages marriage (Lichter and Graefe 2001; Mincy 2001). Mothers may be reluctant to marry fathers with financial and emotional responsibilities to other children, and fathers may be reluctant to take responsibility for nonbiological children.

5. Presence of Domestic Violence

Domestic violence is a clear indicator of a poor-quality relationship. The prevalence of domestic violence is estimated to be higher among low-income populations compared with the general population. Studies suggest the rate of past-year domestic violence among the welfare population ranges from about 10 to 31 percent (Danziger et al. 2000; Raphael and Haennicke 1999). However, only 4 to 5 percent of low-income, new, unwed mothers in the Fragile Families study reported their children’s fathers were sometimes or often violent or often criticized or insulted the mothers in the month before the children were born (Carlson 2002).

In general, the prevalence of domestic violence typically is higher for lifetime occurrence compared with reports of current experience. This is likely to be true for comparisons of past-month and past-year violence, which may partially explain the discrepancy between the lower rate of reported violence in the Fragile Families sample compared with other studies of low-income families. It is also possible that domestic violence is less prevalent in the month preceding a child’s birth.

C. PARENTAL RESOURCES AND CAPABILITIES: DYNAMIC FACTORS

Some parental characteristics that affect family formation and child well-being can be changed (Figure II.1). Interventions may change parents’ relationship skills and attitudes and expectations about marriage. They may increase the potential to become successful mates and parents by, for example, improving parents’ prospects for employment or their physical and mental health. Healthy men and women with steady employment prospects are more likely to be good providers for their children and to be more attractive as mates.

1. Relationship Skills

The quality of relationship skills (such as the ability to deal effectively with interpersonal conflict) among the low-income population nationally is unknown. But evidence from a statewide survey in Oklahoma suggests low-income1 married adults exhibit greater frequency of negative interaction, less commitment, and a lower frequency of going on dates with their spouses compared with the general population (Johnson et al. 2002). Those respondents reported less marital happiness and satisfaction and more marital instability than other survey respondents. It is possible that financial hardship predisposes low-income couples to greater difficulty managing interpersonal conflict because of the higher levels of stress they typically experience. It is also possible that poorer interpersonal skills lead to poorer employment performance and thus financial hardship. In either case, strengthening relationship skills may be especially important for this population.

2. Attitudes Toward Marriage and Cohabitation

New unwed parents typically have a positive attitude toward marriage. At the time of their children’s birth, the majority of unmarried mothers (64 percent) and fathers (77 percent) in the Fragile Families survey agreed or strongly agreed with this statement: “It is better for children if their parents are married” (Carlson 2002). Even the majority of couples in the study who were no longer romantically involved agreed children would be better off if their parents married.

A positive attitude toward marriage is a good predictor of later marriage. The Fragile Families study finds unwed parents who believe marriage is good for children were more likely to be married one year later (Carlson 2002).

Some evidence from the Oklahoma survey suggests low-income adults (not necessarily new unwed parents) may not be as positive about marriage but more accepting of cohabitation and nonmarital childbearing than adults in the general population. Similar to the findings in the Fragile Families study, about 60 percent of low-income respondents in the Oklahoma survey agreed people who have children together ought to be married (Johnson et al. 2002); in comparison, 77 percent of Oklahomans in the general population agreed with this statement. About 50 percent of low-income adults in Oklahoma agreed cohabitation is an acceptable arrangement compared with 35 percent in the state’s general population. Low-income adults also indicated greater openness to divorce, being more likely to believe that parents who do not get along or who no longer love each other should divorce even if they have children (Johnson et al. 2002).

3. Education and Employability

Many low-income unwed parents have low education and/or low earnings. About 45 percent of low-income unmarried mothers and 38 percent of unmarried fathers in the Fragile Families study had not finished high school, and 4 percent had attended college when their babies were born (McLanahan et al. 2001). About 40 percent of mothers and 19 percent of fathers earned less than $5,000 in the 12 months before the study.

People with low education and earnings are not as likely to marry (White and Rogers 2000; Carlson 2002) because they are viewed as less attractive potential spouses. Lack of employment prospects, economic instability, and low income also create stress for families and may be key factors in causing relationships to fail.

4. Physical and Mental Health

Evidence suggests low-income families have poorer health—including chronic conditions, functional limitations, and disabilities—than people in the general population (Zedlewski and Alderson 2001). A nationally representative study that interviewed adult welfare recipients shows 17 percent said health problems limit their ability to work. In addition, many parents who were terminated from TANF because of sanctions reported they were unable to comply with the program’s requirements because of a health condition, illness, or disability (Sweeney 2000). Such problems can affect many spheres of a person’s life, including the ability to contribute to daily household tasks and chores, and to carry out parenting tasks. These limitations can further stress a couple’s relationship.

Substance abuse and dependence bring special challenges to the formation and maintenance of strong and healthy relationships and marriage. Depending in part on the measure used, the percentage of welfare recipients with substance abuse problems ranges from 16 to 37 percent (Center on Addiction and Substance Abuse 1999). Substance abuse and dependence have major implications not only for employment, but also for the quality and stability of family relationships, marital interaction, and parenting behavior.

Mental health conditions are clearly related to marital quality, stability, and satisfaction (Davila et al. 1997; Regier et al. 1993; Whisman et al. 2000), and adults receiving government assistance are more likely to suffer from psychological problems. According to national data, 22 to 28 percent of welfare recipients are estimated to have poor mental health (Zedlewski and Loprest 2001). And many welfare recipients suffer from more than one mental health problem (California Institute for Mental Health 2000). The types of mental health problems most prevalent among welfare recipients are:

• Major Depression. Welfare recipients are at greater risk than the general population for major depression (Moore et al. 1995; Quint et al. 1997). More than one-quarter of a sample of welfare recipients in an urban Michigan county was classified as having a major depressive disorder compared with 13 percent in the general population (Danziger et al 2000). The CalWORKS Prevalence Project finds 22 percent of current recipients in one California county and 36 percent of welfare applicants in another county could be classified as having major depression (Chandler and Meisel 2000).
  
• Generalized Anxiety Disorder (GAD). GAD is estimated to be present in about 4 percent of women in the general population, according to the National Co-Morbidity Study, the largest nationally representative epidemiological study of mental health in the United States. But recent studies of welfare recipients find higher rates of this disorder: about 7 percent in the Michigan study and 9 to 10 percent in the California samples.
  
• Post Traumatic Stress Disorder (PTSD). The rate of past-year PTSD in the Michigan and California studies ranged from 13 to 14 percent compared with a prevalence of 3.4 percent in the National Co-Morbidity Study. State-level studies find welfare recipients are more likely to experience traumas (such as domestic violence and rape) and to have suffered traumas (such as physical or sexual abuse) in their families of origin. These types of events put them at risk of PTSD (Curcio 1996; Butler and Burton 1990; Chandler and Meisel 2000).

5. Parenting Skills

Parenting behavior, parent-child interaction, and aspects of the home environment are important predictors of children’s development (Maccoby and Martin 1983; Bradley et al. 1989; Bornstein 1995; McLoyd 1990; Patterson et al. 1989). Therefore, many interventions for at-risk families focus on strengthening parenting skills, including teaching strategies for providing adequate cognitive stimulation and emotional support to children and teaching positive discipline.

Yet parenting cannot be separated from the overall context of family structure and parental interaction. In a longitudinal study of new parents, negative marital interaction, such as tension and unresolved conflict, was associated with harsher parenting, less warmth, and less encouragement of children’s autonomy (Cowan and Cowan 2002). Parenting skills also may affect the likelihood of family formation. In considering marriage, an unwed parent may take into account a potential mate’s ability to effectively parent the children.

6. Kin Support

The influence of extended family may affect family formation decisions and parenting of low-income couples. Couples may receive support from extended family in the form of money, help with child care, encouragement, or guidance. Alternatively, the extended family could complicate a couple’s decision to stay together and/or marry. Sometimes, family members do not approve of their loved one’s choice of partner or the fact that there is a child out of wedlock. This can be an important issue with regard to teen parents, many of whom are still living at home under the authority of their parents.

The involvement of extended family also can create conflict around childrearing, creating implications for the children’s well-being and development of the parental relationship. A study of young black mothers, their preschool children, and the grandmothers of the same children show more favorable parenting as well as grandparenting when the mothers and grandmothers are not living together (Chase-Lansdale et al. 1994).

7. Religiosity

Religiosity is associated with higher rates of marriage and marital stability (Thornton et al. 1992; Bumpass 2000). Data from the National Survey of Family Growth indicate that women who report their religion is not important are less likely to marry than other women (Bramlett and Mosher 2002). A recent analysis of Fragile Families data supports the idea that religious institutions and observance play a key role in the likelihood of marriage (Wilcox 2002). The study shows unmarried mothers who attend church frequently are 90 percent more likely to marry within 12 months of their children’s birth compared with unmarried mothers who do not attend church frequently.

D. OPPORTUNITIES FOR INTERVENTION

Some of the static parent characteristics discussed in this chapter are factors for which individuals and couples should be assessed (young age, multiple partner fertility, lack of interest in marriage, and presence of domestic violence) because they may define subgroups that may not be appropriate for interventions that seek to strengthen relationships and promote healthy marriages between parents. These subgroups could, however, benefit from other services or interventions that could reduce further nonmarital childbearing and promote cooperation between the parents to maximize the potential for positive involvement of both parents in their child’s life.

The environmental factors and the dynamic personal characteristics that affect family formation discussed in this chapter suggest types of services that could be provided or policies that could be changed for those unwed adult parents who do choose to marry. The following three broad types of interventions are discussed in the next three chapters of this report:

• Programs to Improve Couple Relationships. Two important dynamic factors that determine family formation and relationship success are the quality of relationship skills and attitudes and expectations about marriage. Couple relationship and marriage education programs—discussed in the Next—are intended to improve these skills and to change attitudes and expectations that might otherwise interfere with healthy relationships and marriages.


• Services to Improve Marriageability. Other dynamic factors that affect family formation—education and employability, physical and mental health, and parenting skills—affect family formation by making parents more attractive as marriage partners. Services to improve marriageability of parents are discussed in Chapter IV.
  
• Public Policy Changes. Changes in policies are required to reduce the disincentives embodied in many public policies. Because many unwed parents have low incomes, such policies may be significant deterrents in their decisions to marry. Possible policy interventions are discussed in Chapter V.

---

1Defined as adults who currently were receiving or who had at one time received government assistance in the form of TANF, food stamps, or Medicaid. (back)

 

Table of Contents | Previous | Next