Donate
Independent, objective, nonpartisan research

R 701MWR

Authors

R 701MWR

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3711 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(12) "R_701MWR.pdf" ["wpmf_size"]=> string(6) "659689" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(223209) "Unmarried Parents, Fragile Families: New Evidence from Oakland ••• Maureen R. Waller 2001 PUBLIC POLICY INSTITUTE OF CALIFORNIA Library of Congress Cataloging-in-Publication Data Waller, Maureen Rosamond, 1967– Unmarried parents, fragile families : new evidence from Oakland / Maureen R. Waller. p. cm. Includes bibliographical references. ISBN 1-58213-035-3 1. Single-parent families—California—Oakland. 2. Children of single parents—California—Oakland. I. Title. HQ777.4 .W35 2001 306.85'6'0979466—dc21 2001034951 Copyright © 2001 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword In 1997, PPIC was a leading financial supporter of the Fragile Families and Child Wellbeing (FFCW) survey, a national, longitudinal study of unmarried parents and their children. About 33 percent of all U.S. and California children are currently born to unmarried mothers— about three times the corresponding proportion in 1970. Partly because unmarried parents and their children present difficult public policy challenges—for example, the children are more likely to be poor and to receive welfare—PPIC welcomed the opportunity to support a national study on the topic. In the aftermath of the 1996 Personal Responsibility and Work Opportunity Reconciliation Act, the study seemed especially timely. As one of the first products of that study, Maureen Waller’s Unmarried Parents, Fragile Families: New Evidence from Oakland underscores the importance of these policy challenges to California and the nation. Two results are especially noteworthy. First, less than 10 percent of these Oakland couples married in the year after childbirth. After noting problems ranging from financial and relationship problems to the stress of living in a complex urban environment, Waller concludes that federal efforts to encourage marriage will face serious obstacles. Second, she suggests that cohabiting couples should receive special policy attention. The early evidence indicates that such couples are in relatively stable relationships and that cohabitation is associated with high expectations for marriage. This stability, she suggests, could translate into greater family cohesion and child wellbeing over time. Future waves of the national survey will examine other issues pertaining to fragile families, including paternal involvement and children’s wellbeing. As we at PPIC consider welfare reform in its broadest context, we see no issue more important than the circumstances iii and prospects of these children. We hope that this study, and the national effort of which it is a part, will improve not only public policy but also the lives of children today and for generations to come. David W. Lyon President and CEO Public Policy Institute of California iv Summary About 33 percent of all U.S. births occur outside marriage, up from about 4 percent in 1940 and 11 percent in 1970. In California, too, unmarried women now account for one-third of all births. Although increases in nonmarital births have leveled off in recent years, families headed by unmarried parents continue to present significant policy challenges. Children born outside marriage are more likely than other children to be poor and to receive welfare. They are also less likely to receive child support from their noncustodial parents. To address these challenges, welfare legislation and proposals to encourage responsible fatherhood have sought to promote marriage and to strengthen twoparent families. Despite this growing interest in marriage among policymakers, there is very little representative data on unmarried parents’ relationships. As a consequence, we have limited knowledge about why transitions to marriage do or do not occur or how to strengthen fragile families. Because unmarried parents’ decisions to marry or break up often occur in the first years after their children are born, it is especially important to understand relationship transitions during this time. This report investigates factors that support and discourage efforts on the part of unmarried parents to form stable relationships, including marriage. It focuses in particular on 250 families in Oakland who participated in the Fragile Families and Child Wellbeing (FFCW) survey, a national, longitudinal study of unmarried parents and their children. Oakland was the first city surveyed in this study and the first for which follow-up data are available. Drawing on two waves of surveys with parents, this report characterizes the stability of unmarried parents’ relationships in the first year of their child’s life and tracks their transitions toward increased and decreased involvement. It also examines open-ended interviews with parents to analyze economic, personal, and relationship issues that v parents faced in the first year of their child’s life. Finally, the report tests whether the issues identified in the interviews are related to parents’ assessments in the larger survey sample of whether they will marry their child’s other parent. Results One Year After Birth Of the Oakland parents considered in this report, about half were living together at the time of their child’s birth, 35 percent were in romantic relationships but were not living together, and 15 percent were not romantically involved with the other parent. One year later, about 7 percent of parents were married, almost half were living together, and 40 percent were not in a romantic relationship with the other parent. Cohabitation was the most stable relationship type in the first year. About three-quarters of cohabitors were still living together one year after their child’s birth, and an additional 10 percent had married. At the same time, these cohabiting parents were slightly more likely to decrease their involvement in relationships than to marry. Over half of parents who were romantically involved but not cohabiting ended their relationship between the time of birth and the 12-month follow-up; about 4 percent of these parents married and another 27 percent moved in together. Not surprisingly, the overwhelming majority of parents who were not romantically involved at the time of birth remained uninvolved one year later (see Figure S.1). Parents in Oakland reported high expectations about marriage at the time of their child’s birth. Over one-third of the mothers and almost half the fathers reported an almost certain chance that they would marry. Cohabiting parents had the highest expectations about marriage. Among mothers who were living with the father when their child was born, 53 percent reported an almost certain chance of marriage and only 8 percent indicated that there was little or no chance that they would marry the other parent. These expectations were related to the outcomes observed one year after the birth. Among parents who reported an almost certain chance of marrying the other parent, about 14 percent were married in the first year and an additional 68 percent were living together one year later. About 5 percent of parents who reported a good or 50-50 chance of marriage were married and slightly over half were cohabiting; about vi Percentage 100 13 90 2 80 70 55 60 50 75 97 40 15 30 20 27 10 10 0 43 Cohabiting Involved in Not involved at birth a romantic in a romantic relationship but relationship not cohabiting at birth at birth Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up NOTE: Reports from mothers in both waves of FFCW survey. Percentages may not sum to 100 because of rounding. Figure S.1—Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers one-third did not have romantic relationship 12 months after the birth. Of those parents who reported little or no chance of marriage, 83 percent did not have a romantic relationship with the other parent at the 12month follow-up. The results indicate a clear and statistically significant relationship between parents’ expectations and early outcomes. Insights from Qualitative Interviews The open-ended, qualitative interviews conducted for this report illustrate some of the economic, relationship, and personal challenges parents faced in the first year. The analysis identifies several factors related to the dissolution of parents’ relationships before the birth of their child or in the first 12 months. It also considers obstacles to relationship stability among couples who stayed together and how these obstacles contributed to their reluctance to marry. Parents who ended their relationships often mentioned a specific event (such as a violent incident, vii drug relapse, or deportation) that triggered the breakup. Other parents mentioned general relationship problems such as conflicts about trust, infidelity, and commitment. Parents were more likely to point to specific events or relationship problems than to economic factors when discussing what prompted their breakups. However, their accounts indicate that financial instability, including not having a place to stay together, occurred simultaneously with these problems and probably contributed to them. Parents who broke up before the follow-up interview were typically experiencing multiple problems when their relationships ended. Parents who stayed together in the first year mentioned many of the same problems as those who did not. Although these couples may have experienced less-severe or less-numerous problems than those who dissolved their relationship, they were primarily differentiated from other parents by the fact that they were living together at the time of their child’s birth. Many of the parents who stayed together in the first year regarded their relationships as highly vulnerable to dissolution, and most said that they wanted to wait until the conditions were right for marriage. They were often hopeful that their economic and personal circumstances would improve, making marriage possible. Despite these hopes, these couples are likely to experience continuing economic problems. Couples in the general population share some of the problems reported by these unmarried couples, which may explain why a large number of men and women are delaying marriage; however, the material hardship experienced by most parents at the Oakland site exacerbated common relationship problems and introduced new ones. Factors Related to Early Expectations The report also tracks the relationships between issues identified in the interviews—economic problems, distrust, conflict, physical violence, and drug and alcohol abuse—and how parents assessed their chances of marrying each other. These assessments are important insofar as they reflect parents’ expectations and predict relationship outcomes at year one. This part of the report is based on statistical analyses of responses from 3,712 mothers and 2,775 fathers to the FFCW survey. Because the models control for significant interactions between parents’ responses viii and their residences, the 20-city FFCW results can be applied to the Oakland sample. The analysis includes information from qualitative interviews to illustrate important findings. Cohabitation has the strongest positive effect on marriage expectations. This pattern indicates that unmarried parents view cohabitation as a step toward marriage. Also, the experience of cohabitation may increase parents' expectations that they will marry. Fathers’ employment is positively related to mothers’ reporting high chances of marriage. Also, mothers with higher education—an indicator of employment potential—express more certainty about marriage than other mothers. The results show strong negative relationships between marriage expectations and the belief that the other gender cannot be trusted to be faithful. Likewise, reports of high levels of conflict, violence, and partners with drug or alcohol problems are negatively related to marriage expectations. The effects of these factors are more consistent than those associated with economic factors. Conclusion and Policy Implications These findings provide an early look at relationships between unmarried parents in the first year of their child’s life. About half of unmarried parents in Oakland were living together when their children were born, suggesting a relatively high level of commitment in their relationships. One year later, parents were divided between those living in formal or informal unions and those who were no longer involved in romantic relationships. Despite high expectations at the time of their child’s birth, only 7 percent of unmarried parents had married. Parents who were living together were more likely to break up than to marry, and over half of the parents who were romantically involved but not living together broke up within 12 months of their child’s birth. These results suggest that policies and programs to encourage marriage face serious challenges. Programs that focus on maintaining relationships, stabilizing families, and facilitating the transition to marriage are likely to have the highest chances of success among cohabiting parents. However, even these parents—who, on average, have relatively stable relationships and ix high marriage expectations-—will have difficulty making the transition to marriage because of personal and economic problems. More feasible, perhaps, are programs that help fathers remain emotionally and economically involved in their children’s lives. Since 1996, states have been given more flexibility to initiate or support such programs. Eligibility criteria for cash assistance (such as Temporary Assistance for Needy Families, or TANF) and in-kind assistance (such as food stamps, housing, and health coverage) are primarily geared to single custodial parents. New proposals to broaden eligibility criteria for in-kind assistance and to eliminate distinctions between single-parent and twoparent families in receiving cash assistance may help these fragile families remain intact. Conflict and distrust between mothers and fathers may be less amenable to policy intervention than economic problems. Some community-based fatherhood programs have addressed these issues directly, but such programs tend to target noncustodial fathers with child support obligations, many of whom are no longer cohabiting or romantically involved with the other parent. The state might consider implementing more programs to serve unmarried couples still involved in romantic relationships, particularly couples who are cohabiting. Programs for parents with drug or alcohol problems may allow them to stabilize their families and increase their capacity for employment. Drug treatment could also help some noncustodial fathers emotionally reengage with their children. However, encouraging marriage or increased family involvement for parents with drug or alcohol problems may be unrealistic or detrimental to other family members. In families characterized by domestic violence, this danger is even more pronounced. Knowing more about these parents’ circumstances will help policymakers identify cases for which encouraging marriage would be inappropriate or unfeasible. This preliminary evidence suggests that the personal and economic barriers to marriage and relationship stability among unmarried parents are numerous, significant, and complex. As the FFCW study continues, policymakers will have an even clearer idea of how these complex issues affect unmarried parents’ capacity to form and maintain stable relationships, including marriage. x Contents Foreword ..................................... Summary..................................... Figures ...................................... Tables ....................................... Acknowledgments ............................... iii v xiii xv xvii 1. INTRODUCTION ........................... 2. STUDY DATA AND METHODS ................. Qualitative Study ............................. Characteristics of the Oakland Sample ............... 3. OUTCOMES AND EXPECTATIONS IN THE FIRST YEAR .................................... Changes in Unmarried Parents’ Relationships Between Birth and 12-Month Follow-Up .................... Expectations About Marriage ..................... Relationship Expectations versus Outcomes ............ Conclusion ................................. 4. INSIGHTS FROM QUALITATIVE INTERVIEWS ..... Parents Not in Romantic Relationships at Follow-Up ...... Relationship Ended in the First Year ............... Relationship Ended Before the Birth ............... Parents in Romantic Relationships at Follow-Up ......... Married at Year One ......................... Involved in Other Romantic Relationships at Year One ... Fragile Families in the First Year .................. Conclusion ................................. 5. FACTORS UNDERLYING EARLY EXPECTATIONS ... Men’s Employment and Education ................. Women’s Employment, Education, Public Assistance, and Household Income ........................ Gender Distrust.............................. 1 7 9 10 19 19 26 30 33 35 35 37 40 41 41 42 47 49 53 55 57 60 xi Relationship Conflict .......................... Violence in Relationships ........................ Drug and Alcohol Problems ...................... Race, Ethnicity, and Immigration .................. Age and Other Children ........................ Cohabitation ............................... Conclusion ................................. 6. CONCLUSION AND POLICY IMPLICATIONS ...... Financial and Housing Stability ................... Distrust and Conflict in Relationships ............... Drugs, Alcohol, and Violence ..................... 61 63 64 65 66 67 69 71 72 74 75 Appendix A. The Fragile Families and Child Wellbeing Study ......... 77 B. Sample Means and Regression Results ................ 85 Bibliography .................................. 97 About the Author ............................... 105 Other Related PPIC Publications ..................... 107 xii Figures S.1. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers.................. 2.1. Relationships Between Unmarried Parents at Child’s Birth, As Reported by Mothers ................. 3.1. Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers .......................... 3.2. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers.................. 3.3. Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers, by Race .................... 3.4. Expectations About Marriage at Child’s Birth, As Reported by Oakland Mothers and Fathers.......... 3.5. Expectations About Marriage, As Reported by Oakland Mothers, by Relationship at Child’s Birth........... 3.6. Relationships at 12-Month Follow-Up, As Reported by Oakland Mothers, by Expectations About Marriage .... 5.1. Odds Ratios for Mothers’ Reports of an Almost Certain Chance of Marriage ........................ vii 16 20 23 24 27 29 31 60 xiii Tables 2.1. Demographic Characteristics of Unmarried Parents .... 2.2. Socioeconomic Characteristics of Unmarried Parents ... 2.3. Material Hardship Caused by a Lack of Money Among Unmarried Parents in Oakland Between Child’s Birth and 12-Month Follow-Up .................... 3.1. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers.................. 3.2. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers, by Race ............ 3.3. Means for Marriage Expectations, by Type of Relationship at Child’s Birth, Oakland ............ 3.4. Odds Ratios from Logistic Regressions of Mothers’ Reports of Being Married, Married or Cohabiting, or Not Romantically Involved at 12-Month Follow-Up ... 5.1. Factors Related to Parents’ Expectations About Marriage................................ A.1. Comparison of Demographic and Socioeconomic Characteristics of Unmarried Parents in Oakland in the Sample at 12-Month Follow-Up and Parents Not in Sample at 12-Month Follow-Up ................ B.1. Means for Dependent Variables in Logistic Regressions .. B.2. Means for Independent Variables in Logistic Regressions .............................. B.3. Odds Ratios from Logistic Regressions of Mothers’ Reports of an Almost Certain Chance of Marriage ..... B.4. Odds Ratios from Logistic Regressions of Mothers’ Reports of Little or No Chance of Marriage ......... B.5. Odds Ratios from Logistic Regressions of Fathers’ Reports of an Almost Certain Chance of Marriage ..... 11 14 15 21 25 30 32 54 83 85 86 89 91 93 xv B.6. Odds Ratios from Logistic Regressions of Fathers’ Reports of Little or No Chance of Marriage ......... 95 xvi Acknowledgments This report would not have been possible without the contributions of many people--including the mothers and fathers who kindly agreed to participate in the study. I would like to thank PPIC for its generous support of the Fragile Families and Child Wellbeing Study’s first site in Oakland, California. In particular, David Lyon’s sustained enthusiasm, direction, and backing of the project were instrumental in launching what is now an important, national study of unmarried parents and their children. Michael Teitz provided guidance throughout various stages of the project and Debbie Reed offered important advice on this report. My thanks go to David, Mike, and Debbie for their support of this multiyear effort. I would also like to express my appreciation to the National Institute of Child Health and Development, the Robert Wood Johnson Foundation, and the William and Flora Hewlett Foundation for their contributions to the Oakland site. Thanks to Frank Furstenberg, Laura Hill, Wendell Primus, and Elaine Sorensen for their valuable suggestions for revising the manuscript. The report also benefited from Laura Moye’s diligence, care, and intellectual input into the analysis and production of the report. Renzo Tragsiel provided excellent assistance for the qualitative study and Suzanne Shoemaker expertly transcribed the qualitative interviews. Peter Richardson’s significant editorial skills helped to sharpen the report’s presentation. Nancy Reichman and Ofrira Schwartz at Princeton University made the use of the survey data possible and helped coordinate the research process. Finally, I would like to thank Sara McLanahan, Irv Garfinkel, and Ron Mincy for envisioning this innovative study of fragile families. As principal investigators of the study, Sara and Irv have also made this vision a reality. Despite these contributions, all errors in this report are the responsibility of the author alone. xvii 1. Introduction About 33 percent of all U.S. births occur outside marriage, up from about 4 percent in 1940 and 11 percent in 1970 (Ventura and Bachrach, 2000).1 In California, too, one-third of births occur to unmarried women (Curtin and Martin, 2000). Although increases in nonmarital births have leveled off in recent years, the prevalence of families headed by unmarried parents presents significant policy challenges.2 Children born outside marriage are more likely than other children to be poor and to receive welfare, and they are less likely to receive child support from their noncustodial parents. In light of these challenges, the language of welfare reform legislation and recent proposals to encourage “responsible fatherhood” emphasize marriage and strengthening two-parent families. According to the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, two of the four purposes for which Temporary Assistance for Needy Families (TANF) funds can be used include ____________ 1Daniel Patrick Moynihan’s controversial report, The Negro Family: The Case for National Action (1965), opened the contemporary policy debate about the role of public policy in promoting two parent families. Although its focus was on black families, the report drew public attention to the changing structure of families and called for a national response to this issue. At the time that report was released, however, less than 8 percent of all births in the United States occurred to unmarried women. The Moynihan report also focused on the fact that about one-quarter of black children were born to unmarried parents. This is about same proportion (22 percent) of children who are now born to unmarried white parents. Births to unmarried white women make up the largest percentage of births outside marriage in the United States (40 percent), followed by births to black women (32 percent) and Hispanic women (25 percent). However, the proportion of births outside marriage is higher among black and Hispanic women— about 69 percent and 42 percent, respectively (Ventura and Bachrach, 2000). 2Social scientists have reported that marriage has declined and that the proportion of nonmarital births has risen across social groups in the United States. Not only have recent cohorts of adults postponed marriage, but men and women are also more likely to divorce and less likely to remarry. Therefore, men and women have a higher risk of having a child outside marriage. Furthermore, fewer men and women today are deciding to marry in response to a nonmarital pregnancy (Ventura et al., 1995). 1 promoting marriage and encouraging the formation and maintenance of two-parent families.3 The President’s budget proposes $64 million in 2002 ($315 million over five years) for programs that strengthen the role of fathers in the lives of families. Competitive grants would be provided to faith-based and community organizations that promote marriage (U.S. Office of Budget and Management, 2001).4 Congress is also considering similar legislation primarily targeted at low-income parents.5 The Strengthening Working Families Act of 2001, for example, would provide funds “to promote the formation and maintenance of married, twoparent families, strengthen fragile families, and promote responsible fatherhood” (U.S. 107th Congress, 2001b). ____________ 3The 1996 welfare reform legislation that created TANF is part of PRWORA. PRWORA begins with the findings that “(1) marriage is the foundation of a successful society” and “(2) marriage is an essential institution of a successful society which promotes the interests of children” (U.S. 104th Congress, P. L. 104-193, Sec. 101). According to PRWORA, TANF funding can be used “to end the dependence of needy parents on government benefits by promoting job preparation, work, and marriage” and “to encourage the formation and maintenance of two-parent families” (U.S. Department of Health and Human Services, 1999, pp. 11–12). 4President Bush emphasized marriage during his campaign and since taking office (Government Executive Magazine Daily Briefing, 2001; Milbank, 2001). The nomination of Wade Horn as Assistant Secretary of Health and Human Services for Family Support also signals the importance of marriage in the current administration. Horn, the former president of the National Fatherhood Initiative, was quoted as saying that he viewed the administration as having a four-pronged strategy of “strengthening fatherhood, strengthening marriage, strengthening community organizations that help families, and seeking a role for religious organizations in building communities” (Milbank, 2001). 5The Strengthening Working Families Act of 2001 (S. 685), the Responsible Fatherhood Act of 2001 (H.R. 1300), and the Child Support Distribution Act of 2001 (H.R. 1471) (U.S. 107th Congress, 2001a,b,c) all contain provisions to encourage marriage and the formation and maintenance of two-parent families through fatherhood programs and media campaigns (Center on Fathers, Families, and Public Policy, 2001). The Child Support Distribution Act, for example, would make grants available to fatherhood programs “that promote marriage through such activities as (A) counseling, mentoring, disseminating information about the advantages of marriage, enhancing relationship skills, teaching how to control aggressive behavior, disseminating information on the causes and treatment of domestic violence and child abuse, and other methods; and (B) sustaining marriages through marriage preparation programs, premarital counseling, and marital inventories, and through divorce education and reduction programs, including mediation and counseling.” 2 For low-income unmarried parents, the barriers to forming stable relationships, including marriage, appear to be formidable and multifaceted (Furstenberg, forthcoming). Demographic data indicate that only about two out of five mothers who had a nonmarital birth married their child’s father or anyone else within five years of the birth (Ventura et al., 1995, pp. 58–59). Furthermore, relationships between unmarried parents, and between unmarried fathers and their children, decline significantly in these early years (Sorensen, Mincy, and Halpern, 2000; Lerman and Ooms, 1993). Although there is substantial and growing interest in relationship transitions between unmarried parents, researchers and policymakers have not had access to representative, longitudinal data on unmarried parents that examine where these changes begin and why they occur. This report draws on a unique survey of unmarried parents and their children to provide an early look at changes in relationships between unmarried parents. The report also looks at factors that support and discourage their efforts to form stable relationships, including marriage. Specifically, the report examines data on 250 families headed by unmarried parents living in Oakland, California, who participated in the Fragile Families and Child Wellbeing (FFCW) survey, a national, longitudinal study of unmarried parents and their children. Although Oakland was the first city to participate in the study, the total sample of unmarried parents includes 3,600 families living in 20 U.S. cities. Because most parents in the FFCW study have low incomes, examining a sample of unmarried parents in Oakland primarily means looking at economically disadvantaged families. The study refers to these unmarried parents and their children as fragile families because they have a higher risk of poverty and family dissolution than do families headed by married parents. Unmarried mothers and fathers participated in a baseline interview at the time of their child’s birth and a follow-up interview 12 months later. This report draws on both waves of interviews to provide new information on changes in the relationships between unmarried mothers and fathers in the first year of their child’s life and factors that underlie 3 these transitions. This report also uses information from qualitative interviews with Oakland families who participated in the study to enrich the analysis. Because the study focuses on men and women who have already become unmarried parents, it addresses only a piece of this complicated puzzle involving the changing nature of American families. However, the report uses newly available data to shed light on factors that may discourage union formation and stability. Most of the increase in nonmarital childbearing since the early 1980s has occurred to cohabiting parents, and a significant proportion of children born outside marriage are now living with both biological parents early in life (Bumpass and Lu, 2000).6 In recognition of these changes, the report pays particular attention to the effect of cohabitation on parents’ relationships and outlooks for the future. The report begins by examining early relationship outcomes (including marriage, cohabitation, and dissolution) in the Oakland sample. Chapter 2 describes the sample and research design, and Chapter 3 discusses relationship transitions that begin soon after unmarried couples have a child. In particular, this chapter looks at changes in relationships between parents who live together, parents who are involved in romantic relationships but do not live together, and parents who reported that they did not have a romantic relationship with the other parent at the time of their child’s birth. The report also examines whether marriage is something unmarried parents think will happen and how expectations match early relationship transitions. Chapter 4 draws on evidence from qualitative interviews conducted between the baseline survey and the 12-month follow-up survey to investigate the challenges to relationship stability that unmarried parents face in the first year of their child’s life. This analysis examines factors related to the dissolution of parents’ relationships before the birth of their child or in the first 12 months. It also looks at challenges experienced by couples who stayed together in the first year and explores how they may be related to couples’ reluctance to marry in the first year. ____________ 6Most of the increase in nonmarital births to cohabiting women occurred among white women (Ventura and Bachrach, 2000). 4 Chapter 5 tests whether the challenges identified in the qualitative interviews are related to expectations held by parents in the larger sample. Expectations reflect parents’ own sense of the potential for marriage, and the factors that influence expectations are likely to be the same as those that influence outcomes. Because the size of the follow-up sample in Oakland is not large enough to permit multivariate regression analysis, parents’ assessments of whether they will marry their child’s other parent are used as a proxy for relationship outcomes, such as marriage, cohabitation, and relationship dissolution. Examining parents’ expectations has the added benefit of allowing us to think about how relationships might continue to change beyond the first year. The analysis first considers how different dimensions of unmarried parents’ economic status and security may affect their relationships. A significant body of research shows a positive connection between male employment and marriage. Male employment is of particular concern in low-income communities, especially poor, black communities where, scholars argue, high levels of joblessness render men less “marriageable” in the eyes of women. The analysis next examines how general perceptions of trust in relationships influence unmarried mothers’ and fathers’ orientations toward marriage. Scholars suggest that trust issues may be especially important for couples living in economically unstable environments. The analysis then investigates how conflict affects unmarried parents’ relationships. Relationships with high levels of conflict are vulnerable to dissolution and may be less likely to result in marriage. Also, there is reason to believe that new, unmarried parents with limited economic resources may experience particularly high levels of stress and conflict in their relationships. Finally, the analysis considers the effects of domestic violence and drug or alcohol abuse—indicators of serious and potentially dangerous problems. Although many of these issues contribute to delayed marriage in the general population, they may be more severe or numerous among unmarried parents and present greater barriers to relationship stability.7 ____________ 7Social scientists have identified multiple changes in the economy, culture, and gender expectations that have discouraged marriage and encouraged increases in nonmarital childbearing. Many of these changes have occurred across social groups in the United States and in other industrialized countries. In recent years, men’s earnings have 5 This information and analysis can help policymakers, community leaders, and practitioners identify problems that unmarried parents face at the time of their child’s birth and how these problems affect their relationships during the first year. This information also provides a window into the kind of difficulties families may continue to experience in the coming years. ________________________________________________________ declined relative to women’s, changing the conventional “gender bargain” in marriage (McLanahan and Sandefur, 1994). Theories about culture and gender suggest that these factors had equally important roles in the shift toward delayed marriage (or nonmarriage) as economic factors. For example, young men and women now perceive fewer advantages to marriage than in the past (Thornton and Freedman, 1982) and seem to have higher expectations for realizing cultural ideals, such as compatibility in their relationships. Scholars suggest that changing expectations about gender have also meant that women are now less willing to tolerate extremely unequal or abusive marriages (Stacey, 1991; Luker, 1996). 6 2. Study Data and Methods This study of early relationship transitions focuses on unmarried parents living in Oakland, California—one of 20 cities across the United States in which parents are participating in the FFCW study.1 The FFCW study follows a new birth cohort of approximately 4,700 children to learn more about the circumstances and experiences of unmarried parents and their children in the first four years of their children’s lives.2 In particular, the study investigates such things as the relationships between unmarried parents, the involvement of unmarried fathers in their children’s lives, and the wellbeing of children born outside marriage (see Appendix A). The Oakland sample includes 250 families headed by unmarried parents and a comparison group of 75 families headed by married parents. These families were approached at the time of their children’s birth and will be followed for four years. New mothers were initially interviewed in person at the hospital within 24 hours of giving birth. Fathers were interviewed either at the hospital or elsewhere as soon as possible after the birth. The first interview with the mother lasted about 30 minutes and the interview with the father lasted 40 minutes. This report also includes information from the 12-month follow-up interview in Oakland. This interview was conducted over the telephone and lasted about 50 minutes.3 ____________ 1The other 19 cities, which were selected to represent different policy environments and labor markets, are Austin, Baltimore, Boston, Chicago, Corpus Christi, Detroit, Indianapolis, Jacksonville, Milwaukee, Nashville, New York, Newark, Norfolk, Philadelphia, Pittsburgh, Richmond, San Antonio, San Jose, and Toledo. 2The total sample includes 4,700 births (3,600 nonmarital and 1,100 marital). The data are representative of nonmarital births to parents residing in cities with populations over 200,000. The data are also representative of nonmarital births in each of the cities that make up our sample. See Reichman et al. (2001) for more information about the study’s methodology. 3Two more follow-up interviews will be conducted by phone, and the final follow-up interview will include an in-home assessment of the child’s school readiness at age four. 7 The data presented in this report are representative of nonmarital births to women in Oakland hospitals between February and June 1998.4 To generate a representative sample of births, all mothers who delivered during this time were approached in the hospitals and asked to participate in the study until both the nonmarital and marital quotas were reached. Approximately 90 percent of unmarried mothers agreed to participate in the study. Mothers were also asked to provide locating information on the fathers, and fathers were contacted either in the hospitals or as soon as possible after the child’s birth. About 75 percent of unmarried fathers participated in the study. The total baseline survey in Oakland includes interviews with 248 mothers and 189 fathers. Because most mothers we approached in the hospitals agreed to participate in the study, and because mothers in both the city’s birthing hospitals were interviewed, we are confident that the sample is representative of the population of unmarried women giving birth in Oakland hospitals.5 We are less certain that the sample of unmarried fathers is representative, as only 75 percent of these men were interviewed. Compared to average unmarried fathers, the men who were interviewed in this study are more strongly attached to the mothers of their children and differ in other ways as well.6 In the follow-up survey, 85 percent of unmarried mothers who were interviewed at baseline were also interviewed at 12 months. Seventy-one percent of unmarried fathers who were interviewed at baseline were interviewed again one year later. Therefore, the follow-up survey includes interviews with 212 unmarried mothers and 144 unmarried ____________ 4The research design assumes that families of children born in this four-month period are the same as families who have children at other times of the year. 5The sample includes unmarried mothers who gave birth in Oakland hospitals in early 1998, not the entire population of Oakland mothers who gave birth in this year. The sample does not include mothers living in Oakland who gave birth in hospitals outside the city. 6See Teitler, Reichman, and Sprachman (2001) for a description of how fathers who were interviewed differ from those who were not interviewed. 8 fathers.7 Appendix A describes how parents who participated in both surveys compare to those who dropped out of the survey at 12 months. Qualitative Study This report also includes information from qualitative interviews I conducted with a subsample of Oakland parents who participated in the survey. These interviews allowed me to explore survey themes in greater depth with the FFCW participants and to interpret their responses. All qualitative interviews were conducted between the two survey interviews (i.e., after the birth but before the child was one year old). The total qualitative sample includes 60 parents (37 unmarried parents and 23 married parents). The sample of unmarried parents had a response rate of about 86 percent for mothers and 88 percent for fathers.8 To select parents for qualitative interviews, I divided the survey sample into three groups: black mothers, Hispanic mothers of Mexican descent, and white mothers.9 I then drew a random sample of births to unmarried and married parents within race and ethnic groups.10 Although the percentages of black and Hispanic births selected for the qualitative interviews roughly reflect the proportions of these births in the larger survey, I oversampled white parents to allow for group ____________ 7Nine fathers who had not been contacted at baseline entered the sample at the 12month follow-up survey. These fathers were not included in the analysis. 8The qualitative sample is drawn from the survey sample. Of the 70 married and unmarried parents contacted for qualitative interviews, ten parents could not be located or refused to participate. I did not approach three families in cases where (1) the baby died shortly after birth, (2) the baby was put up for adoption, and (3) the parents told survey researchers that they did not want to participate in a follow-up interview. Furthermore, I did not attempt to locate fathers who did not participate in the baseline survey. 9Because only a small number of unmarried, Asian parents participated in the Oakland survey and because of the diversity of Asians within this group, I did not draw a qualitative sample of these parents. Unfortunately, Asians who did not speak English were not interviewed in the FFCW study because the expense of hiring multilingual interviewers for Asian parents was prohibitive. 10When mothers’ and fathers’ racial or ethnic characteristics differed, I selected cases on the basis of mothers’ characteristics. Five of these couples are interracial or interethnic. 9 comparisons.11 A final qualification is that the subsample selected only Mexican-American mothers who were born in the United States or had lived in this country since age 12.12 Because most of the parents of Mexican origin in my sample were born in the United States and most others had attended school in this country and had been exposed to its culture, I believed that this strategy would allow for better comparisons among parents. This analysis focuses on the 37 unmarried parents who participated in the qualitative interviews. Twenty-eight of the parents interviewed were part of a coupled sample of 14 families in which both the mother and father participated. For the remaining nine interviews, I was unable to obtain an interview with the other parent. In all, 23 families are represented in the qualitative sample of unmarried parents.13 The qualitative interviews were conducted with mothers and fathers separately (as they were in the survey) and lasted about one and one-half hours. Because questions asked in the interview were semi-structured and open-ended, the interview encouraged parents to elaborate on some of the questions and to introduce new issues for discussion. I conducted the majority of interviews in the parents’ homes. Finally, all interviews were recorded and transcribed for data analysis. Characteristics of the Oakland Sample The remainder of this chapter presents a profile of Oakland unmarried parents’ demographic, economic, and relationship ____________ 11The total qualitative sample is made up of 42 percent black mothers and 44 percent black fathers, 33 percent Hispanic mothers and 41 percent Hispanic fathers, and 24 percent white mothers and 15 percent white fathers. 12Two of these five mothers arrived at age one, two arrived at age nine, and one arrived at age 12 (she had been in the United States for eight years). Four of their partners, however, had not been in the United States since age 12. Three of them had moved to the United States within two years of the time they were interviewed; the fourth father had lived in the United States for eight years, but he moved to this country at age 26. Renzo Tragsiel (a FFCW survey interviewer) conducted these four interviews with fathers in Spanish. 13The sample of unmarried parents consists of 52 percent black mothers and 63 percent black fathers, 29 percent Hispanic mothers and 31 percent Hispanic fathers, and 19 percent white mothers and 6 percent white fathers. 10 characteristics.14 It also reports information from the combined survey sites to compare unmarried parents in Oakland with parents in the survey as a whole. Appendix A describes in more detail how the parents who were reinterviewed one year after their child’s birth compare to those parents who did not participate in the follow-up survey. Table 2.1 presents information about the demographic characteristics of unmarried parents in the Oakland sample and the total 20-city sample of the FFCW survey. Although this report focuses on parents in Oakland, it is useful to look at the two samples side by side to compare the characteristics of Oakland parents with a nationally representative sample of unmarried parents living in other large cities. Table 2.1 Demographic Characteristics of Unmarried Parents (in percent) Agea Under 20 21–25 26–30 Over 30 Race/ethnicitya White (non-Hispanic) Black (non-Hispanic) Hispanic Asian Native American Other/not classified Immigrant Have other children Total number of respondents Oakland Sample 20-City Sample Mothers Fathers Mothers Fathers 26 12 33 18 34 32 35 37 24 29 18 21 16 28 14 23 3 1 14 12 58 60 55 55 33 34 28 29 4 3 11 1 1 <1 1 1 2 12 31 33 15 16 69 57 61 56 248 189 3712 2775 NOTE: The table includes some parents who were unmarried at the time of their child’s birth but were married at year one. aPercentages may not sum to 100 because of rounding. ____________ 14See McLanahan, Garfinkel, and Waller (1999) for more information about sample characteristics at baseline. 11 As the table indicates, the largest group of unmarried mothers (34 percent) and fathers (32 percent) in the Oakland sample were between the ages of 21 and 25 at the time of their child’s birth. Slightly over onequarter of unmarried mothers were age 20 or under and slightly under a quarter were between the ages of 26 and 30. Fewer fathers than mothers were age 20 or under and more were over age 30. Unmarried parents in the Oakland sample were older than those in the total sample. The majority of parents in the Oakland and other sites had at least one biological child in addition to the child who was just born.15 The majority of parents in both samples are black or Hispanic. About six out of ten unmarried parents in Oakland are black and about one-third are Hispanic. A small percentage of unmarried parents in Oakland are white, Asian, or Native American.16 Less than 3 percent of parents identified their race as “other” or reported racial or ethnic ____________ 15On average, Oakland mothers have 2.5 children and fathers have 2.1 children (these means include the child just born). Information on the parent of previous children was not available at baseline. About 50 percent of mothers in both waves of the survey had additional children with their new baby’s father only or with the baby’s father and another partner. The other 50 percent had additional children with another partner only. For fathers in Oakland, 49 percent had additional children with the mother only or with the mother and another partner. About 51 percent had additional children with another partner only. 16Our racial and ethnic percentages are somewhat different from those reported by California Vital Statistics for the same hospitals, which show a higher proportion of births to blacks and a lower proportion of births to Hispanics and whites. The figures for unmarried births from California Vital Statistics indicate that 62 percent of children delivered in these hospitals in early 1998 were born to black mothers, 22 percent to Hispanic mothers, 10 percent to white mothers, and 7 percent to mothers of other races. The FFCW study excluded births from the sample if the parents planned to put their child up for adoption, if the father died before the birth of the child, and if the mother did not speak English or Spanish. Differential response rates by race or ethnicity were not recorded in Oakland. The differences between our sample and that of California Vital Statistics may result, in part, because our sample was likely to miss some mothers whose children were born with serious health problems and who were moved to other hospitals for care shortly after birth. We would also miss mothers whose children died immediately after birth. Although we do not know the racial composition of children with serious health problems in the two hospitals we sampled, we do know that unmarried, black mothers are more likely to have low-birth-weight children than are Hispanic and white mothers. Such children have much greater health risks, including infant mortality. See Spetz (2001, p. 82) for information about how birth weight varies by race and ethnicity in California. See Padilla and Reichman (2001) for an analysis of factors related to low birth weight in the FFCW sample. 12 information that could not be classified within one of the larger categories. The Oakland site includes a smaller percentage of whites and a somewhat larger percentage of Asians than the total sample. Compared to the larger sample, Oakland has more black and Hispanic parents. Most of the Hispanic parents in Oakland are of Mexican descent (72 percent of mothers and 84 percent of fathers). Furthermore, over 30 percent of Oakland parents are immigrants to the United States—about double the percentage in the total sample.17 Table 2.2 shows the socioeconomic characteristics of unmarried parents in the Oakland survey and total sample. The median household income of unmarried mothers and fathers in the Oakland site was close to $17,500 at the time of their child’s birth, similar to the incomes of mothers in the combined sample.18 Fathers in the larger sample, however, had higher household incomes than Oakland fathers. After household size is taken into account, over half of Oakland mothers and close to half of fathers had incomes at or below the poverty line. Again, more fathers in Oakland than fathers in the combined sample could be classified as poor. More than eight out of ten mothers and seven out of ten fathers in Oakland have incomes at or below 200 percent of the poverty line. Over 10 percent more mothers and fathers in Oakland than parents in the total sample fall into the poor or near-poor category. The results also show that the educational attainment of unmarried parents in Oakland and the total sample is generally low but that Oakland parents report lower levels of education than do parents in the whole survey. For example, about half of unmarried parents in Oakland did not have a high school diploma compared to about 40 percent in the total survey. Despite a strong economy at the time of the interview and an unemployment rate in Oakland of about 3.9 percent, only two-thirds ____________ 17The majority of unmarried, Hispanic immigrants (77 percent of mothers and 85 percent of fathers) in the Oakland study are also from Mexico. 18For cases in which household income was missing, it was imputed from the following variables: parents’ age, education, race or ethnicity, immigrant status, whether or not they worked in the previous year, number of hours worked in the previous year, earnings in the previous year, whether or not they received welfare in the previous year, and number of adults in the houshold. Alternative ways of imputing household income are currently being evaluated. 13 Table 2.2 Socioeconomic Characteristics of Unmarried Parents Education (%)a Less than high school High school only Some college College Oakland Sample 20-City Sample Mothers Fathers Mothers Fathers 52 47 41 40 31 33 34 36 15 19 23 21 2 23 3 Median household income at child’s birth ($) 17,300 17,500 17,500 22,500 Mothers who had earnings year before 35 — 67 — child’s birth (%) Mothers employed at 12-month follow-up (%) 39 — — — Poverty status (%) < 50 of line 50–100 of line 100–200 of line 20 18 23 17 31 28 20 16 31 26 28 26 Fathers employed week before child’s birth (%) — 67 — 76 Fathers employed at 12-month followup (%) — 71 — — Total number of respondents at baseline 248 189 3,712 2,775 aPercentages may not sum to 100 because of rounding. NOTE: The table includes some parents who were unmarried at the time of their child’s birth but were married at year one. of unmarried fathers in Oakland were employed in the week before their child’s birth compared to over three-quarters of unmarried fathers in the combined sites. The employment rate for unmarried fathers in Oakland who were reinterviewed was about 4 percent higher at the 12-month follow-up interview, when the general unemployment rate for Oakland had dropped to about 3.6 percent. Surprisingly, only about 35 percent of mothers in Oakland had earnings in the year prior to their child’s birth compared to 67 percent in other locations. Oakland mothers 14 continued to have low employment rates one year later (only about 39 percent were employed the week before the follow-up interview).19 Table 2.3 presents information about mothers’ and fathers’ reports of material hardship experienced between the time of their child’s birth and the 12-month follow-up interview. The most common types of hardship include having to borrow money from family and friends to pay bills, not paying the full amount of a gas or electricity bill, receiving free food or meals, not paying full amount of mortgage or rent, moving in with other people, and having phone service disconnected. Less than 10 percent of parents reported having their gas or electric service turned off, staying at a shelter, not seeing a doctor, or being evicted because there was not enough money. Although a small percentage of parents say their Table 2.3 Material Hardship Caused by a Lack of Money Among Unmarried Parents in Oakland Between Child’s Birth and 12-Month Follow-Up (in percent) Received free food or meals Children went hungry Parent went hungry Did not pay full amount of mortgage or rent Evicted from home for not paying rent/mortgage Did not pay full amount of a gas or electricity bill Service turned off by the gas or electricity company Service disconnected by the phone company Borrowed money from family/friends to help pay bills Moved in with other people Stayed at a shelter/some other place for at least a night Someone in the house did not see a doctor Mothers 17 5 11 16 2 25 9 13 32 13 6 6 Fathers 13 1 5 17 2 17 3 14 27 13 6 8 NOTE: The table includes parents who were unmarried at the time of their child’s birth but were married at year one. ____________ 19Statistical analyses show that Oakland mothers with lower levels of education were significantly less likely to have earnings the year before their baby was born or to be working the week before the 12-month follow-up interview. Other characteristics of mothers such as race/ethnicity, immigration status, and having another biological child were not associated with mothers’ employment status at either interview. Age was significant at the .10 level for employment at follow-up, indicating that older mothers were less likely than younger mothers to be employed. 15 children went hungry because of a lack of money, a larger percentage of parents reported going hungry. The majority of parents (about 58 percent) reported experiencing at least one type of material hardship in the past year. Of those experiencing hardship, most had experienced two or more types. For example, although the indicators of housing-related hardships individually range between 2 percent and 17 percent, about 26 percent of unmarried mothers and 29 percent of unmarried fathers reported some type of hardship related to housing.20 Figure 2.1 shows unmarried parents’ relationship characteristics at the time of their child’s birth in the Oakland site and combined sites of the FFCW survey. The most striking result is that the largest group of unmarried parents in Oakland were living with their baby’s other parent 60 50 49 48 40 30 38 34 Oakland 20-city Percentage 20 17 14 10 0 Cohabiting Involved in romantic Not romantically relationship not involved cohabiting NOTE: Percentages may not sum to 100 because of rounding. Figure 2.1—Relationships Between Unmarried Parents at Child’s Birth, As Reported by Mothers ____________ 20In addition, about 45 percent of unmarried parents said they had moved since the birth of their child. These parents moved 1.6 times on average between the birth and first year follow-up. For Oakland parents who moved in the first year, the number of moves ranged from one to seven times. 16 at the time of the birth. Almost half of Oakland parents were cohabiting, similar to the total sample. The next-largest group—parents who were in romantic relationships at the time of the birth but were not living together—consisted of about 38 percent of unmarried parents in Oakland and 34 percent in the total sample. A minority of parents said that they either “hardly ever” talked or did not talk to their child’s other parent; others described themselves as “just friends.” These parents were categorized as not being romantically involved at the time of birth. A somewhat smaller group fell into this category in Oakland (about 14 percent) than in the survey as a whole (about 17 percent). Therefore, we see that 87 percent of parents in Oakland were involved in romantic relationships at the time of their child’s birth compared to 82 percent in the total sample. The majority of these parents were living together. In sum, unmarried parents in Oakland with new babies had limited socioeconomic resources. About half reported incomes at or below the poverty line for their family size and about half did not have a high school diploma. Most were in their twenties and had another biological child. The majority were also black or Hispanic; over 30 percent were immigrants. Oakland parents tended to be in committed romantic relationships at the time of their child’s birth, with about half living together in informal unions. 17 3. Outcomes and Expectations in the First Year As we saw in the last chapter, unmarried parents’ involvement in relationships was high at the time of their child’s birth. This chapter examines how relationships between unmarried parents at birth compare to their relationships one year later. It then investigates how unmarried mothers and fathers assess their chances of marrying their child’s other parent and how these expectations about marriage correspond to relationship outcomes. Given the fragility of relationships between unmarried parents, it is important to understand where these changes emerge and how they are related to parents’ hopes for the future. Changes in Unmarried Parents’ Relationships Between Birth and 12-Month Follow-Up In the last chapter, we saw that most unmarried parents were involved in romantic relationships with their child’s other parent at the time of their child’s birth. About half were living together, about 35 percent were involved in romantic relationships with each other but were not living together, and only 15 percent were not romantically involved at the time of birth (Figure 3.1).1 Figure 3.1 also shows some important changes in the overall composition of Oakland parents’ relationships in the first 12 months. One year after the birth, parents are divided to a greater extent between those living in cohabiting or marital unions and those who do not have a romantic relationship with the other parent. About 7 percent of parents who were unmarried at the time of birth were married one year later (Figure 3.1). A similar percentage of parents ____________ 1The percentages differ slightly from those presented in Chapter 2 because they are shown only for families who participated in the baseline and 12-month follow-up surveys. 19 15% 50% 40% 7% 47% 35% 7% Married Cohabiting Involved in a romantic relationship but not cohabiting Not involved in a romantic relationship NOTE: Percentages may not sum to 100 because of rounding. Figure 3.1—Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers (about half) were living together at the baseline and at the 12-month interview. Cohabitation continues to be the most common relationship type at follow-up. A much larger percentage of parents reported not being in a romantic relationship with the other parent at the 12-month follow-up than at the child’s birth. Fully 40 percent of parents reported no romantic relationship with the other parent one year after their child’s birth.2 Another large change occurred in the category of unmarried parents who were not living together but had romantic relationships at birth. This group accounted for about 35 percent of parents at birth but only 7 percent of parents one year later. An important caveat is that the follow-up survey gave parents the option of reporting whether they were living together “some of the time” or “all of the time,” whereas the baseline survey simply asked whether ____________ 2About one-quarter of parents who reported no romantic relationship with the baby’s other parent at the follow-up said they had romantic relationships with other partners at that time (results not shown). 20 parents lived together.3 Therefore, we may not be comparing exactly the same type of cohabiting relationships at birth and at 12 months. Because the cohabitation variables differ, they are collapsed into one category (i.e., the baseline variable). We next look at how relationships between parents who were cohabiting, romantically involved but not cohabiting, and not romantically involved changed between birth and follow-up (Table 3.1). This analysis indicates which relationships changed the most in the first year and where these changes occurred. Among those who were living together when their child was born, about 85 percent maintained or increased their involvement with the other parent. Specifically, threequarters were still living together and an additional 10 percent were married one year later. However, about 15 percent of parents had decreased their involvement in relationships. Most of these changes occurred through Table 3.1 Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers (in percent) Married Relationship at 12-month follow-up Involved in a romantic Not relationship, romantically Cohabiting not cohabiting involved Relationship at birth Cohabiting 10 75 2 13 Involved in a romantic relationship, not cohabiting 4 27 15 55 Not romantically involved 0 0 3 97 NOTE: Percentages in rows may not sum to 100 because of rounding. ____________ 3At follow-up, about 11 percent of cohabitors were living together some of the time and 89 percent were living together all of the time. 21 relationship dissolution. About 13 percent of cohabitors reported no relationship at the 12-month follow-up interview. In comparison, about 4 percent of parents who were married at the time of birth were separated or divorced 12 months later (results not shown).4 Only 2 percent of former cohabitors were still involved in romantic relationships other than marriage or cohabitation. Since parents may be living apart from each other for a variety of reasons at 12 months, those parents did not necessarily decrease their commitment to the relationship. Unlike parents who were living together at the birth, those who were romantically involved but not living together experienced considerable changes in their relationships. In fact, only 15 percent of these parents reported being in the same type of relationship at both interviews. The greatest movement among non-cohabitors was toward relationship dissolution in the first year. Over half of these parents ended their relationship between the time of birth and the 12-month follow-up. Close to one-third of these parents appear to have increased their involvement in the relationship, with 4 percent marrying and another 27 percent moving in together. The overwhelming majority of parents who were not romantically involved at the time of birth remained uninvolved one year later.5 About 3 percent (i.e., one couple) seem to have re-initiated their relationships one year after their child’s birth, but these parents did not marry or move in together. Figure 3.2 depicts movement and non-movement in each type of relationship reported in Table 3.1. We notice that cohabiting relationships (represented in the first column) are generally stable in the first year. However, cohabitors were slightly more likely to decrease their involvement in relationships than to marry during this time. Although this group is more likely to marry than other groups of parents, marriage ____________ 4An additional parent who reported being married at the time of the birth characterized her relationship status as “cohabiting” at the follow-up interview. 5This group includes parents who said that they were just friends and those who said that they hardly or ever talked at the time of birth. The wording of the question at 12 months changed somewhat. Therefore, parents in the “no romantic relationship” category include those who said they were just friends and those who said they were not in any kind of a relationship with the father. 22 Percentage 100 13 90 2 80 70 55 60 50 75 97 40 15 30 20 27 10 10 0 43 Cohabiting Involved in Not involved at birth a romantic in a romantic relationship but relationship not cohabiting at birth at birth Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up NOTE: Percentages may not sum to 100 because of rounding. Figure 3.2—Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers rates among cohabitors are low. It will be important to watch the rate at which these relationships lead to marriage or dissolution in the coming years. Among parents who were not cohabiting at birth, the deterioration in many relationships seems to occur rapidly. Although some of these couples married and others may do so in the future, their relationships appear to be extremely fragile. For parents not in romantic relationships at birth, the likelihood of marriage, or even resuming a romantic relationship with the other parent, appears to be negligible. When we look at relationships at birth and at 12 months by race and ethnic subgroups, we see that a lower percentage of black parents (44 percent) were cohabiting at the time of the birth than Hispanic parents 23 Percentage 100 90 16 80 70 41 60 50 40 30 44 20 9 29 62 25 49 1 9 70 37 Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up 10 0 54 Black Hispanic Black Hispanic Relationship at birth Relationship at 12 months NOTE: Because only five unmarried white mothers were in Oakland sample at birth and 12 months, the percentages are not displayed. Percentages may not sum to 100 because of rounding. Figure 3.3—Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers, by Race (62 percent) (Figure 3.3).6 Conversely, more black parents (41 percent) than Hispanic parents (29 percent) were involved in romantic, noncohabiting relationships at this time. A small portion in each group were not involved in romantic relationships with the other parent. At the follow-up interview, a similar percentage of black and Hispanic parents were married. However, a larger proportion of Hispanic parents were living together at the follow-up interview than at birth, and a smaller proportion of black parents were living together. Almost half of black parents were not involved in romantic relationships one year after their child’s birth compared to one-quarter of Hispanic parents. When we look at relationship transitions, we see that a similar percentage of black parents (6 percent) and Hispanic parents (7 percent) ____________ 6The percentages are based on reports of mothers’ race. However, 12 percent of Oakland mothers who participated in both waves of the survey were in interracial or interethnic relationships. 24 who were cohabiting at baseline married in the first year after their child’s birth (Table 3.2). Although cohabiting relationships are stable for both groups, they appear to be particularly stable for Hispanic parents. Over 90 percent of Hispanic parents who were cohabiting at baseline were still living together at the 12-month interview compared to about two-thirds of black parents. Furthermore, about one-quarter of black cohabitors broke up in the first year, but few Hispanic cohabitors did so. About 20 percent of black parents moved in together in the year after their child was born; 45 percent of Hispanic parents did so. More non-cohabiting black parents retained this status in the first year than Table 3.2 Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers, by Race (in percent) Relationship at birth Cohabiting Black Hispanic Relationship at 12-month follow-up Involved in a romantic Not relationship, romantically Married Cohabiting not cohabiting involved No. 6 66 7 91 4 0 25 53 2 43 Involved in a romantic relationship, not cohabiting Black Hispanic 6 — 20 45 18 5 55 50 Not romantically involved Black — — — 100 Hispanic — — — 100 NOTE: Percentages in rows may not sum to 100 because of rounding. 49 20 19 6 25 Hispanic parents. About half of non-cohabiting, romantic relationships in both groups ended in the first year. Finally, none of the black and Hispanic parents who did not have romantic relationships at baseline reunited in the first year. Expectations About Marriage In addition to looking at transitions in relationships between unmarried parents, we are also interested in how parents view the future of their relationships at the time of birth and how closely expectations match relationship transitions. The analysis considers how expectations about marriage are related not only to marriage but also to other relationship outcomes. Previous research suggests that unmarried parents in poor communities have low expectations about marriage. Reasons for low expectations include young men’s reluctance to break from their peer groups and women’s reluctance to commit to men with negative socioeconomic or personal attributes (e.g., Anderson, 1989; Wilson, 1996). It seems likely, therefore, that unmarried parents in Oakland would have low expectations about marriage. At the same time, research on cohabiting couples has shown that most couples who live together expect to marry each other (Bumpass, Sweet, and Cherlin, 1991, p. 922). Because about half of unmarried parents in Oakland were living together, parents’ relationship status may affect overall expectations in this sample. At the time of birth, new mothers and fathers in the FFCW survey were asked what they thought the chances were that they would marry their child’s other parent. The question originally had five response categories that allowed mothers and fathers to indicate whether there was an “almost certain,” “pretty good,” “50-50,” “little,” or “no chance” of marriage in the future. For the analysis, responses are recoded into high (almost certain), medium (good/50-50), and low (little/no) categories.7 ____________ 7Conceptually, an “almost certain” chance of marriage is distinguished from the other categories because this response indicates much more confidence that a marriage will occur. The two remaining categories, “good/50-50 chance” and “little/no chance,” were grouped to reflect the distribution of responses in Oakland. 26 Figure 3.4 displays parents’ response in three bar graphs. The first bar in each set presents the most comprehensive set of responses—those of all unmarried mothers in the survey. The second bar in each set shows the responses of the subsample of unmarried mothers whose partners were interviewed (this is the same size as the number of fathers interviewed).8 Again, we were able to interview only 75 percent of unmarried fathers compared to about 90 percent of unmarried mothers. The last bar in each set presents the responses of unmarried fathers. Bars one and three compare the entire sample of mothers (including mothers whose partners were not interviewed) to a smaller and more select sample of fathers.9 Bars two and three show responses from cases where both the mother and father completed the interview. Therefore, we would expect the percentages in bars two and three (i.e., the coupled interviews) 60 50 46 42 40 34 30 37 39 38 All mothers Mothers with father interviews Fathers 29 Percentage 20 20 16 10 0 Almost certain Good/50/50 chance Little/no chance chance of marriage of marriage of marriage NOTE: Percentages may not sum to 100 because of rounding. Figure 3.4—Expectations About Marriage at Child’s Birth, As Reported by Oakland Mothers and Fathers ____________ 8One father who was in the survey did not answer this particular question, but the mother of his child was left in the analysis. 9See Teitler, Reichman, and Sprachman (2001) for more description of how fathers who were interviewed differ from those who were not interviewed. 27 to more closely resemble each other than the percentages in bars one and three. These findings show that unmarried parents in Oakland reported fairly high expectations about marriage at the time of their child’s birth. For example, about 34 percent of all unmarried mothers, 42 percent of the subsample of mothers with father interviews, and 46 percent of unmarried fathers in Oakland thought there was an almost certain chance that they would marry the other parent in the future (Figure 3.4). The fact that mothers reported lower expectations about marriage than fathers may reflect the selectivity of the fathers’ sample. These fathers seem to be in more highly involved relationships with their child’s mother than fathers not interviewed. We would expect this more select group of fathers in the survey to have higher expectations than the full sample of mothers. However, even in matched pairs, fathers appear to be more optimistic than their partners. The analysis has so far indicated that cohabiting relationships show more stability than non-cohabiting relationships in the first year. Cohabiting parents also have higher expectations about marriage (Figure 3.5). Among mothers who were living with the father when their child was born, 53 percent reported an almost certain chance of marriage (again, this group represents over half of all families in the survey). About 38 percent said there was a good or 50-50 chance of marriage. Only 8 percent of cohabitors indicated that there was little or no chance they would marry the other parent. Parents involved in romantic relationships with each other but who were not living together assessed their chances of marriage lower than cohabitors. The responses of almost half of mothers in these noncohabiting relationships appear in the middle category, indicating a good or 50-50 chance of marriage. Only 22 percent of these mothers reported an almost certain chance of marriage and about 32 percent reported little or no chance of marriage. Consistent with parents’ relationship transitions, those mothers who were not involved with the father at the time of the interview had very low expectations about marriage. Almost all mothers in this group indicated that there was little or no chance that they would marry the father. 28 Percentage 100 8 90 80 38 70 32 60 97 50 47 40 Little/no chance of marriage Good/50/50 chance of marriage Almost certain chance of marriage 30 53 20 10 22 03 Cohabiting Involved in Not involved at birth a romantic in a romantic relationship but relationship not cohabiting at birth at birth NOTE: Percentages may not sum to 100 because of rounding. Figure 3.5—Expectations About Marriage, As Reported by Oakland Mothers, by Relationship at Child’s Birth The responses to the question asking parents to rank their chances of marriage (with 1 indicating the lowest chance of marriage and 5 the highest chance) show that cohabiting parents’ mean chances of marriage are somewhat higher than 4, whereas the means of parents not romantically involved are close to 1 for mothers or 2 for fathers (Table 3.3). Means tests indicate that responses for cohabitors are significantly higher than for non-cohabitors. In this table, we are particularly interested in whether both samples of mothers give different responses than fathers within each relationship group, to see whether fathers are significantly more optimistic than mothers. The columns in Table 3.3 display parents’ expectations in three groups—the sample of all unmarried mothers, the sample of unmarried mothers whose partners were also interviewed, and the sample of unmarried fathers (as in Figure 3.4). Reading across the columns, we can again see that the responses of mothers whose partners were interviewed more closely resemble those of their partners. Although fathers’ expectations appeared to be higher than 29 Table 3.3 Means for Marriage Expectations, by Type of Relationship at Child’s Birth, Oakland All unmarried parents Cohabiting All Mothers 3.40 (1.49) Mothers with Father Interviews 3.73 (1.34) Fathers 3.88 (1.33) 4.13 (1.10) 4.19 (1.05) 4.39 (.92) Involved in a romantic relationship, not cohabiting 3.20 (1.39) 3.41 (1.29) 3.32 (1.43) Not involved in a romantic relationship 1.30 (.53) 1.31 (.48) 2.08* (1.26) No. 245 188 187 NOTES: Scale: 1 = no chance, 2 = little chance, 3 = 50/50 chance, 4 = good chance, 5 = almost certain chance. Standard errors in parentheses. *p ≤ .05. mothers’, results from means tests indicate that only fathers who were not involved with the child’s mother at the time of birth (but who agreed to be interviewed) had significantly higher expectations about marriage than the mothers of their children. Again, the fathers who were interviewed are a somewhat more select group than those who were not. Relationship Expectations versus Outcomes Given the fact that parents generally believed their chances of marriage to be high but that only 7 percent of unmarried parents were actually married one year later, it appears that unmarried parents do not anticipate well what happens in their relationships. However, parents were asked whether they expect to marry in the future not in the coming year. Although we would expect many of these changes to occur relatively soon after the birth, presumably more relationship transitions will take place after the first year. Parents’ expectations may also be an indicator of overall relationship stability, not just of the transition to marriage. In fact, parents’ expectations are closely related to whether they will marry and or live in cohabiting unions and to whether they will have no 30 romantic relationship with the other parent one year after the birth (Figure 3.6). About 82 percent of unmarried parents who reported an almost certain chance of marriage were in marital or cohabiting unions at the follow-up interview. Specifically, about 14 percent married in the first year and 68 percent were cohabiting. Only 14 percent did not have a relationship at the 12-month interview. In comparison, about 5 percent of parents who reported a good or 50-50 chance of marriage were married and slightly over half were cohabiting. About one-third did not have romantic relationship 12 months after the birth. Of those parents who reported little or no chance of marriage, about 83 percent did not have a romantic relationship with the other parent at the 12month follow-up. These results point to a clear relationship between expectations and early outcomes. These correlations are also statistically significant (Table 3.4). Results from a simple logistic regression show that the odds of Percentage 100 90 14 4 80 32 70 60 11 83 50 68 40 52 30 20 4 10 14 14 5 0 Almost certain Good/50/50 Little/no chance of chance of chance of marriage marriage marriage Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up NOTE: Percentages may not sum to 100 because of rounding. Figure 3.6—Relationships at 12-Month Follow-Up, As Reported by Oakland Mothers, by Expectations About Marriage 31 Table 3.4 Odds Ratios from Logistic Regressions of Mothers’ Reports of Being Married, Married or Cohabiting, or Not Romantically Involved at 12-Month Follow-Up Married at 12-month follow-up Almost Certain Chance of Marriage 7.44** (5.47) Little or No Chance of Marriage — Married or cohabiting at 12-month follow-up 5.63*** (2.12) .07*** (0.03) Not romantically involved at 12month follow-up .18*** (.07) 19.14*** (8.64) No. 201 201 NOTES: Reporting a little or no chance predicts failure to marry perfectly. These models control for age, education, race/ethnicity, and other biological children. Standard errors in parentheses. ** p ≤ .01. ***p ≤ .001. being married among mothers who report an almost certain chance of marriage increase over sevenfold. 10 The odds of reporting little or no chance of marriage are perfectly correlated with not marrying (i.e., no one with low expectations married in the first year). The odds of being ____________ 10These results and results in Chapter 5 are presented in terms of odds ratios. Odds are the ratio of favorable to unfavorable outcomes. Odds are expressed in terms of the probability of an outcome divided by one minus the probability of the outcome. For example, 40 percent of unmarried mothers are not involved in a romantic relationship at the 12-month follow-up. Therefore, we could say that the odds of not being involved are .40/(1-.40) = .67, or .67 to one. This means that there are .67 mothers who are not in a romantic relationship for each mother in a romantic relationship of any kind (i.e., marriage, cohabiting, romantic non-cohabiting). Odds ratios allow us to compare odds between groups we are interested in analyzing. For example, we can take the ratio of the odds for mothers who report an almost certain chance of marriage and compare it to those who do not by dividing the odds of the former group by the odds of the latter group. In the logistic regressions models examined in this report, odds ratios are interpreted net of other variables controlled for in the models. In Table 3.4 we see, for example, that the odds of being married or cohabiting at the follow-up among mothers who report an almost certain chance of marriage are 5.6 times the odds of mothers who do not report an almost certain chance of marriage. 32 married or cohabiting at the 12-month follow-up were about 5.6 times greater for mothers who reported an almost certain chance of marriage than for other mothers. However, the odds of marriage or cohabitation were about 93 percent lower for mothers who reported little or no chance of marriage. Similarly, reporting little or no chance of marriage increased the odds of not being in romantic relationship at follow-up by about 19-fold. The odds of not being in a relationship at the 12-month interview are reduced by about 82 percent for mothers who reported an almost certain chance of marriage at the time of birth. 11 In other words, higher expectations are negatively related to being uninvolved with the other parent one year after the birth. Because parents’ expectations not only show us how they assess the future of their relationships but also predict early relationship transitions, this variable serves as a proxy for relationship outcomes in Chapter 5. Conclusion Data from the FFCW survey indicate that the majority of unmarried parents in Oakland were romantically involved with each other when their children were born. In fact, close to half of parents were living together at this time. About 7 percent of parents married in the first year after their child’s birth. About half were cohabiting, similar to the proportion in the previous year. Although only 15 percent of parents ____________ 11Brown (2000) shows that cohabiting couples in which both partners (or the male partner only) reported high expectations for marriage were significantly more likely to marry than couples who did not hold these expectations. I also analyzed the outcomes of couples who held different or similar expectations about marriage. Information on transitions in the first year is available only for two cities in the FFCW survey (i.e., Oakland and Austin). This analysis is also limited to couples so that we can include both mothers’ and fathers’ reports. Logistic regression analysis shows that when mothers report a certain chance of marriage and fathers do not, when fathers report a certain chance of marriage and mothers do not, and when both parents report a certain chance of marriage (compared to neither reporting a certain chance of marriage), the odds of being married or being married or cohabiting at 12 months are increased. The effect of both parents’ reporting an almost certain chance of marriage are statistically significant. Similarly, the odds of not being in a relationship with the father are decreased when mothers are more optimistic than fathers, fathers are more optimistic than mothers, and both are optimistic about marriage. The effects of fathers’ having higher expectations than mothers and both having high expectations are statistically significant in this model. 33 said that they were not involved in a romantic relationship at the time of birth, 40 percent had no romantic relationship one year one later. Parents living in cohabiting relationships at the time of birth tended either to remain in stable relationships or to marry, whereas most parents in non-cohabiting romantic relationships dissolved their relationships. About 85 percent of parents living together at the time of birth had either maintained this relationship or had married; but cohabitors were slightly more likely to end their romantic relationship than to marry. In contrast, over half of those who were involved in romantic relationships but not living with the other parent reported no romantic relationship one year later. The number of non-cohabiting, romantic relationships that dissolved in the first year indicates the fragility of these relationships. Only 3 percent of the parents (i.e., one couple) who reported no relationship at year one seem to have reunited. As a result of these changes, parents’ relationships were more likely to be divided between those that showed a high level of involvement and those with no romantic involvement one year after the birth than they were at baseline. Cohabitation at birth seems to be the key indicator of which way unmarried parents’ relationships will move in the first year. Given parents’ high expectations for marriage at birth, the fact that only 7 percent of parents married in the first year indicates that parents’ expectations were inflated—perhaps because parents were especially optimistic at the time of their child’s birth. At the same time, parents were asked about their expectations for marriage in the future, not simply in the first year, and more transitions will likely occur. Expectations show us how parents view the future of their relationships. They are significantly related to whether parents will be living in married or cohabiting union or will not have a romantic relationship with the other parent one year after the birth. This suggests that expectations are a meaningful proxy for relationship outcomes, including, but not limited to, marriage. 34 4. Insights from Qualitative Interviews As we saw in Chapter 3, parents were divided between those in nonromantic relationships and those involved with each other one year after the birth of their child. This chapter turns to the qualitative data to examine the experiences of these two groups. For the purposes of this chapter, the category of intact romantic relationships includes parents who married before the follow-up interview, those living together at follow-up, and those involved in non-cohabiting, romantic relationships one year after the birth. As discussed in Chapter 2, I conducted openended, qualitative interviews with 37 unmarried parents who participated in the FFCW survey at the time of their child’s birth and at the 12month follow-up interview. Most of these interviews were completed for both the mother and father (this group includes 28 parents, or 14 families). In other cases, I was able to interview only one parent (this group includes nine parents, or nine families). All interviews with unmarried mothers and fathers were conducted separately. This chapter draws on qualitative interviews to illustrate some of challenges to relationship stability or increased commitment parents faced in the first year of their child’s life. The first part of this chapter asks what factors contributed to the dissolution of unmarried parents’ relationships by the time of the 12-month interview, as reported by parents themselves. The second part of the chapter investigates factors that influenced parents’ plans for the future of their relationships. In particular, it explores why couples romantically involved may be reluctant to increase their involvement through marriage. Parents Not in Romantic Relationships at Follow-Up About two out of five parents were not romantically involved with the other parent at the time of the follow-up interview. Thirty-five 35 percent of these relationships ended before the birth (or were never established), and 65 percent ended in the first year. I conducted qualitative interviews with parents in each group to find out more about what happened in their relationships to precipitate the breakup. Before turning to parents’ qualitative responses, I first look at how parents in the qualitative sample responded to a survey question about relationship dissolution. In the FFCW survey, parents were asked why their relationship with the other parent ended.1 Among parents who ended their relationships in the first year, mothers from the qualitative sample reported in the survey that their relationships dissolved because of violence/abuse, “relationship reasons,” alcohol or drugs problems, and the deportation of the father.2 In the survey, all fathers in my qualitative sample reported that their involvement with the mother ended for “relationship reasons.”3 In addition, I interviewed another couple who had already broken up when I talked to them but who did not participate in the follow-up interview. In the qualitative interview, both parents agreed that they did ____________ 1Respondents were coded in up to seven categories for why their relationship ended, including (1) financial reasons, (2) distance, (3) incarceration, (4) relationship reasons, (5) drug/alcohol problems, (6) violence/abuse, and (7) other. Parents were allowed to give more than one reason why the relationship ended. Of the 37 Oakland mothers who ended their relationship between their child’s birth and the follow-up and who answered this question in the FFCW survey, 11 percent cited distance, 73 percent cited relationship reasons, 5 percent cited drug or alcohol problems, and 11 percent cited violence or abuse as reasons why they were no longer involved with the father. Of those 21 Oakland fathers who responded to this question in the FFCW survey, 5 percent cited financial reasons, 5 percent cited the mother’s incarceration, 95 percent cited relationship reasons, and 5 percent cited abuse or violence as reasons why the relationship ended. 2 One mother I interviewed was not asked why the relationship ended because she said she was uncertain about who the father was. “Deportation” was not a response category but was written in for two mothers. 3One father was skipped from this question because the couple was still together at the time of the interview, but he was deported shortly after. Two additional fathers were not asked why their relationship ended because they revised their relationship status at birth when asked about it in the follow-up. That is, they reported that they were romantically involved with the mother in the baseline interview. However, when asked one year later about their relationship status at birth, they said they did not have a romantic relationship with the mother at that time. Their former partners, however, did classify the initial relationship as romantic in both interviews. One of these mothers said the relationship ended because of violence/abuse and the other mother attributed the breakup to “relationship issues.” 36 not have a serious romantic relationship at the time of birth and that the father reunited with the mother of his older children in the first year. Among the two parents in the qualitative sample who had broken up before the birth of their child, the first said their involvement ended for “relationship reasons” and the second parent’s response was coded “other.”4 Relationship Ended in the First Year The reasons parents gave in the qualitative interviews about why their relationships ended were basically consistent with the survey responses. The qualitative data provide more in-depth information about why couples broke up and illustrate the multiple challenges parents faced as their relationships were dissolving. These open-ended interviews are particularly helpful in allowing us to unpack the “relationship reasons” response most frequently reported in the follow-up survey. According to qualitative reports, “relationship reasons” point to high conflict in the relationship around issues such as trust, fidelity, and commitment. These interviews also help put other responses about what triggered the breakup, such as drug use, violence, and deportation, into a larger context. Furthermore, the qualitative data point to additional problems with financial and housing instability, which appear to have contributed to the dissolution of parents’ relationships. Unlike most parents who ended their relationship in the first year, Marian and Daniel were living together at the time of their child’s birth.5 In the qualitative interview, Marian attributed the couple’s breakup to an argument in which Daniel hit her. Although this serious event motivated the final breakup, Marian explains that it was preceded by prior breakups, the last of which resulted from Daniel’s “cheating.” ____________ 4Of the 24 Oakland mothers whose relationship ended before the birth and who responded to this question in the FFCW survey, about 13 percent cited financial reasons, 17 percent cited distance, 20 percent cited relationship problems, 13 percent cited drug or alcohol problems, 13 percent cited abuse or violence, and 30 percent cited other reasons why the relationship ended in the FFCW survey. Of the 11 Oakland fathers who responded to this question, 36 percent of fathers cited financial reasons, 18 percent cited distance, 9 percent cited the mother’s incarceration, 36 percent cited drug or alcohol problems, and 27 percent cited other reasons the relationship ended. 5All names used in this report are pseudonyms. 37 Marian recounted what was happening in their relationship at the time she decided to end the relationship for good: Once I got this place and was pregnant with my daughter, [the problems] started. He stayed here and it was like he would start leaving for weeks . . . and we got into an argument one night, and he hit me. . . . Before he cheated on me, so we broke up because of that. Then we were talking about getting back together and we got into an argument and he hit me, and that’s when I decided [to end things]. Marian also found it difficult to trust the father because he had been in and out of jail for selling drugs. Although she says that he promised he would not sell drugs again, he began leaving for “weeks at a time” to do so. Katherine reported that her relationship with the father ended because of problems related to drugs and alcohol. Both she and the father were sentenced to jail at the same time on drug-related charges. In the qualitative interview, Katherine explained that she did not find out she was pregnant until she was already incarcerated. The couple corresponded during the pregnancy and planned to marry when they were released from jail. Having spent half her sentence in drug rehabilitation, Katherine hoped to start a new, drug-free life when they got out. However, she broke up with the father (and broke off their engagement) two weeks after he was released, explaining that he clearly “wanted to continue in the drug scene” and she couldn’t risk going to prison again, for her own and her children’s sakes. In the survey, Leticia and Ricardo both reported that they broke up for relationship reasons. When I spoke to these parents, they had recently had a serious argument and said that their relationship was “pending.” By the follow-up interview, the relationship had dissolved. Leticia’s two major concerns about the relationship revolved around trust issues and Ricardo’s reluctance to commit to a more serious relationship. The couple had not been dating long before Leticia became pregnant, and concerns about trust went both ways. Leticia said they argued frequently about Ricardo’s relationship with another woman, whom, she believes, he was romantically involved with. During the pregnancy, she said they also broke up for a while when Ricardo expressed some doubt about whether he was the baby’s father. 38 At the time of the interview, Leticia had given Ricardo an ultimatum that she would leave him unless he agreed to settle down with her. According to Leticia, Ricardo did not act committed to the relationship and would not make plans for them to live together as a family. Leticia was struggling to raise another child alone and did not want to go through this experience again with another father. She explains, “It’s stressful because you don’t know what he really wants. . . . I don’t know if he just expects me to wait around.” Although Ricardo considered their relationship serious, he thought they argued too frequently about insignificant issues. More important, he thought it would be pointless to reconcile because they were incompatible as partners and their arguments would likely continue. Other parents ended their relationships by the time of the 12-month interview but after I interviewed them. When I spoke to these parents, however, they were experiencing obstacles to keeping their relationships together. In the survey, Norma attributes her breakup with Miguel to his recent deportation. Miguel was interviewed for the follow-up survey shortly before Norma and before he had been deported. Norma did not anticipate this event, but she discussed how his immigration status contributed to other issues that made relationship stability difficult. When I spoke to Norma and Miguel, she was working part time and Miguel was working under the table. The couple met when they were staying in the same building (as neighbors), but they were forced to move out before their child’s birth. Norma and Miguel were currently living with their families. She explained that the couple could not get their own place because she did not make enough money to get an apartment and could not put Miguel on the lease because he was not in the country legally. She applied for a Section 8 housing voucher but said she is on a very long list with other low-income families. Because the couple could not find a way to live together, Norma did not feel as if they had a foundation upon which to build a marriage or family. She explained: “We were supposed to be getting married last month. We postponed it cause we had no money. How can we be married if he lives there and I live over here?” Norma stressed that the couple argued frequently and that many of their arguments arose because Miguel was not living with her and the baby. Because they lived apart, 39 she felt as if the baby was not getting to know him and that Miguel was not helping her take care of the baby. Furthermore, she believed they had more problems with trust and communication because they lived in separate households: “He [does] not know what I’m doing, and I don’t know what he’s doing. We don’t know how to trust each other. And we got a lot of he said/ she said interactions going between us. A lot of conflicts.” Although Miguel also thought living apart was an issue in their relationship, he did not seem as dissatisfied with their living arrangements as Norma. Miguel agreed that the couple could not move in together because of financial difficulties. However, Miguel did not think that the couple argued frequently. Because Miguel’s deportation forced an end to their relationship, we do not know whether this couple could have overcome these problems and established a stable relationship. This example reminds us that many unmarried parents in Oakland are foreign-born and may face distinctive challenges in their relationships. Relationship Ended Before the Birth Lashae, who was living in a drug rehabilitation center when I interviewed her, said her relationship with her baby’s father ended two months before the baby was born. Lashae explained that they had seen each other for a few years before the pregnancy and “it got serious for a minute, but we kind of drifted away, and he was cheating.”6 Lashae described the father’s response when she found out she was pregnant: He was like—he wasn’t ready. I wasn’t really ready either, but I didn’t really agree on abortion. So I just went on and had her whether he was gonna do something or not cause I’m already a single parent, so I figured I can do it— I’m already doing it. But he said he wasn’t ready, wasn’t financially stable. But from what I’ve heard, he has had jobs before. Lashae says that the father has seen the baby only once or twice since the birth and has not offered any financial support. Although she does not want a relationship with the father, Lashae would like him to be ____________ 6In the survey, Lashae’s response about why the relationship ended did not fit into one of the response categories in the survey. Therefore, it was coded “other.” 40 involved with their child.7 Lashae did not have a job and was not receiving financial support from the baby’s father. Therefore, she was primarily concerned about how she would support her family when she lost her welfare benefits in 6 months.8 Parents in Romantic Relationships at Follow-Up As we saw in the previous chapter, about 60 percent of parents who were unmarried at birth were romantically involved with the other parent at the time of the follow-up interview. This section explores the obstacles parents experienced in their relationships during the first year. Given that many parents were in stable relationships, such as cohabitation, in the first year, it also investigates why parents postpone marriage or do not anticipate marriage. Married at Year One Only one of the unmarried couples I spoke to in the qualitative interviews married between the baseline and 12-month follow-up. Monica and Jack had lived together for about nine years before marrying and had three children together. Initially, Jack moved in with Monica and her father because he did not have a permanent place to stay (Monica was a teenager at this time). For much of their relationship, Monica and Jack had been struggling with problems such as methamphetamine addiction and financial instability. Monica and Jack each described their marriage as the result of a particular set of circumstances that converged when they “bottomed out” on drugs. Intervention from his family, Child Protective Services (CPS), their church, and their drug rehabilitation program prompted them to stop using drugs and “get themselves straight.” Because the couple already considered themselves in a common law marriage, they said it was not difficult for them to formalize their relationship. According to Jack, ____________ 7Unmarried fathers’ involvement with children is an important issue and will be the focus of a future report on the Oakland FFCW study. 8Lashae was facing time limits on welfare. 41 We’ve always wanted to get married, you know. From day one, you know. But it took us this long to finally get married, you know, because of financial reasons. . . . We’ve never been really financially stable. You know. Bank account, things like that. We’ve never had things like that . . . and then we had another child, and then another child. And we kind of put that off. We knew we were together as a couple. Common law marriage, and that was fine. . . . We were probably going through this crisis or that crisis in our life . . . every time we’d say, “Hey, let’s get married next year.” . . . Something was always going on in our life at that time, and it just never happened. I think when we finally got our acts straight last year, then we decided to set things right. But it was never, you know, like I told people, it was never a thing of I didn’t want to marry her or she didn’t want to marry me. We always felt like there were other things to do at the time. Although it is hard to generalize from Monica’s and Jack’s experience, their transition to marriage was probably greatly facilitated by the fact that they had been living together in an informal union for many years. When they stopped using drugs, Monica and Jack were also able to become more financially secure. In this case, overcoming their drug addiction seems to have had a larger effect on their lives than marrying. However, their reasons for delaying marriage may shed light on why other parents in committed relationships do not marry in the first year. Involved in Other Romantic Relationships at Year One Tammy and Jeff were living together when their child was born. When I contacted them for the qualitative interview, they were staying at a shelter with their baby until they could find affordable housing (the apartment building where they had been living was condemned). During the qualitative interview, the couple expressed high hopes for marriage. When I asked Jeff if he considered his relationship with Tammy to be serious, he replied, “I’m gonna marry her.” Jeff added that he would like to be married by next year but wanted to have more financial and residential stability before taking this step. At the same time, Jeff was optimistic that he would get a good job, similar to the ones he held in the past: I want to have things right before we marry. You know. I want to have our own place, and [be] financially established. . . . I’d love to be able to do what I want to do now. But, I look at it like this: I’ve had all these things before . . . and I’ll have it again. It’s just a matter of time. You know. And I’m a patient 42 man. Patience is a virtue. It’s nothing that can hurt you. Everything can make you stronger. Huh buddy? [to baby]. Although Tammy had no reservations about staying with Jeff, she also wanted to be sure she was “ready” and that the circumstances were “right” before they married.9 My main thing is just making sure everything’s right, and we’re ready for it. Not just one of us, but both of us. . . . Jeff makes me feel like he’s the right person. And I know he’s ready for it. I believe in time I will be ready for it. I’m just still a little young, and I got my baby. And I don’t know, saying “I do” is a big thing . . . because it’s like saying that you’re going to be with him for the rest of your life. I would love to be financially ready, you know. If I had my own place, you know. My own car instead of the bus. You know. Things like that. Despite this hesitation, Tammy felt that Jeff was the right person because he is a good father to their child and has stood by her through difficult times. For example, the couple met after Tammy had been kicked out of her own family’s home. Soon after, Jeff offered to take her in. Jeff also stayed committed to Tammy, despite his estranged wife’s attempts to reconcile with him. Susan and Timothy were also living together at both interviews in public housing. When I spoke to them, Susan had gone back to work full-time and Timothy occasionally worked under the table. Timothy also took care of the baby and Susan’s other children while she worked. Susan and Timothy talked about how financial problems and conflict in their relationship affected their outlooks on marriage. Recounting how they went back and forth about this subject, Susan says: ____________ 9In qualitative interviews, I asked parents what they believed were the most important things to consider before getting married and what makes a marriage or relationship work. In addition to love, parents emphasized the importance of financial stability (including housing stability), communication, compatibility (including the ability to get along), and trust (including fidelity). For parents with children by other partners, an additional concern was whether their baby’s other parent would accept the extra emotional and economic responsibilities they bring to the relationship. These general considerations came into play when parents talked about their own plans for the future. 43 We just talked about it [marriage]. . . . He had give me a ring, and proposed to me. And I got mad, and I broke it, and I gave it back to him, and he threw it away somewhere on the creek or somewhere. . . . And he’s talking about it again. He wants us to get married. So, I was like, okay, I want to. In the interview, Susan expressed a great deal of love for Timothy and thought they were highly compatible, referring to Timothy as her “other part.”10 However, she admits that the relationship has been rocky at times. At the end of the pregnancy, she said, “our relationship just started falling apart” and had just recently improved. Both parents attribute their interpersonal conflict to things like the other person’s “attitude,” lack of patience, sensitivity, temper, and jealousy. However, when asked about their plans to marry, Susan emphasized the couple’s lack of economic resources. Susan is optimistic about their relationship and chances that things will improve for them: “hopefully next year, something will change. . . . Maybe next year we won’t have to struggle as much.” Timothy expressed concerns about their relationship following some recent conflicts. However, when I later asked him whether they were planning to marry, he agreed with Susan that they would be married by next year. Although he was nervous about marrying someone with other children to support, he explained: “we going to probably [get married], when we get some more money. That’s the only thing keeping us from getting married.” And when asked if he could imagine changing anything to make it easier to have a relationship, Timothy responded simply: “Her a better paying job, and me a job.” Similarly, Timothy thought the most important thing to consider when thinking about marriage is: “Income. . . . The whole household income. Period.” Other parents also described how problems in their relationships, personal problems, and the stress of being new parents occurred simultaneously with economic problems. Marta and Roberto were living together at the time of their child’s birth and 12 months later. Like Tammy and Jeff, Marta took Roberto in to live with her when he had no place to stay. Roberto was hopeful about their future together, saying, “I ____________ 10Susan says “my other part” to interpret a common Spanish phrase “mi media naranja.” The English translation of this phrase is “my half orange.” 44 already feel married.” However, he acknowledged: “We got our little knick-knacks, we gotta straighten up like any relationship. But we’re gonna straighten those up.” Roberto was unemployed and Marta was receiving welfare when I spoke to them. Roberto explained the decision to delay marriage in terms of these financial difficulties saying, “Just like the position we’re at now. It’s just hard to really budget and have a really good marriage.” Like other parents, he wanted to wait for the right time to marry. However, he acknowledged that this decision could be put off indefinitely if they tried to wait for an ideal economic opportunity: We want to have a real good marriage. . . . It’s never going to be good enough once you start having kids. It doesn’t matter. They’re in school, they’re out of school. . . . It’s never the right time. Sometimes, maybe it is a bad time. There’s a lot of bills . . . all the time. You still got to budget and get through it, you know. . . . I mean, some people, they got it made. They just got it made, you know what I’m saying? Some people got it bad, real bad. . . . You just don’t leave somebody because life’s bad. Marta—who said she had been depressed since the baby was born— expressed more uncertainty about their relationship than did Roberto. Marta identified specific problems in the relationship, such as frequent arguments and Roberto’s prior problem with drugs, which concerned her. She also discussed the history of their relationship with some ambivalence: I don’t know why, but we’re still here. Still together. Roberto has a good heart. He has a lot of bad habits. A lot of things that I can’t even under . . . explain, but the main thing that I do know for a fact is that he does have a good heart. Marta characterized their relationship as highly volatile and says that Roberto is like “two different people” sometimes. According to Marta, his emotional unpredictability affected their plans for the future: At first he was like, yeah, he wanted to marry me. Then it’s like—when he turns into that other person that I don’t know, he’s like “who would want to marry you?” and then when he’s back to [being himself], he’s like “I want to marry you” again. It’s like a yo-yo. 45 Because she wanted to have a marriage that lasted “forever,” Marta said that she would consider how they are getting along, whether or not they were employed, and whether Roberto was using drugs before making any decisions about marriage. We’d have to make sure that the baby’s in a good environment as far as arguing. And, jobs - we’re working. No drugs at all. . . . If I’m going to get married, I want to stay married. I don’t want to go through the divorce. I want to be sure that it’s going to be forever, you know. Although Marta wanted things to improve for them financially and emotionally, she was unsure about this happening. Finally, Gina was romantically involved with her child’s father but was not living with him one year after the birth. Although they lived together before, they could not afford to keep the apartment. When I asked her about whether they had thought about marriage, Gina said she would prefer to be cautious rather than jump into a marriage that could easily fall apart because of financial or personal problems. She had some doubts about the father’s character and temperament and wanted to make sure that he was responsible, stable, and mature before considering marriage. She also wanted to have more confidence in her feelings for him. Gina talked about her own considerations about marriage in reference to the high incidence of divorce: Me speaking personally, I’m seeing all these divorces, and I don’t want to end up like that. I don’t want to have to do any of that whole thing. I don’t want to have to go through drag[ging] my kids to court, or whatever. And it’s just a scary thing. I want to be sure, you know. . . . People are taking more time to see if this is the person that they really want to spend the rest of their life with. And too, financially, people aren’t financially set like they want to be. . . . Not to say I couldn’t gain that after I get married, I just don’t want to go into it with a whole bunch of problems. Although couples in intact relationships were often considering marriage, and some were even planning to marry in the next year, parents suggested that they wanted to move into marriage cautiously. Because parents believed that their relationships could easily fail, they wanted to make sure their partner and the conditions were “right” for marriage and they were emotionally and financially “ready” to marry. In time, parents 46 were often hopeful that their economic situations, relationships, and personal circumstances would improve, making marriage possible. Although waiting appears to be a rational decision for couples experiencing multiple challenges to their relationship, as Roberto acknowledged, low-income, unmarried parents may have trouble ever reaching a point where they feel economically prepared for marriage. And more doubts may arise as they wait. These findings are consistent with previous research on unmarried parents’ orientations toward marriage (Luker, 1996; Wilson, 1996; Waller, 1999; Edin, 2000; Furstenberg, forthcoming). Fragile Families in the First Year This chapter concludes with a more complete description of a couple in the study who, like the majority of parents in Oakland, were living together at the time of their child’s birth and one year later. Cynthia and James had been together for close to four years and were trying hard to succeed as a family. Despite their economic difficulties, Cynthia and James talked enthusiastically about their relationship and their future together. James recounts the story of how the couple met and why their relationship is special: The first day, my momma enrolled me [in school], she was in with her momma, being enrolled, too. And I saw her that day, and we looked, and like something clicked. . . . We saw a movie, chilled one time. . . . Took her back to her house. . . . I gave her a little kiss on the cheek, gave her my phone number to call. We talked till about four in the morning. The next thing I know, I wake up with the phone still on my ear, but it’s going “uh uh uh.” I was like, man. So the next morning, I went and picked her up for school, we walked to school together. So from that day on, everything we did is how we do it now—with each other. So there ain’t no flaws to me in the relationship. It’s cool ever since that day. Christmas, it’ll be four years. All my friends look at me like “I don’t see how you do it.” You just gotta have patience. If it ain’t real love, you ain’t gonna do it. . . . All the time, my friends going “Man, you leavin’ us—you ain’t been with us no more.” “Hey man, I just trying to make this work right here. You don’t want to try to make it work?” Cause my friends, none of their relationships gonna be successful. Although the couple has been living together for a number of years, they have had difficulty maintaining a stable residence. James moved out 47 of his family’s home and into Cynthia’s home in high school, explaining that it was too crowded. When it became difficult for the couple to stay with Cynthia’s mother after it became too crowded there, the couple’s families helped them move into their own apartment. At this time, Cynthia had a one year-old son from a previous relationship and they were both attending school. However, they had to move out of one apartment because Cynthia’s welfare check was cut and they could not afford the rent. They were then evicted from another apartment after losing a lawsuit filed with other tenants against the landlord. Cynthia’s mother allowed them to use her house as a fallback, but it was difficult for them to stay indefinitely, given that this added four members (Cynthia, James, their baby, and Cynthia’s older child) to the household. Her mother’s rent was also subsidized by a Section 8 voucher and required her to report the other tenants living in the home. Cynthia explains their current living situation: He lives here, then sometimes he goes back to his mom’s, but then he comes over here every day cause he like to be around little James. . . . [My mom] say we gotta be married in order to stay together. . . . She say it alright for him to stay here . . . only if he helping me with the baby. . . . He gotta sleep in here, and I got to sleep in there with the baby. Both Cynthia and James want to keep their family together. James already acts as a step-father towards Cynthia’s older son and they think of themselves as being in a marriage-like relationship. According to James, We’re planning on getting married on her birthday. We planned it like a year ago. I really don’t think it’s gonna be too different, cause I feel like we married right now. Just by . . . how everything operate—how we work. It feel like that now. It wouldn’t really be no different. It would be a good moment, that’s probably about it. Like other parents, however, Cynthia and James said they wanted to be ready to take this step. Cynthia is serious about the relationship and thinks James’ actions show that he is too: It seems like James want to be with me and be willing to help me out. He’s doing a lot to try and get a job and make sure we get us a place to stay. But I ain’t seen no doubt [from him] cause he be around me. He want us to be together, and he make sure everyone’s happy. 48 Before marrying, Cynthia would ideally like them both to have jobs, child care, and a place to live on their own. Cynthia suggests that their educational limitations prevented them from finding steady work in the past.11 Cynthia described what she wanted for the future: My mom say, “I think you guys can make it. You all be a happy family.” . . . I think we might make it. . . . We talk about that [getting married] a lot. . . . First I want to get in school, get me a nice job, find me a nice apartment where I can [be] set and not worry about nothing. . . . The best thing for us would be [if] James would find a good paying job, and they would be willing to help us out so everybody has the benefits—dentist or whatever—and for us to have a nice house. [And] for little James to have child care. . . . There’s a lot of jobs out there, but it’s like you gotta have a lot of things in order to get that job. Like a high school diploma or GED. Or you gotta have some experience in that job. I think that would be best for us. James and Cynthia have a strong commitment to each other and want to raise their children together. However, we see that they have both positive and negative forces affecting their relationship. On the positive side, they have a close relationship, a shared commitment to raising their children, and support from their families. On the negative side, they are young, do not have education beyond high school, have difficulty finding jobs, and cannot afford to live independently as a couple. Their future may depend upon whether these positive or negative factors tip the balance of their relationship. Conclusion Unmarried parents who ended their relationships before the 12month follow-up interview sometimes mentioned a specific event, such as an incidence of violence, a drug relapse, or deportation, that triggered the breakup. Other parents talked about more general problems in their relationships, such as conflicts around issues such as trust, infidelity, and commitment. Although parents were more likely to point to a particular event or problem that prompted the breakup than to talk about economic problems, their accounts indicate that financial instability, including not having a place to stay together, occurred simultaneously ____________ 11James called me a month after the interview to report that he had found a job. 49 with these problems and most likely contributed to them. In most cases, parents were facing multiple challenges in their relationship during the time that it ended. Parents who stayed together in the first year mentioned experiencing many of the same kinds of economic, relationship, and personal problems (such as financial and housing instability, frequent conflict, trust issues, and drug problems) as parents who had did not have intact relationships one year after their child’s birth. Financial problems were mentioned more often by parents in intact relationships (when they described their plans for the future) than by those in nonintact relationship (when they identified why the relationship ended). This is perhaps because the financial problems are endemic to both groups and are not seen as prompting a breakup. Because parents who stayed together in the first year believed that their relationships could easily fail, they wanted to make sure that their partners and the conditions were “right” for marriage and they were “ready” to do so. These parents—particularly unmarried mothers who may have to assume primary responsibility for parenting—also wanted reassurance that their partners are reliable and will be there through difficult times. Often they were hopeful that their economic and personal circumstances would improve, making marriage possible. Despite these hopes, unmarried couples are likely to experience continuing economic problems. In addition, more doubts may arise as they wait. Of course, couples in the general population share some of the problems reported by these unmarried couples, which may explain why a large number of men and women are delaying marriage. However, the material hardship experienced by most parents at the Oakland site exacerbated common relationship problems and introduced new ones. It is possible that couples who stayed together had less severe or immediate problems in their relationships than couples who broke up. It is also possible that they were better able to negotiate these problems during trying times in the first year. We do know that most of the couples who remained intact during the first year were living together. It 50 is likely that these parents established closer emotional or financial ties before moving in together or after they set up a household together.12 The next chapter returns to the survey data to examine how expectations about marriage of parents in the larger sample are related to issues parents say contributed to relationship dissolution or influenced their decisions to become more involved with the other parent. ____________ 12This increased interdependence may also inhibit breakups in the future. Thanks to Laura Moye for observing that cohabiting couples are more interdependent than noncohabiting couples. 51 5. Factors Underlying Early Expectations This chapter investigates factors that underlie parents’ expectations about marriage. In particular, the analysis considers how issues identified in the qualitative interviews, such as parents’ economic circumstances, distrust, conflict and physical violence in their relationships, and problems with drugs and alcohol influence how they assess their chances of marrying each other. Expectations reflect parents’ own sense of the potential for marriage, and the factors that influence expectations are likely to be the same as those that influence outcomes. Because the question about parents’ expectations asks them to assess their chances of marriage to the other parent in the future, these results may also point to changes in parents’ relationships beyond year one. Ideally, I would also model transitions as measured by relationship status several years after the birth. These data will ultimately be part of the FFCW study, but at this point, only the 12-month follow-up data have been collected and these are the only data available for two cities.1 Because the size of the follow-up sample in Oakland is not large enough to permit multivariate regression analysis, parents’ assessments of whether they will marry their child’s other parent in the future serve as a proxy for relationship outcomes.2 The following analysis uses data from the first wave of the FFCW survey conducted at the time of birth. It also brings in information from qualitative interviews to discuss important findings. This analysis includes responses of 3,712 unmarried mothers and 2,775 unmarried fathers living in Oakland and the 19 other cities in the FFCW survey. ____________ 1Follow-up data are available in Oakland and Austin. 2When I regressed the right-hand-side variables from the expectations models on relationship outcomes at year one for the two cities for which data are available, none of the results were invalidated. 53 Because the models control for significant interactions between parents’ responses and their residence in Oakland, the 20-city results can be applied to the Oakland sample.3 The analysis compares the responses of parents who report an almost certain chance of marriage to those of parents with lower expectations, using separate logistic regression models for mothers and fathers. Regressions were also run to examine the responses of parents who report little or no chance of marriage; differences between the findings in these models are indicated at points in the text and in the footnotes. Regression models were run with and without parents’ cohabitation status. Living with the other parent is expected not only to raise marriage expectations but also to mediate the effects of other variables because parents with certain characteristics are more likely to live together. The discussion primarily focuses on the model that does not include cohabitation (summarized in Table 5.1). However, the effect of parents’ living together on their marriage expectations is addressed in the final section of this chapter and in footnotes. The means for variables Table 5.1 Factors Related to Parents’ Expectations About Marriage Father’s employment Mother has high school diploma Mother has more than high school diploma Women/men cannot be trusted to be faithful Frequent arguments (scale) Mother/father hit or slapped Mother/father has drug/alcohol problem Mothers: Fathers: Certain Chance Certain Chance + n/s + + –– –– –– –– NOTE: These results are from the models without cohabitation. + = significant, positive. – = significant, negative. n/s = not significant. ____________ 3The models were run with full interactions of parents living in Oakland and the independent variables. Significant interactions were left in the models. See Appendix Table B.2 for the means of these interaction variables. Appendix Tables B.3–B.6 indicate which interaction variables are used in each model. 54 examined in the models and the odds ratios from the logistic regressions are reported in Appendix B. 4 Men’s Employment and Education A large body of research shows a positive connection between male employment and marriage. Researchers also argue that high levels of joblessness in poor communities have rendered men less “marriageable” in the eyes of women (Wilson, 1996, 1987).5 Because most unmarried parents in the FFCW Oakland sample are poor or near-poor, we might expect women with partners in the labor force to view them as having greater potential for marriage. Fathers with jobs might also feel like they are more economically prepared for marriage. Education is an indicator of future earnings, and fathers with higher levels of education should also have higher expectations about marriage. In Oakland, about 57 percent of unmarried mothers said their child’s father was employed the week before the child’s birth, whereas about 67 percent of unmarried fathers reported being employed. 6 Fathers who were not working but were in school most of the time during the week before the interview (about 5 percent of unmarried fathers in Oakland) were examined separately because they could be considered to be better marriage candidates than those neither employed nor in school. Because almost one out of ten mothers did not report whether the father was working, a missing variable was created for this response. About 90 percent of unmarried mothers who did not report employment information on the father responded that they did not know how he spent most of his time the week before the interview. These mothers are likely to be out of contact with the father. ____________ 4See Chapter 3, footnote 11 for a description of odds and odds ratios. 5Exacerbating the shortage of employed men is an unequal sex ratio between lowincome black men and women resulting, in part, from high male incarceration and mortality rates (Tucker and Mitchell-Kernan, 1995). 6Although the models for unmarried fathers include fathers’ own reports of employment, models for unmarried mothers use women’s reports of men’s employment. This allows all unmarried mothers to be included, rather than just those with a father interview. Mothers’ perception of whether the father was employed may also influence their expectations about marriage more strongly than men’s actual employment status, to the extent these differ. 55 The model for fathers also includes information about their educational status. Fathers without a high school diploma are compared to parents with a high school diploma or GED, fathers with education beyond high school but not a college degree, and fathers with a college degree. As we saw in Chapter 2, about 47 percent of Oakland fathers had less than a high school diploma, about 33 percent had a high school diploma or GED, about 19 percent had more than a high school diploma (but not a college degree), and about 2 percent had a college degree. The analysis shows that unmarried mothers who said that their baby’s father was employed in the week before the interview have 38 percent higher odds of reporting an almost certain chance of marriage than those whose partners were not working (see Table 5.1 for a summary of regression results and Appendix B, Tables B.3–B.6, for all odds ratios).7 Among mothers, not reporting employment information for the father is negatively related to marriage expectations. The odds of indicating an almost certain chance of marriage among mothers who did not report father’s employment information are about 76 percent lower. Again, most of these mothers say that they do not know the father’s employment status. Fathers’ being in school the week before the interview does not significantly affect mothers’ expectations. Unlike the case for mothers, employment is not related to fathers’ having high expectations. However, fathers’ employment status is significantly related to fathers’ reporting little or no chance of marriage, indicating that fathers with jobs do not have low expectations about marriage. Fathers’ education does not have a significant effect in this model (which excludes cohabitation).8 Qualitative interviews shed light on how fathers’ economic characteristics influence unmarried parents’ assessments of marriage. When I asked parents about their considerations for marriage, they often emphasized the need for financial security. Whether the father has a stable job seems to represent an important part of parents’ feeling ____________ 7Fathers’ employment status is not significantly related to whether mothers report little or no chance of marriage. 8Fathers’ education is significant in the model with cohabitation. 56 financially secure. For example, Leslie lived with her baby’s father before their baby was born. He did not have a job at that time, and she explains their decision to stop living together “was more so [for] financial [reasons] . . . because he was hurt, and he wasn’t working, and I didn’t want to take on the strain of trying to keep the apartment.” She explains that she would like to make sure this job lasts before they make a decision about marriage: He has a good job, now . . . but I want him to be on the job for a while. Not to say, “Okay, I have a good job” or then when I’m tired, or “I quit, I’m gonna get another job.” I just want him to be stable, you know. Somewhere stable. And I think a lot of people are starting to think like me. . . . Once you’re married, you take on their problems . . . as far as financial and otherwise. And now I’m looking at the total picture, and long term the things you have to [think about]. . . . And it’s hard for me. I just can’t say “yes”. . . since I have kids, something’s telling me to wait. I don’t know what it is. It’s just telling me, wait. Leslie’s response is consistent with those of parents in the previous chapter: She would like to be financially ready for marriage. However, Leslie expresses doubts about the father’s financial stability despite the fact that he recently found a job, suggesting that she has additional reservations about marriage. Women’s Employment, Education, Public Assistance, and Household Income The analysis also considers how various dimensions of mother’s economic status—their employment, education, welfare use, housing assistance, and household income—affect parents’ expectations about marriage. Mothers with earnings and higher earnings potential may believe that their chances of marrying the father are better because they could contribute to the household income. Because women’s employment allows them be more independent, however, working may make marriage less desirable to them. A related hypothesis is that welfare and housing assistance acts as a disincentive to marriage, if mothers believe that they can keep benefits only by remaining single (e.g., Murray, 1984). Because women in the baseline survey had just given birth, they were not asked whether they were working during the previous week. 57 Therefore, the variable measuring women’s participation in the workforce is based on whether they had earnings in the last year. Similarly, the variable for welfare receipt indicates whether women received income from public assistance, welfare, or food stamps in the year before the birth and the household income variable is reported for the prior year. The variable for housing assistance asks if the mother is currently living in a housing project or receiving help from the federal, state, or local government to pay rent. As we saw in Chapter 2, about 35 percent of mothers in Oakland had earnings from work the year before the birth.9 About 52 percent had less than a high school diploma, 31 percent had a high school diploma or GED, 15 percent had some education beyond high school, and 2 percent had a college degree. In addition, about 53 percent of Oakland mothers received welfare or food stamps and about 24 percent lived in public or subsidized housing. Mothers’ average household income was about $19,600.10 The regression results indicate that mothers’ receipt of earnings in the previous year was not significantly related to how they assessed their own chances of marriage. Because mothers were pregnant in the prior year, their educational level may be a better measure of how they view their earnings potential. The findings show that mothers who have education beyond high school believe that their chances of marriage are higher than other mothers. The odds of reporting an almost certain chance of marriage are about 25 percent higher for mothers with a high school diploma and about 39 percent higher for mothers with more than a high school diploma but not a college degree (see Table 5.1). Results from the regression models indicate that mothers’ receipt of welfare or housing assistance is not significantly related to how parents assess their chances of marriage. The household income variable was included in the regression primarily as a control to see if there was an additional effect of receiving welfare or housing assistance after income ____________ 9In the total sample, about 67 percent of unmarried mothers had earnings the prior year. 10Household income is a continuous variable and is divided by $10,000 in this analysis. 58 was taken into account. Although mothers with higher household incomes are significantly more likely to report an almost certain chance of marriage, there is only a modest change in the odds of mothers’ giving either response. For example, the odds of reporting an almost certain chance of marriage increase about 6 percent for every $10,000 increase in mothers’ household income. Consistent with the findings for mothers, receipt of earnings, welfare, and housing assistance is not significantly related to fathers’ marriage expectations.11 Furthermore, mothers’ household income has no significant effect. Because mothers were pregnant in the last year, education may be the best indicator of future economic stability. In particular, education that leads to good paying jobs may also increase the chance that couples will be more financially prepared to marry. Before making definite plans to marry the father, Rene says she would like to receive a high school diploma: I’m going to go back to school and finish. . . . I’d rather take my high school diploma than my GED cause most of these jobs tell me to go get my diploma, then they give me a call back. . . . [By next year] I want to be out of school and I want to be able to work. In addition to looking at how economic factors influence marriage, we are also interested in the effect of trust, conflict, violence, and drug/alcohol problems on parents’ expectations. Figure 5.1 presents the odds ratios of some of the significant variables in the model for mothers who report an almost certain chance of marriage. It is important to keep in mind that positive odds ratios may look larger than negative odds ratios in this figure because odds cannot go below 100 percent in the negative direction. We can see that the effects of noneconomic factors ____________ 11Mothers’ earnings are significantly related to whether fathers report little or no chance of marriage. This finding could indicate that fathers consider mothers with earnings to be less-desirable spouses (perhaps because they are more independent). At the same time, fathers who report a little or no chance of marriage are less likely to be in romantic relationships with or living with the mother. Therefore, mothers who had children with these men may be working more often because they are not sharing resources with the father. 59 60 40 38 39 20 Percentage 0 –20 –40 –60 –80 –100 –55 –67 –88 Father employed Mother has more than high school diploma Men cannot Frequent be trusted arguments to be faithful Father hit/ slapped mother –69 Father has drug/ alcohol problem Figure 5.1—Odds Ratios for Mothers’ Reports of an Almost Certain Chance of Marriage are large. In this model, the variable indicating frequent arguments has the largest effect. Gender Distrust Parents’ general beliefs about how men and women are likely to act in relationships should also be associated with marriage expectations. In particular, scholars suggest that feelings of distrust between men and women may be heightened in economically unstable and unpredictable environments (Furstenberg, forthcoming).12 Although academic discussions of “gender distrust” have primarily focused on women’s trust issues, this analysis examines whether mothers and fathers who do not trust members of the opposite sex to be faithful—an important ____________ 12According to Furstenberg (forthcoming), a culture of gender distrust encourages women to “monitor relationships closely and detect the early warning signs of distress.” He observes that men are aware that they are being watched closely for signs of failure, and they, in turn, have a lack of confidence about their own abilities and low expectations for success in their relationships. 60 component of gender distrust—have lower expectations about marriage. Given the salience of trust in the qualitative interviews, it is likely that this issue goes beyond individual relationships. Results from the FFCW survey indicate that 36 percent of unmarried mothers in Oakland do not believe that men can be trusted to be faithful compared to about 14 percent of fathers. The regression results also show that parents who report distrust have significantly lower expectations about marriage: The odds of mothers’ reporting an almost certain chance of marriage are about 55 percent lower (see Figure 5.1). Distrust is also an important issue for fathers in the FFCW survey. Among fathers who do not trust women to be faithful, the odds of believing there is a certain chance of marriage are about 42 percent lower. In qualitative interviews, parents identified trust as one of the most important things to consider before marrying. Mothers and fathers who did not believe potential partners could be trusted were also more pessimistic about marriage. Jamila generalizes from her experiences with her children’s fathers to explain why she believes men cannot be trusted to be faithful. In part, because of these experiences, she does not want to get married. I ain’t getting married. Cause if you can’t commit, if you gonna cheat on me, if you cheated on your woman, you’ll cheat on me. . . . I’m not just saying with him [the father] . . . just in general. That’s what I’m saying, just in general. Fathers also emphasized the importance of trusting the other partner in a marriage. Like other parents, John views trust as the most important thing to consider before deciding to marry: You gotta be able to trust, you know, the person, you have to trust who you love. That’s the main thing, the trust . . . not to cheat. Not to lie about anything, you know, be open, and you know, I think that’s just the main thing right there. Relationship Conflict Couples that argue often are likely to have lower expectations about marriage, in part because they believe that their relationships are likely to dissolve and that they are incompatible as a couple. Married couples often experience a great deal of stress in their relationships after the birth 61 of a child (Cowan and Cowan, 1997). It is likely that new unmarried parents with limited economic resources would experience a particularly high level of stress and conflict, because they tend to be in less-secure interpersonal, social, and economic situations. In the regression equations, a scale is used to measure whether parents report a high level of conflict in their relationship. This scale is constructed from a series of six questions asking if couples argue often about money, spending time together, sex, the pregnancy, drinking or drug use, and being faithful—issues often cited as sources of conflict by parents in qualitative interviews.13 Oakland mothers have a mean score of .11 and fathers have a mean score of .07 on a scale from 0 to 1.14 The regression results indicate that the level of conflict in parents’ relationships does influence their expectations a great deal. Among unmarried mothers, the odds of reporting an almost certain chance of marriage decrease by about 88 percent for every unit increase in the conflict scale (see Figure 5.1). This means that as mothers’ scores on the scale increase, mothers’ expectations decrease. Among fathers, the odds ____________ 13Couples who said that they argue often about these issues were coded 1 on each variable; those who argue sometimes or never were coded 0. A scale was then created from these variables. A scale is the sum of the scores of the individual variables. In generating the scale, a score was created for every observation for which there was a response to at least one variable. The summative score was divided by the six variables over which the sum was calculated (StataCorp, 1997). The new scale variables that were generated have scores between 0 and 1. For this scale, the alpha is .60 for all mothers and .57 for all fathers in the FFCW survey. Cronbach’s alpha measures the reliability of a scale. Specifically, it measures how well the variables measure a single, unidimensional latent construct. As the average interitem correlation increases, the alpha will also increase. The alpha is also affected by the number of variables in the scale. (See the UCLA Academic Technology Services website at www.ats.ucla.edu/stat/spss/faq/alpha. html). 14Although this scale is intended to indicate an overall level of conflict, parents reported the most frequent arguments about money, spending time together, and being faithful to each other (results not shown). Examining arguments about money is another way to look at how parents’ economic circumstances affect their expectations. When frequent arguments about money were used as an independent variable (rather than as part of a scale), logistic regression results indicate that these arguments significantly reduce the odds of mothers’ reporting an almost certain chance of marriage. Among fathers, arguments about money are significantly related to diminished expectations about marriage, decreasing the odds of reporting an almost certain chance and increasing the odds of reporting little or no chance of marriage. 62 of reporting an almost certain chance of marriage decrease by about 87 percent for every unit change in the conflict scale. In the qualitative interviews, we saw that parents also expressed hesitancy about marriage when they argued frequently. According to Patricia, the frequency with which she and the father argued, and their inability to resolve these conflicts, caused her to re-think the possibility of marriage. We talked about getting married, maybe a year ago. A year ago marrying him—I probably would have. Now —I’m—I’m more mm. No. Not now. . . . Cause it’s like—we can’t get through little things, and if we’re married—we’re not ready to be married. . . . It’s just the little arguments we have—stuff that we can’t even get through. Like I said, sometimes we don’t even settle arguments. I wouldn’t want to go into a marriage like that. I would want to see things change between us first, and then maybe. You know, we could think about it, or talk about it, or maybe even do it. Just basically little stuff. Just little stuff that add up to be a lot. Violence in Relationships Qualitative research suggests that physical violence in relationships may be associated with the decision not to marry, particularly among women (Waller, 1999; Edin, 2000). To examine relationship-specific experiences, a variable was used to indicate whether the mother or father reported that the other parent has hit or slapped them in an argument. In Oakland, about 8 percent of unmarried mothers and 17 percent of unmarried fathers said that this had happened. Because this variable asks about any single incidence of violence, it does not indicate prevalence. Therefore, this measure may represent varying types of conflict—ranging from a single event to more serious patterns of abuse. The variable can, however, indicate heightened conflict and actual, or potential, danger in parents’ relationships. The results show that reporting that the other parent has hit or slapped them in an argument reduces both unmarried mothers’ and fathers’ expectations about marriage. The odds of reporting an almost certain chance of marriage are almost two-thirds lower for mothers who report an incidence of violence (see Figure 5.1). Among fathers, the odds of reporting an almost certain chance of marriage are about 39 percent 63 lower for fathers who say that the mother has hit or slapped them.15 Jackie describes her relationship with her baby’s father before they had a violent argument and how their relationship has changed since this time: We had been together for so many years, and then we had talked about getting married, talked about starting a family—well, we had already started a family. It meant a lot to me, [to be] together. . . . When he first hit me, he write me a letter saying he was sorry, he never meant for it to happen. But when he called me, he said he did it on purpose. He know what he did. He’d do it again if he has to. He’ll come up to my job to make sure I lose my job. . . . He threatened to take my kids . . . and make sure I never see them again. So that’s exactly why I just keep my distance. I had a restraining order on him . . . but I dropped it so he could see his kids. And his mother said she would come over with him to make sure nothing happened. Drug and Alcohol Problems Qualitative results in Chapter 4 suggest that having a partner with drug or alcohol problems affects the other parents’ expectations for the future. These problems may, for example, prevent the other parent from entering the workforce or sustaining relationships. Therefore, a variable was included in these models to measure whether the other parent has a problem with alcohol or drugs that interfered with work or personal relationships. A missing variable was also created for parents who did not respond to this question. About 6 percent of unmarried mothers and fathers in Oakland said that their baby’s other parent had drug or alcohol problems. Another 3 percent of Oakland mothers did not respond to this question. Most of these mothers said that they did not know whether the father had a problem. Like mothers who did not report employment information for fathers, some of these mothers are likely to be out of contact with the fathers; others may not have wanted to report this sensitive information.16 For mothers whose partners have drug or alcohol ____________ 15Unlike measures of conflict and distrust which predict high and low expectations among mothers and fathers, this measure has a more consistent effect in the models for mothers than fathers. Specifically, experiencing violence is significantly associated with mothers’ but not fathers’ reporting little or no chance of marriage. 16The correlation between not reporting employment information and not reporting drug/alcohol problems is about .26 for unmarried mothers in Oakland. 64 problem, the odds of reporting an almost certain chance of marriage are about 69 percent lower (see Figure 5.1).17 Among unmarried fathers, the odds of reporting an almost certain chance of marriage are about 38 percent lower if their child’s mother has a drug or alcohol problem.18 In the qualitative interviews, Kelly illustrates the connection between marriage expectations and fathers’ drug use. Kelly explains, “I know that’s who I’m going to marry. I just, I want him to get off drugs, and get a more stable job.” But when I asked Kelly where she saw their relationship going, she replied: If it keeps going the way it is going? Absolutely nowhere. If . . . I actually get the strength to do what I know I need to do and am supposed to do, I see our relationship as going into marriage, and being a long-term relationship. Her plan is to get her own apartment and prevent her boyfriend from living there if he continues to use drugs. She continues: His Reverend wants to marry us and whatnot. We have to get a license. But things are really rocky at this point. . . . The drugs and whatnot. I just, I totally want them out of my life and all those people out of my life. Race, Ethnicity, and Immigration In the general population, we know that blacks are less likely to marry than whites or Hispanics (see Tucker and Mitchell-Kernan, 1995). In this analysis, the expectations of unmarried white and Hispanic parents are compared to those of unmarried black parents. Whether parents were born in the United States is also included in the analysis, as the majority of Hispanics in the Oakland sample are immigrants.19 As we saw in Chapter 2, about 58 percent of unmarried mothers and 60 percent of unmarried fathers in the Oakland sample are black, about 3 ____________ 17Not reporting fathers’ drug or alcohol use is significantly associated with mothers’ indicating little or no chance of marriage, and less strongly to reducing certainty about marriage. 18Mothers’ having drug or alcohol problems is not significantly related to fathers’ reporting little or no chance of marriage. Like reporting the mother has hit or slapped them, the effect of this variable is also less consistent for fathers than mothers. 19Qualitative findings in Chapter 4 indicated that some couples’ relationships broke up in the first year because the father was deported. Although we would like to look at the effect of immigrants’ being documented, this question was not asked in the survey. 65 percent of mothers and 1 percent of fathers are white, and about 33 percent of mothers and 34 percent of fathers are Hispanic. Immigrants represent about 31 percent of mothers and 33 percent of fathers in the Oakland sample of unmarried parents. The findings indicate that the odds of reporting a high chance of marriage are about two times higher among white mothers and 1.5 times higher among Hispanic mothers than among black mothers. Being born in another country does not have a significant effect on mothers’ expectations. In models for fathers, we see some similar patterns. White and Hispanic fathers are significantly more likely than black fathers to report an almost certain chance of marriage. Unlike Hispanic parents, however, white mothers and fathers are not significantly less likely than black parents to report little or no chance of marriage.20 Again, fathers’ immigration status has no effect. Because parents did not discuss their race or ethnic status in relation to their expectations in the interviews, I do not draw on qualitative data to interpret these findings. Age and Other Children Demographic variables are included for parents’ age and having other biological children (in addition to the child that was just born). Mothers and fathers over age 20 are compared to parents age 20 and under. As we saw in Chapter 2, about 26 percent of Oakland mothers and 12 percent of Oakland fathers are age 20 or younger, about 34 percent of mothers and 32 percent of fathers are age 21–25, about 24 percent of mothers and 29 percent of fathers are age 26–30, and about 16 percent of mothers and 28 percent of fathers are over age 30. About 69 percent of Oakland mothers and 57 percent of fathers have another biological child. We might expect parents with additional children to have more motivation to form a stable partnership with the other parent for themselves and their children.21 However, the results show that age ____________ 20In fact, white mothers report significantly lower expectations about marriage in the little or no chance model after cohabitation is controlled for. White parents’ expectations appear to be divided, perhaps because more white parents report high expectations and fewer report mid-range expectations. 21Parents who have additional children with the baby’s other parent may already have a marriage-like relationship and, therefore, have higher expectations for marriage. 66 and having other children are not associated with not reporting an almost certain chance of marriage among mothers or fathers.22 Cohabitation In Oakland, about 49 percent of unmarried mothers and 60 percent of unmarried fathers were living with the other parent at the time of their child’s birth. Earlier chapters show the importance of cohabitation to the stability of unmarried parents’ relationships in the year after their child was born. Previous research suggests that cohabiting couples in the general population typically expect to marry each other and that couples who hold these expectations are more likely to marry than others (Bumpass, Sweet, and Cherlin 1991; Brown, 2000).23 Although the results of this chapter are presented for a model without cohabitation, this variable was added to a second model because it was expected to strongly influence parents’ expectations. For example, cohabitation may indicate a relatively high commitment to the relationship or the ability to set up an independent household. Perhaps not surprisingly, the second model shows that cohabitation is the strongest predictor of parents’ reporting high expectations about ________________________________________________________ Although it is important to understand the effect of having additional children with the baby’s parent rather than another parent, this question was asked only in the 12-month interview. To date, information is available only for the two cities that have completed follow-up surveys (i.e., Oakland and Austin). The follow-up data for Oakland indicate that about 50 percent of mothers had additional children with only the baby’s father or with the baby’s father and another partner, and 50 percent had additional children with another father only. About 49 percent of fathers had additional children only with the baby’s mother or with the mother and another mother, and about 51 percent had children with another mother only. Regression results from the Oakland and Austin sample indicate that having additional children with the baby’s other parent is positively related to parents’ marriage expectations. 22Mothers age 26–30 report significantly lower chances of marriage and those who have other children report significantly higher chances of marriage in the little or no chance model. 23Researchers have found that cohabiting black couples are just as likely as cohabiting white couples to expect to marry. About 70 percent of white and black cohabitors and about 64 percent of Mexican-American cohabitors expect to marry (Bumpass, Sweet, and Cherlin, 1991). However, white cohabitors are more likely than black cohabitors to actually make the transition (Manning and Smock, 1995; Brown, 2000). 67 marriage (see Appendix Tables B.3–B.6). Cohabiting increases mothers’ odds of reporting an almost certain chance of marriage over fourfold and fathers’ odds almost threefold. 24 However, cohabitation is itself an outcome and should be determined by many of the same factors that affect marriage expectations. Therefore, adding cohabitation to the model dilutes the effect of factors shown to be important determinants of parents’ marriage expectations. For example, Hispanic parents are more likely to be cohabiting, and cohabitors have higher expectations about marriage than other parents. As a result, the effect of being Hispanic is no longer significant after controlling for cohabitation status.25 Similarly, employed fathers are significantly less likely to report low expectations for marriage before but not after cohabitation is included in the model. This suggests a selection into cohabitation among employed fathers. Unmarried parents may decide to live together to test whether their relationship would be successful as a marriage. Sandra explains this decision: I think we wanted to move in together kind of to test it out, because we knew we wanted to be together, you understand what I’m saying? But at the same time, I don’t think we were ready for marriage. . . . It [moving in together] was serious. It’s still serious, but as a test. Other parents did not move in together with the intention of marrying. However, their expectations for marriage were heightened after they established a home together. One mother said that she and the father did not make an explicit decision to live together but that this happened as he began spending the night on a regular basis. ____________ 24Because information about length of cohabitation before the birth was asked only in the follow-up survey, these data are available in only two cities. The follow-up data for Oakland indicate that parents who were cohabiting at birth had been living together on average for about 1.4 years. Regression results from the Oakland and Austin sample indicate that parents who had lived together longer had higher expectations about marriage. 25Looking at both models for high and low expectations, we see that the effects of fathers’ having missing information about a drug or alcohol problem, mothers’ household income, and mothers’ having another biological child also lose significance after controlling for cohabitation. In addition, the effect of mothers’ having a drug or alcohol problem loses significance in models for fathers. 68 Like a couple nights he had spent the night, and I was getting ready for work, and I was like “You don’t have to leave, you can stay here or whatever.” . . . And then after a couple of months . . . I said, “You can spend the night, go get you a pair of clothes and you can spend the night.” His clothes started building up, building up, building up. And I said “Shoot, you might as well move in.” Kind of joking, you know, joking about it. And then he did move in. But it was like, I don’t even know how it came about. But I know he just started spending the night, spending the night. . . . It was like he gradually moved in. It wasn’t like: “Okay, go get your stuff, I want you to move in.” It wasn’t like that. Since living together, she says they have begun to think of their relationship as a marriage: “It’s more like husband and wife instead of boyfriend and girlfriend. That’s how it feels.” Several parents in Oakland noted the difficulty of marrying and raising children together when they could not afford an apartment together. About 45 percent of unmarried parents in the FFCW survey had moved since the birth of their child. On average, parents moved 1.6 times between the birth and first-year follow-up, with the number of moves ranging from one to seven. For some parents, being able to afford a place together was a prerequisite for marriage.26 The results suggest that effect of fathers’ employment may work in part through allowing parents to move in together. Mark explains that he could not afford an apartment for him and his baby’s mother after losing his job. He feels it is critical to have a place to live together before marrying: “I feel like if you take a woman’s hand in marriage, you should at least have a place for that woman to go stay. . . . We got a kid involved in this. I want to be able to take care of them.” Conclusion Consistent with the qualitative findings, the results from the logistic regression models show that unmarried parents face multiple barriers in their relationships. Socioeconomic factors, distrust between men and ____________ 26We would expect the number of times parents moved in the previous year—a measure of housing stability—to be negatively related to parents’ outlooks on marriage. However, this question was asked at the 12-month interview rather than at the time of the birth. Therefore, the regression analysis cannot look at the effect of this variable on expectations at birth. 69 women, relationship conflict and violence, and drug and alcohol problems all play a role in explaining parents’ expectations about marriage. The results show that fathers’ employment is positively related to mothers’ reporting high chances of marriage. Mothers with higher education—an indicator of employment potential—express more certainty about marriage than other mothers. Compared to the qualitative findings, the effects of economic variables in the statistical analysis are not as pronounced. This difference could result because unmarried parents in this survey show little variation in their economic and educational characteristics—that is, most are poor and have not attained high levels of education. The statistical analysis, however, is intended to examine how variations in parents’ economic status affect outcomes. As in the qualitative findings, these results show strong negative relationships between parents’ expecting to marry and certain other factors. Parents who believe that the other gender cannot be trusted to be faithful, who report a high level of conflict in their relationship, who report that the other partner has hit them, and have partners with drug or alcohol problems have lower expectations about marriage. In the regression models, these effects are more consistent than the effects of economic factors. Taken together, these economic and non-economic factors explain more of the variation in mothers’ expectations than fathers’ expectations about marriage. When cohabitation is added to the model, it is the strongest predictor of parents’ expectations about marriage. Living with the other parent consistently increases the odds that mothers and fathers will report higher chances of marrying the other parent. Because some of the same determinants of cohabitation also underlie parents’ expectations about marriage, adding cohabitation to the model also reduces the effects of variables. These findings suggest that unmarried parents perceive cohabitation to be a step toward marriage, whether or not this transition eventually occurs. Men and women who expect to marry or want to test whether their relationship is right for marriage may move in together before the birth of their child. The experience of living with the other parent may also increase parents’ expectations that they will marry. 70 6. Conclusion and Policy Implications These findings provide an early look at the stability of relationships between unmarried parents in the first year of their child’s life. They show that about half of unmarried parents in Oakland were living together when their children were born, suggesting a relatively high level of commitment in their relationships. One year later, we observed the following outcomes: • Parents were divided between those living in formal or informal unions and those who were no longer involved in romantic relationships. • Only 7 percent of unmarried parents had married, despite high expectations about marriage at the time of their child’s birth. • Relationships between parents who were cohabiting were relatively stable. About 75 percent of these parents were still living together, 15 percent had broken up or stopped living together, and 10 percent had married. • Relationships between parents who were romantically involved but not cohabiting at the time of their child’s birth experienced considerable dissolution. Over half of these relationships had ended, and although over one-quarter of these parents moved in together, few married. • With one exception, couples who were not romantically involved at the time of their child’s birth did not reunite. Taken together, these results suggest at least two conclusions. First, policies and programs to encourage marriage face serious challenges. Second, efforts to strengthen two-parent families should probably target cohabiting parents. Although these couples were not inclined to marry in the first year, their relationships seem to have the best chance of 71 providing stable families and making the transition to marriage. Programs that help couples still involved in romantic relationships overcome common obstacles may be effective in promoting relationship stability. Policymakers should also be especially careful not to implement programs that unintentionally interfere with family stability or discourage paternal involvement (National Conference of State Legislators, 2000a; Sorensen, Mincy, and Halpern, 2000).1 There seems to be little chance that parents who are not in romantic relationships with one another will marry or even reunite. However, programs could help noncustodial parents (typically fathers) remain emotionally and economically involved in their children’s lives and help unmarried parents raise their children cooperatively in the absence of a romantic relationship. These efforts are particularly important because fathers’ involvement with their children often declines sharply after the romantic relationship with the mother ends (Mincy and Pouncy, 1999; Primus and Daugirdas, 2000; Sorensen, Mincy, and Halpern, 2000).2 Fatherhood programs could also help unmarried fathers who are involved in romantic relationships with their child’s mother gain access to jobs. Under TANF, states have been given more flexibility to support programs for fathers that promote paternal support and involvement (National Conference of State Legislators, 2000b). Financial and Housing Stability Compared to the effects of personal and relationship issues in the statistical analysis, the effects of economic variables were moderate. However, the qualitative interviews show that unmarried parents were experiencing serious financial problems that made them feel less secure about the future of their relationships and more hesitant about marriage. There is also some evidence from the study that these economic effects ____________ 1See Waller and Plotnick (1999) for a discussion of the mismatch between child support policies and the situations of low-income parents. See Sorensen, Mincy, and Halpern (2000) for an explanation of how some cash and in-kind assistance programs may be disadvantageous to two-parent families. 2The issue of paternal support and involvement will be the subject of a future report on the FFCW study in Oakland. 72 work, in part, by impeding cohabitation, which has the strongest positive effect on expectations. Some state and community programs provide coordinated services to promote employment and increase earnings among low-income parents, which may assist fragile families in overcoming economic insecurity.3 Although primarily directed toward low-income custodial mothers, some county- and community-based programs in California and elsewhere provide employment services to noncustodial parents of low-income children (California Department of Social Services, 1999).4 Cash assistance programs (such as TANF) and in-kind assistance (food stamps, housing, and health coverage) for families headed by unmarried parents has also been directed primarily toward single custodial parents. To support two-parent families, new proposals call for further eliminating distinctions between single-parent and two-parent families in receiving cash assistance and broadening eligibility criteria for in-kind assistance to ____________ 3For example, states offer educational and training opportunities to allow parents to obtain better-paying jobs and to provide transitional jobs to parents who lack job skills. Child care, health care, and transportation are important supportive services that allow parents to enter the workforce and retain jobs. Improving access to Medi-Cal, Healthy Families (i.e., the state Child Health Insurance Program), food stamps, and child care for non-TANF families are regarded as important strategies to help low-income families cover their expenses while working at low-paying jobs. In addition, programs such as a state Earned Income Tax Credit and increases in the state minimum wage are intended to boost the incomes of low-income families and reward work. Finally, state efforts to improve education (especially through lowering dropout rates and increasing access to education beyond high school) could also assist fragile families in bettering their economic situations over the long run. Many of these state strategies are described at greater length in Lazere, Fremstad, and Goldberg (2000). 4The goal of helping noncustodial parents obtain employment is consistent with the purpose of promoting marriage in PRWORA, according to the Department of Health and Human Service’s guide to states: “Under this purpose, a state could help any needy parent, including a non-custodial parent or a working parent, by providing employment, job preparation, or training services. Examples of potential services include job or career advancement activities, marriage counseling, refundable earned income tax credits, child care services, and employment services designed to increase the non-custodial parent’s ability to pay child support. Activities that promote any one of the three objectives—job preparation, work, and marriage—would be consistent with this purpose” (U.S. Department of Health and Human Services, 1999, p. 11). 73 include all poor parents and their children, including noncustodial parents (Sorensen, Mincy, and Halpern, 2000).5 In the qualitative interviews, parents emphasized housing problems. These problems may be more pronounced in the Bay Area, where rents are particularly high and housing is limited, than they are in other locations. Programs that increase parents’ employment and earnings would also allow them to better afford housing. Although the statistical analysis did not show a connection between mothers’ receiving housing subsidies and parents’ expectations about marriage, there is also some evidence from welfare reform experiments that government housing subsidies have contributed to positive employment and earnings outcomes in California and elsewhere.6 In particular, housing subsides may help families stabilize their lives, enabling them to find and retain employment (Sard and Lubell, 2000). However, if parents do not cohabit or do not report that they are cohabiting because they fear losing eligibility for housing, these subsidies are unlikely to stabilize fragile families. Distrust and Conflict in Relationships Parents reported conflicts around trust, infidelity, and commitment that contributed to the dissolution of their relationships in the first year. They also expressed concerns about conflict and trust when discussing reasons for delaying marriage. In the statistical analysis, the negative effects of these variables were more consistent than those associated with economic factors, but the qualitative analysis suggests that these issues may be especially pronounced when parents are living in conditions of poverty. Compared to economic problems, conflict and distrust between parents may be less amenable to policy intervention. However, some ____________ 5See Knox, Miller, and Gennetian (2000) for a discussion of findings from the Minnesota Family Investment Program evaluation. Among other things, the Minnesota Family Investment Program had less-restrictive eligibility rules for two-parent families. The evaluation of the overall program shows small but positive effects on increasing marriage rates and larger effects on increasing marital stability and reducing domestic violence among recipients of Aid to Families with Dependent Children. 6Most housing assistance, including public housing, publicly assisted housing, and tenant-based assistance (such as Section 8 vouchers) is funded at the federal level and allocated to localities (Kingsley, 1997). 74 community-based fatherhood programs (and programs for noncustodial parents) have addressed these issues directly. Although these programs are designed primarily to increase parents’ ability to secure a job and meet their child support obligations, they also provide services to increase parents’ skills in building or maintaining family relationships. For example, some programs help unmarried mothers and fathers develop a “team parenting” plan for raising their children cooperatively. Programs also offer peer support sessions for fathers that help them better communicate with the mother and address issues of distrust between men and women more generally. TANF funds can be used for relationship skills training for noncustodial parents as well as for employment services. California’s plan for providing assistance under PRWORA states its intention to do so.7 Although fatherhood programs offer promising ways to help parents develop and maintain family relationships, they are limited in number. Furthermore, they tend to serve noncustodial fathers with child support obligations, who are more likely than other fathers to have ended their romantic relationship with their child’s mother. To be effective on a larger scale, more programs may be needed to serve unmarried parents who are still in romantic relationships. Drugs, Alcohol, and Violence Reports of drug and alcohol abuse were negatively related to relationship outcomes and marriage expectations. These effects were ____________ 7The Department of Health and Human Services’ guide on funding TANF services suggests that states could address PRWORA’s goal of encouraging two-parent family formation and maintenance by providing services such as parenting skills training, premarital and marriage counseling, mediation services, activities to promote parental access and visitation, job placement and training services for noncustodial parents, initiatives to promote responsible fatherhood and increase the capacity of fathers to provide emotional and financial support for their children, and crisis or intervention services (U.S. Department of Health and Human Services, 1999). According to the California state plan (California Department of Social Services, 1999), “Services will be provided to unemployed or underemployed noncustodial parents whose children are receiving public assistance such as CalWORKs, Food Stamps, Medi-Cal, Healthy Families or Child Support Assurance benefits. Services may include employment training, job finding skills, family mediation services, parental skill building, supportive services and applicable child support program services.” 75 more consistent for mothers than for fathers, suggesting that these issues may be a particular concern for women. As with conflict and distrust, these problems tend to coincide with economic problems. Currently, some state programs are available to help welfare recipients deal with substance abuse issues that may interfere with their ability to work or maintain employment. TANF money can also be used to fund programs for fathers that provide substance abuse services (National Conference of State Legislators, 2000b). Helping parents overcome these problems may allow them to stabilize their relationships and families. However, it would be important to proceed cautiously in encouraging parents with serious drug and alcohol problems to marry or become more involved with their families. Similarly, some community and state programs are available to help victims of violence repair their lives and support their families. Because domestic violence is considered a barrier to workforce participation, some TANF recipients now have access to domestic violence services. Other programs (such as child support enforcement) can also help ensure the safety of custodial parents and their children (National Women’s Law Center and the Center on Fathers, Families, and Public Policy, 2000). Moreover, TANF funding can also be used for fatherhood programs that address issues of domestic violence (National Conference of State Legislators, 2000b).8 However, encouraging marriage or increased involvement in relationships characterized by domestic violence would be dangerous. By knowing more about the circumstances of unmarried parents, policymakers can better discern when and how to support these fragile families. This report has focused on a critical time in family formation, but future results will deepen our understanding of the challenges these families face and guide program development and implementation. ____________ 8Proposed legislation would also encourage fatherhood programs and campaigns to address these issues. For example, the Child Support Distribution Act of 2001 (U.S. 107th Congress, 2001a) would require “a written commitment by the entity that the entity will make available to each individual participating in the project education about the causes of domestic violence and child abuse and local programs to prevent and treat abuse, education about alcohol, tobacco, and other drugs and the effects of abusing such substances.” 76 Appendix A The Fragile Families and Child Wellbeing Study This report presents information from Oakland, California—the first city surveyed in the FFCW study. The study follows families living in 20 cities from the birth of their child through age four. New mothers were interviewed at the hospital after giving birth. Fathers were interviewed either at the hospital or elsewhere as soon as possible after the birth. Follow-up interviews are being conducted 12, 30, and 48 months later; the final follow-up interview will include an in-home assessment of the child’s school readiness at age four. The data are representative of nonmarital births to parents residing in cities with populations over 200,000. The data are also representative of nonmarital births in each of the cities in the study. The total sample of 4,700 families includes 3,600 births to unmarried parents and 1,100 births to married parents. The study addresses four sets of related questions: • What are the conditions and capabilities of new unmarried parents, especially fathers? How many of these men hold steady jobs? How many want to be involved in raising their children? • What is the nature of the relationships between unmarried parents? How many of these couples are involved in stable relationships? What proportion expects to marry? What proportion is exposed to high levels of conflict or domestic violence? • What factors push new unmarried parents together? What factors pull them apart? In particular, how do public policies affect parents’ behaviors and living arrangements? • What are the long-term consequences for parents, children, and society of new welfare regulations, stronger paternity establishment and stricter child support enforcement, and changes in health care and child care financing and delivery? 77 This report focuses on a subsection of questions in the baseline and follow-up survey. These questions measure parents’ relationship status at baseline and at year one as well as their expectations about marriage and the factors that predict expectations. The text of these survey questions appears below. Survey Questions Marriage expectations M1b14 and F1b14. What do you think the chances are that you will marry [baby’s father]/[baby’s mother] in the future? No chance ........................................1 A little chance .................................. 2 A 50-50 chance ................................3 A pretty good chance ........................4 An almost certain chance .................. 5 Relationship (at birth) M1b3. Which of the following statements best describe your current relationship with [baby’s father]? We are romantically involved on a steady basis .......................1 We are involved in an on-again and off-again relationship......2 We are just friends..................................................................3 We hardly ever talk to each other ...........................................4 We never talk to each other ....................................................5 M1b8. Are you and the [baby’s father] living together now? Yes .....................................................1 No .....................................................2 Relationship (12-month follow-up) M2a7. What is your relationship with the father now? Married .............................................1 Romantically involved .......................2 Separated/divorced ............................3 Just friends ........................................4 Not in any kind of relationship .........5 M2a7a. Are you and the father currently living together? All or most of the time ......................1 Some of the time ...............................2 Rarely or never ................................. .0 78 Employment (at birth) M1j1a. In the last year did you have income from earnings? Yes..................................................... 1 No..................................................... 2 M1i6. What was [baby’s father] doing most of last week (working, going to school, else)? Working at a regular job .................... 1 Looking for work ............................... 2 Going to school ................................. 3 Unable to work .................................. 4 Other ................................................ 5 F1j1. Last week, did you do any regular work for pay? Yes..................................................... 1 No..................................................... 2 Drugs and alcohol M1g5. In the past year, has drinking or using drugs ever interfered with your work on a job or with your personal relationships? Yes..................................................... 1 No..................................................... 2 M1i9. Does [baby’s father] have problems such as keeping a job or getting along with family and friends because of alcohol or drug use? Yes..................................................... 1 No..................................................... 2 Don’t know ....................................... 3 Physical violence M1b7b & M1b13b and F1b7b & F1b13b. Thinking about your past relationship with [baby’s father]/[baby’s mother], how often would you say that: Often Sometimes Never He/she hit or slapped you when he/she was angry?......................................1............2.............3 Conflict in the relationship M1b6a-f & M1b12a-f and F1b6a-f & F1b12a-f. The following is a list of subjects on which couples often have disagreements. How often, if at all, in the last month have you and [baby’s father] had open disagreements about each of the following: 79 Often Sometimes Never Money? Would you say often, sometimes, never?...............................1..........2...............3 Spending time together?........................1..........2...............3 Sex?.......................................................1..........2...............3 The pregnancy?.....................................1..........2...............3 Drinking or drug use?............................1..........2...............3 Being unfaithful?...................................1..........2...............3 Gender distrust M1d3e and F1d3e. Please tell me whether you strongly agree, disagree, or strongly disagree with the following statement: Strongly Strongly Don’t Agree Agree Disagree Disagree Know Men/Women cannot be trusted to be faithful..............................4.............3..........2..............1...........98 Education M1i1 and F1i1. Now I’d like to ask some questions about your education and work experience. What is the highest grade or year of regular school that you have completed? No formal schooling.................................................... 1 8th grade or less........................................................... 2 Some high school (grades 9, 10, 11, and 12) ............... 3 High school diploma (completed 12th grade) .............. 4 GED ........................................................................... 5 Some college or two-year degree .................................. 6 Technical or trade school ............................................ 7 Bachelor’s degree ......................................................... 8 Graduate professional school ....................................... 9 Welfare Use M1j1. In the last 12 months, did you have income from public assistance, welfare, or food stamps? Yes .....................................................1 No .....................................................2 80 Public/Subsidized Housing M1f3. Do you live in a public housing project? Yes.................................................... 1 No.................................................... 2 M1f4. Is the federal, state, or local government helping to pay for your rent? Yes.................................................... 1 No.................................................... 2 Household Income M1j3. Thinking about your income and the income of everyone else who lives with you, what was your total household income before taxes in the past 12 months? Under $5,000 ....................................1 $5,000 to $9,999 ...............................2 $10,000 to $14,999 ...........................3 $15,000 to $19,999 ...........................4 $20,000 to $24,999 ...........................5 $25,000 to $34,999 ...........................6 $35,000 to $49,999 ...........................7 $50,000 to $74,999 ...........................8 Greater than $75,000 ........................9 Refused ............................................97 Don’t know ..................................... 98 Age M1h1 and F1h1. What is your date of birth? (month, day, year) Race/ethnicity M1h3 and F1h3. Which of these categories best describes your race. White ...............................................1 Black, African-American ...................2 Asian or Pacific Islander ...................3 Native American, Eskimo, Aleut.......4 Other ...............................................5 Don’t know ....................................98 M1h3a and F1h3a. Are you of Hispanic or Latino origin or descent? Yes.................................................... 1 No.................................................... 2 Don’t know .................................... 98 81 Country of origin M1h2 and F1h2. Were you born in the United States? Yes .....................................................1 No .....................................................2 Other children M1a12 and F1a6. Do you have any other biological children? Yes .....................................................1 No .....................................................2 Between the interview at birth and at the 12-month follow-up, the Oakland sample lost 15 percent of unmarried mothers and 29 percent of unmarried fathers. Table A.1 shows the percentage of parents retained and lost by selected demographic and socioeconomic characteristics. Reviewed below are some notable changes in characteristics of the sample between the two interview periods. About 16 percent of mothers age 21–25 and 19 percent of mothers age 26–30 who participated at baseline were not interviewed at the 12month follow-up, compared to only 12 percent of mothers age 20 and under and 8 percent of mothers age 30 and over. In addition, the study lost more mothers with less than a high school diploma (16 percent) and mothers with a high school diploma or GED (19 percent) than mothers at higher educational levels. It is important to note that 19 percent of mothers who were romantically involved with the father at the time of birth but not living with him did not participate in the follow-up interview. Among unmarried fathers in the study who were interviewed at baseline but not at 12 months, more fathers age 20 or under (30 percent) and age 26–30 (35 percent) did not participate in the follow-up than fathers in other age groups. Between the baseline and the follow-up, 32 percent of black fathers had dropped out of the study compared to 22 percent of Hispanic fathers. Another change to note is that the median household income was higher for fathers who remained in the study than for those who dropped out: $17,500 versus $12,700, respectively. Finally, 40 percent of fathers who were romantically involved with the mother at birth but not cohabiting dropped out of the study, and 31 82 Table A.1 Comparison of Demographic and Socioeconomic Characteristics of Unmarried Parents in Oakland in the Sample at 12-Month Follow-Up and Parents Not in Sample at 12-Month Follow-Up 83 20 and under 21–25 26–30 Over 30 White Hispanic Black Immigrant Another biological child Less than high school diploma High school diploma/GED More than high school diploma College degree Mother had earnings Father employed Mother receives welfare/food stamps Cohabiting Romantic relationship Not in a romantic relationship Median household income, $ Total Mothers Fathers Baseline 12-Month Sample Not in Sample Baseline 12-Month Sample Not in Sample No. of % of No. of % of No. of No. of % of No. of % of No. of Obs. Baseline Obs. Baseline Obs. Obs. Baseline Obs. Baseline Obs. 65 88 57 12 8 23 70 16 30 7 85 84 71 16 14 60 73 44 27 16 59 81 48 19 11 54 65 35 35 19 39 92 36 8 3 52 77 40 23 12 7 71 5 29 2 2 100 2 0 80 86 69 14 11 63 78 49 22 14 140 86 121 14 19 112 68 76 32 36 75 89 67 11 8 62 76 47 24 15 168 85 143 15 25 105 71 75 29 30 127 84 107 16 20 88 72 63 28 25 77 81 62 19 15 63 68 43 32 20 38 97 37 3 1 35 77 27 23 8 4 100 4 — 0 3 67 2 33 1 85 85 72 15 13 — — — —— — — — — — 125 75 94 25 31 130 87 113 13 17 — — — —— 121 88 106 12 15 113 78 88 22 25 93 81 75 19 18 63 60 38 40 25 34 91 31 9 3 13 69 9 31 4 17,300 17,300 212 16,400 36 17,500 17,500 135 12,700 54 248 85 212 15 36 189 71 135 29 54 NOTES: All parents were in the baseline sample. Nine unmarried fathers participated in the 12-month follow-up survey but not the baseline survey. These fathers were taken out of the 12-month total when calculating the percentages for characteristics of fathers in the survey at 12-month. percent who reported no romantic relationship with the mother at birth were not interviewed at follow-up. As a result, the follow-up sample of fathers is more selective of fathers in committed relationships than the baseline sample. 84 Appendix B Sample Means and Regression Results This appendix presents the means for the dependent and independent variables used in the logistic regression models (Tables B.1 and B.2). It also shows the results from these logistic regressions as odds ratios. Table B.3 reports the odds that mothers will indicate an almost certain chance of marriage and Table B.4 reports the odds that mothers will indicate little or no chance of marriage. Similarly, Table B.5 shows the odds that fathers will indicate an almost certain chance of marriage and Table B.6 shows the odds that fathers will indicate little or no chance of marriage. In Tables B.3–B.6, Model 1 is run without cohabitation and Model 2 adds cohabitation to the regressions. Table B.1 Means for Dependent Variables in Logistic Regressions Mothers Mother reported an almost certain chance of marriage Oakland .34 20-City .33 Fathers Father reported an almost certain chance of marriage Oakland 20-City .46 .47 Mother reported a Father reported a good or 50/50 good or 50/50 chance of marriage .37 .40 chance of marriage .38 .41 Mother reported a little or no chance of marriage .29 Father reported a little or no chance of .27 marriage .16 .12 85 Table B.2 Means for Independent Variables in Logistic Regressions Mothers Father employed (mother’s report) Oakland .57 Father’s employment missing (mother’s report) .09 Father in school, not employed (mother’s report) .05 Mother had earnings last year .35 Mother has less than high school diploma .52 Mother has high school diploma or GED .31 Mother has more than high school diploma .15 Mother has college degree .02 Mother receives welfare/food stamps .53 Mother lives in public/subsidized housing .24 Mother’s household income (divided by 10,000) 1.96 20-City .67 .11 .02 .67 .41 .34 .23 .03 .44 .21 2.42 Fathers Father employed Father has less than high school diploma Father has high school diploma or GED Father has more than high school diploma Father has college degree Mother had earnings last year Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Women cannot be trusted to be faithful Frequent arguments (scale) Mother hit/slapped father Oakland .67 .47 .33 .19 .02 .35 .53 .26 1.91 .14 .07 .17 20-City .76 .40 .36 .21 .03 .68 .44 .21 2.50 .14 .11 .15 Men cannot be trusted to be faithful .36 Mother has drug/ .24 alcohol problem .06 .03 86 Table B.2 (continued) Mothers Frequent arguments (scale) Father hit/slapped mother Oakland .11 .08 20-City .12 .05 Fathers Black White Hispanic Father has drug/ alcohol problem Father’s drug/ alcohol missing Immigrant .06 .06 Age 20 and under .03 .04 Age 21–25 Black .58 .55 Age 26–30 Oakland 20-City .60 .55 .01 .12 .34 .29 .33 .16 .12 .18 .32 .37 .29 .21 White Hispanic Immigrant Age 20 and under Age 21–25 Age 26–30 Age over 30 Mother has another biological child .03 .33 .31 .26 .34 .24 .16 .69 .14 Age over 30 .28 .28 Father has another biological child .57 .15 Cohabiting with .33 mother (mother’s report) .60 .35 Lives in Oakland — .18 Interaction between .16 living in Oakland and father having more than high .14 school diploma — .23 .56 .58 .07 .01 Cohabiting with father Lives in Oakland Interaction between .49 .48 living in Oakland and father having — .07 another child — .04 Interaction between living in Oakland and father having drug problem — Interaction between living in Oakland .004 and mother receiving welfare — .04 Interaction between living in Oakland and mother having earnings — Interaction between living in Oakland and mother having .02 earnings — .02 87 Table B.2 (continued) Mothers Interaction between living in Oakland and mother’s household income Oakland — 20-City .13 Fathers Interaction between living in Oakland and mother living in public/subsidized housing Oakland — 20-City .02 88 Table B.3 Odds Ratios from Logistic Regressions of Mothers’ Reports of an Almost Certain Chance of Marriage Father employed (mother's report) Father’s employment missing (mother’s report) Father in school, not employed (mother’s report) Mother had earnings last year High school diploma More than high school diloma College degree Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s Household Income (divided by 10,000) Men cannot be trusted to be faithful Frequent arguments (scale) Father hit/slapped mother Father has drug/alcohol problem Father’s drug/alcohol missing White Hispanic 1.379** (.146) .236*** (.058) 1.056 (.288) .941 (.086) 1.248* (.121) 1.389** (.156) 1.432 (.336) .942 (.082) .860 (.092) 1.058** (.019) .454*** (.049) .118*** (.033) .331*** (.098) .311*** (.085) .393* (.156) 2.031*** (.233) 1.449*** (.145) 1.318** (.146) .395*** (.099) 1.289 (.373) .910 (.087) 1.301** (.132) 1.508*** (.179) 1.703* (.422) .985 (.090) .960 (.107) 1.043* (.020) .493*** (.056) .149*** (.043) .329*** (.101) .372*** (.103) .506 (.205) 1.599*** (.194) 1.171 (.123) 89 Table B.3 (continued) Immigrant Model 1 1.210 (.156) Model 2 1.075 (.146) Age 21-25 1.147 (.117) 1.088 (.115) Age 26-30 1.012 (.126) .982 (.129) Age more than 30 .832 (.120) .787 (.119) Mother has another biological child 1.188 (.110) 1.010 (.098) Cohabiting with father — 4.253*** (.377) Lives in Oakland 1.196 (.201) 1.129 (.199) Interaction between living in Oakland and father having drug or alcohol problem 9.554** (7.072) 10.627** (8.269) No. 3,487 3,487 Pseudo R2 0.127 0.191 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of mother living in Oakland and father having drug/alcohol problem. Model 2 controls for the interaction of mother living in Oakland and father having drug/alcohol problem. *p < = .05. **p < =. 01. ***p < = .001. 90 Table B.4 Odds Ratios from Logistic Regressions of Mothers’ Reports of Little or No Chance of Marriage Father employed (mother's report) Father’s employment missing (mother’s report) Father in school, not employed (mother’s report) Mother had earnings last year High school diploma More than high school diploma College degree Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Men cannot be trusted to be faithful Frequent arguments (scale) Father hit/slapped mother Father has drug/alcohol problem Father’s drug/alcohol missing White Model 1 .893 (.100) 7.287*** (1.175) 1.181 (.333) 1.199 (.124) .831 (.088) .862 (.110) 1.171 (.319) 1.131 (.110) 1.121 (.122) .929** (.022) 2.210*** (.219) 4.043*** (.935) 2.976*** (.563) 2.548*** (.437) 3.310*** (.893) 1.010 (.137) Model 2 .971 (.117) 4.637*** (.796) .935 (.277) 1.132 (.121) .788* (.090) .772 (.106) .984 (.294) 1.077 (.113) .981 (.114) .960 (.024) 2.009*** (.216) 2.926*** (.730) 3.258*** (.698) 2.366*** (.439) 2.691*** (.761) 1.555** (.238) 91 Table B.4 (continued) Hispanic Immigrant Age 21–25 Age 26–30 Age more than 30 Mother has another biological child Cohabiting with father Lives in Oakland Interaction between living in Oakland and mother having earnings Interaction between living in Oakland and mother’s household income No. Model 1 .782* (.091) .969 (.143) 1.004 (.115) 1.674*** (.229) 1.289 (.203) .707*** (.074) — .689 (.200) .407* (.165) 1.302** (.130) 3,487 Model 2 1.026 (.131) 1.201 (.198) 1.084 (.134) 1.753*** (.263) 1.334 (.228) .890 (.102) .100*** (.012) .618 (.181) — 1.208 (.124) 3,487 Pseudo R2 0.176 0.294 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of mother living in Oakland and having earnings and mother living in Oakland and household income. Model 2 controls for the interaction of mother living in Oakland and household income. *p < = .05. **p < = .01. ***p < = .001. 92 Table B.5 Odds Ratios from Logistic Regressions of Fathers’ Reports of an Almost Certain Chance of Marriage Father employed Father has high school diploma or GED Father has more than high school diploma Father has college degree Mother had earnings Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Women cannot be trusted to be faithful Frequent arguments (scale) Mother hit/slapped father Mother has drug/alcohol problem White Hispanic Immigrant Age 21–25 Age 26–30 Age over 30 Model 1 1.161 (.122) 1.041 (.105) 1.195 (.147) .831 (.204) 1.050 (.100) .842 (.076) .837 (.093) 1.022 (.020) .581*** (.075) .125*** (.035) .610*** (.076) .615* (.155) 2.448*** (.345) 1.438*** (.154) .959 (.127) 1.210 (.150) 1.241 (.178) 1.276 (.186) Model 2 1.121 (.122) 1.071 (.111) 1.272* (.158) .924 (.235) 1.024 (.100) .877 (.084) .905 (.104) 1.013 (.020) .590*** (.078) .161*** (.046) .601*** (.076) .684 (.176) 1.914*** (.280) 1.202 (.133) .874 (.119) 1.124 (.143) 1.129 (.166) 1.146 (.171) 93 Table B.5 (continued) Father has another biological child Model 1 .982 (.094) Model 2 .898 (.088) Cohabiting with mother (mother’s report) — 2.781*** (.254) Lives in Oakland .729 (.202) 1.074 (.339) Interaction between living in Oakland and having more than high school diploma 2.374 (1.161) — Interaction between living in Oakland and having another child 1.692 (.592) 1.939 (.714) Interaction between living in Oakland and mother — .495* receiving welfare (.181) No. 2,543 2,543 Pseudo R2 0.069 0.106 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of father living in Oakland and having more than a high school diploma and father living in Oakland and having another child. Model 2 controls for the interaction of father living in Oakland and having another child and father living in Oakland and mother receiving welfare. *p < = .05. ***p < = .001. 94 Table B.6 Odds Ratios from Logistic Regressions of Fathers’ Reports of Little or No Chance of Marriage Father employed Father has high school diploma or GED Father has more than high school diploma Father has college degree Mother had earnings Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Women cannot be trusted to be faithful Frequent arguments (scale) Mother hit/slapped father Mother has drug/alcohol problem White Hispanic Immigrant Age 21–25 Age 26–30 Age over 30 Model 1 .721* (.106) .747 (.116) .972 (.177) 1.549 (.517) 1.369* (.212) .965 (.133) .987 (.171) .981 (.031) 1.735*** (.284) 13.077*** (4.167) 1.212 (.198) .846 (.319) .713 (.163) .706* (.122) 1.262 (.261) .831 (.155) .927 (.198) 1.011 (.217) Model 2 .785 (.120) .740 (.119) .996 (.187) 1.280 (.448) 1.358* (.216) .933 (.133) .999 (.168) .995 (.033) 1.601** (.273) 9.842*** (3.273) 1.259 (.212) .642 (.252) 1.109 (.270) .954 (.172) 1.476 (.320) .930 (.179) 1.058 (.236) 1.200 (.270) 95 Table B.6 (continued) Father has another biological child Model 1 .920 (.130) Model 2 1.040 (.154) Cohabiting with mother (mother’s report) — .182*** (.028) Lives in Oakland 1.574 (.578) 2.391** (.737) Interaction between living in Oakland and mother having earnings .222** (.136) .259* (.162) Interaction between living in Oakland and mother living in public/subsidized housing 3.010* (1.554) — No. 2,543 2,543 Pseudo R2 0.071 0.146 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of father living in Oakland and mother having earnings and father living in Oakland and mother living in public/subsidized housing. Model 2 controls for the interaction of father living in Oakland and mother having earnings. *p < = .05. **p < = .01. ***p < = .001. 96 Bibliography Anderson, Elijah, “Sex Codes and Family Life Among Poor Inner-City Youths,” Annals of the American Academy of Political and Social Science, Vol. 501, 1989, pp. 59–78. Brown, Susan L., “Union Transitions Among Cohabitors: The Significance of Relationship Assessments and Expectations,” Journal of Marriage and the Family, Vol. 62, No. 3, 2000, pp. 833–846. Bumpass, Larry, and Hsien-Hen Lu, “Trends in Cohabitation and Implications for Children’s Family Contests in the U.S.,” Population Studies, Vol. 54, 2000, pp. 29–41. Bumpass, Larry L., James A. Sweet, and Andrew Cherlin, “The Role of Cohabitation in Declining Rates of Marriage,” Journal of Marriage and the Family, Vol. 53, 1991, pp. 913–927. California Department of Social Services, State Plan for the Provision of Public Assistance Under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, October 10, 1999, available at www.dss.cahwnet.gov/whatnew/default.htm (September 12, 2000). Center on Fathers, Families, and Public Policy, Policy Briefing, Vol. 3, No. 3, April 2000, available at www.cffpp.org/faxbriefings/ apriltwoonefaxbriefing.htm (May 9, 2001). Cowan, Carolyn Pape, and Philip A. Cowan, “Working with Couples During Stressful Transitions,” in Soolly Dreman (ed.), The Family on 97 the Threshold of the 21st Century: Trends and Implications, Lawrence Erlbaum Associates, Mahwah, New Jersey, 1997. Curtin S. C., and J. A. Martin, Births: Preliminary Data for 1999, National Vital Statistics Reports, Vol. 24, No. 14, National Center for Health Statistics, Hyattsville, Maryland, 2000. Edin, Kathryn, “What Do Low-Income Single Mothers Say About Marriage?” Social Forces, Vol. 47, No. 1, 2000, pp. 112–133. Furstenberg, Frank F., “The Fading Dream: Prospects for Marriage in the Inner-City,” in Elijah Anderson and Dave Massey (eds.), Problem of the Century: Racial Stratification in the United States at Century’s End, Russell Sage, New York, Forthcoming. Government Executive Magazine Daily Briefing, “Beefing Up the Bureaucracy to Promote Marriage,” February 5, 2001, available at www.govexec.com (February 7, 2001). Kingsley, G. Thomas, Federal Housing Assistance and Welfare Reform: Uncharted Territory, The Urban Institute, Washington, D.C., 1997, available at www.urban.org. Knox, Virginia, Cynthia Miller, and Lisa A. Gennetian, Reforming Welfare and Rewarding Work: Final Report on the Minnesota Family Investment Program, Manpower Demonstration Research Corporation, 2000, available at www.mdrc.org/RecentPublications. htm. Lazere, Ed, Shawn Fremstad, and Heidi Goldberg, States and Counties Are Taking Steps to Help Low-Income Working Families Make Ends Meet and Move Up the Economic Ladder, Center on Budget and Policy Priorities, November 27, 2000, available at www.cbpp.org/ 11-27sfp.htm (January 1, 2001). 98 Lerman, Robert, and Theodora J. Ooms (eds.), Young Unwed Fathers: Changing Roles and Emerging Policies, Temple University Press, Philadelphia, Pennsylvania, 1993. Luker, Kristin, Dubious Conceptions: The Politics Of Teenage Pregnancy, Harvard University Press, Cambridge, Massachusetts, 1996. Manning, Wendy D., and Pamela J. Smock, “Why Marry? Race and the Transition to Marriage among Cohabitors,” Demography, Vol. 32, No. 4, 1995, pp. 509–520. McLanahan, Sara, Irv Garfinkel, and Maureen Waller, “The Fragile Families and Child Wellbeing Study: Oakland, California,” 1999, available at www.ppic.org/publications/occasional.html. McLanahan, Sara, and Gary Sandefur, Growing Up with a Single Parent: What Hurts, What Helps, Harvard University Press, Cambridge, Massachusetts, 1994. Milbank, Dana, “Fixing Families Is a Priority,” The Washington Post, Weekly Addition, April 23–29, 2001. Mincy, Ronald B., and Hillard Pouncy, “There Must Be 50 Ways to Start a Family,” in Wade F. Horn, David Blankenhorn, and Mitchell B. Pearlstein (eds.), The Fatherhood Movement: A Call To Action, Lexington Books, Lanham, Maryland, 1999. Moynihan, Daniel Patrick, The Negro Family: The Case for National Action, U.S. Department of Labor, Office of Policy Planning and Research, Washington, D.C., 1965. Murray, Charles, Losing Ground: American Social Policy, 1950–1980, Basic Books, New York, 1984. 99 National Women’s Law Center and the Center on Fathers, Families, and Public Policy, Improving Paternity Establishment Practices and Procedures for Low-Income Mothers, Fathers, and Children, Washington, D.C., and Madison, Wisconsin, 2000. Padilla, Yolanda C., and Nancy E. Reichman, “Low Birthweight: Do Unwed Fathers Help?” Children and Youth Services Review, Vol. 23, No. 4, 2001, pp. 505–530. Primus, Wendell, and Kristin Daugirdas, Improving Child Well-Being by Focusing on Low-Income Non-Custodial Parents in Maryland, The Abel Foundation, Baltimore, Maryland, 2000. Reichert, Dana, “Why Focus on Fathers?” in Connecting Low Income Fathers and Families: A Guide to Practical Policies, National Conference of State Legislators, 2000a, available at www.calib.com/ peerta/policies/pdf/whyfocus.pdf (April 19, 2001). Reichert, Dana, “Financing Fatherhood Programs,” in Connecting Low Income Fathers and Families: A guide to Practical Policies, National Conference of State Legislators, 2000b, available at www.calib.com/ peerta/policies/pdf/financin.pdf (April 19, 2001). Reichman, Nancy, Julien Teitler, Irwin Garfinkel, and Sara McLanahan, “The Fragile Families and Child Wellbeing Study: Background, Research Design, and Sampling Issues,” Children and Youth Services Review, Vol. 23, No. 4, 2001, pp. 303–326. Sard, Barbara, and Jeff Lubell, The Value of Housing Subsidies to Welfare Reform Efforts, Center on Budget and Policy Priorities, February 24, 2000, available at www.cbpp.org/2-24-00hous.htm (January 22, 2001). 100 Sorensen, Elaine, Ronald Mincy, and Ariel Halpern, Redirecting Welfare Policy Toward Building Strong Families, The Urban Institute, Washington, D.C., 2000, available at http://www.urban.org. Spetz, Joanne, “Health Outcomes,” in Belinda Reyes (ed.), A Portrait of Race and Ethnicity in California, Public Policy Institute of California, San Francisco, California, 2001. Stacey, Judith, “Backward Toward the Postmodern Family: Reflections on Gender, Kinship, and Class in the Silicon Valley,” in A. Wolfe, (ed.), America at Century’s End, University of California Press, Berkeley, California, 1991. StataCorp, Stata Statistical Software: Release 5.0, Stata Corporation, College Station, Texas, 1997. Teitler, Julien O., Nancy E. Reichman, and Susan Sprachman, “CostBenefit Analysis of Improving Response Rates for a Hard-to-Reach Population,” presentation at the American Association of Public Opinion Research Annual Meeting, Montreal, May 17–20, 2001. Thornton, Arland, and Deborah Freedman, “Changing Attitudes Toward Marriage and Single Life,” Family Planning Perspectives, Vol. 14, 1982, pp. 297–303. Tucker, Belinda M., and Claudia Mitchell-Kernan (eds.), The Decline in Marriage Among African-Americans, Russell Sage, New York, 1995. UCLA Academic Technology Services, SPSS Frequently Asked Questions: What Does Cronbach’s Alpha Mean? available at www.ats.ucla.edu/stat/ spss/faq/alpha.html (March 26, 2001). 101 U.S. 104th Congress, The Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Public Law 104-193, U.S. Government Printing Office, August 22, 1996, available at www.thomas.loc.gov (August 29, 2000). U.S. 107th Congress, The Child Support Distribution Act of 2001, H.R. 1471, April 4, 2001a, available at www.thomas.loc.gov (May 9, 2001). U.S. 107th Congress, The Strengthening Working Families Act of 2001, S. 685, April 3, 2001b, available at www.thomas.loc.gov (May 9, 2001). U.S. 107th Congress, The Responsible Fatherhood Act of 2001, H.R. 1300, March 29, 2001c, available at www.thomas.loc.gov (May 9, 2001). U.S. Department of Health and Human Services, Helping Families Achieve Self-Sufficiency: A Guide on Funding Services for Children and Family Through the TANF Program, Washington, D.C., December 21, 1999, available at www.acf.dhhs.gov/programs/ofa/funds2.htm (September 11, 2000). U.S. Office of Budget and Management, Budget of the United States, Fiscal Year 2000, available at www.whitehouse.gov/omb/budge/ index.html (April 12, 2001). Ventura, S. J., and C. Bachrach, Nonmarital Childbearing in the United States, 1940–99, National Vital Statistics Reports, Vol. 48, No. 16, National Center for Health Statistics, Hyattsville, Maryland, 2000. Ventura, S., C. Bachrach, L. Hill, K. Kaye, P. Halcomb, and E. Koff, “The Demography of Out-of-Wedlock Childbearing,” in Report to Congress on Out-of-Wedlock Childbearing, U.S. Department of Health and Human Services, (PHS) 95-1257, Hyattsville, Maryland, 1995. 102 Waller, Maureen R., “Meanings and Motives in New Family Stories: The Separation of Reproduction and Marriage Among Low-Income, Black and White Parents,” in M. Lamont (ed.), The Cultural Territories of Race: Black and White Boundaries, University of Chicago Press and the Russell Sage Foundation, Chicago and New York, 1999. Waller, Maureen, and Robert Plotnick, Child Support and Low-Income Families: Perceptions, Practices, and Policy, Public Policy Institute of California, San Francisco, California, 1999. Wilson, William Julius, The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy, University of Chicago Press, Chicago, Illinois, 1987. Wilson, William Julius, When Work Disappears: The World of the New Urban Poor, Alfred A. Knopf, New York, 1996. 103 About the Author MAUREEN WALLER Maureen Waller is a sociologist with research interests in poverty, the family, culture, and social policy. Her previous work uses qualitative methods to examine how low-income families reconcile informal support with the requirements of the welfare and child support systems. She holds a B.A. in political science from the University of Dayton and an M.A. and Ph.D. in sociology from Princeton University. 105 Other Related PPIC Publications Child Support and Low-Income Families: Perceptions, Practices, and Policy Maureen Waller, Robert Plotnick Expensive Children in Poor Families: The Intersection of Childhood Disabilities and Welfare Marcia K. Meyers, Henry E. Brady, Eva Y. Seto The Fragile Families and Child Wellbeing Study: Oakland, California Sara McLanahan, Irwin Garfinkel, Maureen Waller PPIC publications may be ordered by phone or from our website (800) 232-5343 [mainland U.S.] (415) 291-4400 [Canada, Hawaii, overseas] www.ppic.org 107" } ["___content":protected]=> string(102) "

R 701MWR

" ["_permalink":protected]=> string(103) "https://www.ppic.org/publication/unmarried-parents-fragile-families-new-evidence-from-oakland/r_701mwr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8144) ["ID"]=> int(8144) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:35:13" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3267) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 701MWR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_701mwr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_701MWR.pdf" ["wpmf_size"]=> string(6) "659689" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(223209) "Unmarried Parents, Fragile Families: New Evidence from Oakland ••• Maureen R. Waller 2001 PUBLIC POLICY INSTITUTE OF CALIFORNIA Library of Congress Cataloging-in-Publication Data Waller, Maureen Rosamond, 1967– Unmarried parents, fragile families : new evidence from Oakland / Maureen R. Waller. p. cm. Includes bibliographical references. ISBN 1-58213-035-3 1. Single-parent families—California—Oakland. 2. Children of single parents—California—Oakland. I. Title. HQ777.4 .W35 2001 306.85'6'0979466—dc21 2001034951 Copyright © 2001 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword In 1997, PPIC was a leading financial supporter of the Fragile Families and Child Wellbeing (FFCW) survey, a national, longitudinal study of unmarried parents and their children. About 33 percent of all U.S. and California children are currently born to unmarried mothers— about three times the corresponding proportion in 1970. Partly because unmarried parents and their children present difficult public policy challenges—for example, the children are more likely to be poor and to receive welfare—PPIC welcomed the opportunity to support a national study on the topic. In the aftermath of the 1996 Personal Responsibility and Work Opportunity Reconciliation Act, the study seemed especially timely. As one of the first products of that study, Maureen Waller’s Unmarried Parents, Fragile Families: New Evidence from Oakland underscores the importance of these policy challenges to California and the nation. Two results are especially noteworthy. First, less than 10 percent of these Oakland couples married in the year after childbirth. After noting problems ranging from financial and relationship problems to the stress of living in a complex urban environment, Waller concludes that federal efforts to encourage marriage will face serious obstacles. Second, she suggests that cohabiting couples should receive special policy attention. The early evidence indicates that such couples are in relatively stable relationships and that cohabitation is associated with high expectations for marriage. This stability, she suggests, could translate into greater family cohesion and child wellbeing over time. Future waves of the national survey will examine other issues pertaining to fragile families, including paternal involvement and children’s wellbeing. As we at PPIC consider welfare reform in its broadest context, we see no issue more important than the circumstances iii and prospects of these children. We hope that this study, and the national effort of which it is a part, will improve not only public policy but also the lives of children today and for generations to come. David W. Lyon President and CEO Public Policy Institute of California iv Summary About 33 percent of all U.S. births occur outside marriage, up from about 4 percent in 1940 and 11 percent in 1970. In California, too, unmarried women now account for one-third of all births. Although increases in nonmarital births have leveled off in recent years, families headed by unmarried parents continue to present significant policy challenges. Children born outside marriage are more likely than other children to be poor and to receive welfare. They are also less likely to receive child support from their noncustodial parents. To address these challenges, welfare legislation and proposals to encourage responsible fatherhood have sought to promote marriage and to strengthen twoparent families. Despite this growing interest in marriage among policymakers, there is very little representative data on unmarried parents’ relationships. As a consequence, we have limited knowledge about why transitions to marriage do or do not occur or how to strengthen fragile families. Because unmarried parents’ decisions to marry or break up often occur in the first years after their children are born, it is especially important to understand relationship transitions during this time. This report investigates factors that support and discourage efforts on the part of unmarried parents to form stable relationships, including marriage. It focuses in particular on 250 families in Oakland who participated in the Fragile Families and Child Wellbeing (FFCW) survey, a national, longitudinal study of unmarried parents and their children. Oakland was the first city surveyed in this study and the first for which follow-up data are available. Drawing on two waves of surveys with parents, this report characterizes the stability of unmarried parents’ relationships in the first year of their child’s life and tracks their transitions toward increased and decreased involvement. It also examines open-ended interviews with parents to analyze economic, personal, and relationship issues that v parents faced in the first year of their child’s life. Finally, the report tests whether the issues identified in the interviews are related to parents’ assessments in the larger survey sample of whether they will marry their child’s other parent. Results One Year After Birth Of the Oakland parents considered in this report, about half were living together at the time of their child’s birth, 35 percent were in romantic relationships but were not living together, and 15 percent were not romantically involved with the other parent. One year later, about 7 percent of parents were married, almost half were living together, and 40 percent were not in a romantic relationship with the other parent. Cohabitation was the most stable relationship type in the first year. About three-quarters of cohabitors were still living together one year after their child’s birth, and an additional 10 percent had married. At the same time, these cohabiting parents were slightly more likely to decrease their involvement in relationships than to marry. Over half of parents who were romantically involved but not cohabiting ended their relationship between the time of birth and the 12-month follow-up; about 4 percent of these parents married and another 27 percent moved in together. Not surprisingly, the overwhelming majority of parents who were not romantically involved at the time of birth remained uninvolved one year later (see Figure S.1). Parents in Oakland reported high expectations about marriage at the time of their child’s birth. Over one-third of the mothers and almost half the fathers reported an almost certain chance that they would marry. Cohabiting parents had the highest expectations about marriage. Among mothers who were living with the father when their child was born, 53 percent reported an almost certain chance of marriage and only 8 percent indicated that there was little or no chance that they would marry the other parent. These expectations were related to the outcomes observed one year after the birth. Among parents who reported an almost certain chance of marrying the other parent, about 14 percent were married in the first year and an additional 68 percent were living together one year later. About 5 percent of parents who reported a good or 50-50 chance of marriage were married and slightly over half were cohabiting; about vi Percentage 100 13 90 2 80 70 55 60 50 75 97 40 15 30 20 27 10 10 0 43 Cohabiting Involved in Not involved at birth a romantic in a romantic relationship but relationship not cohabiting at birth at birth Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up NOTE: Reports from mothers in both waves of FFCW survey. Percentages may not sum to 100 because of rounding. Figure S.1—Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers one-third did not have romantic relationship 12 months after the birth. Of those parents who reported little or no chance of marriage, 83 percent did not have a romantic relationship with the other parent at the 12month follow-up. The results indicate a clear and statistically significant relationship between parents’ expectations and early outcomes. Insights from Qualitative Interviews The open-ended, qualitative interviews conducted for this report illustrate some of the economic, relationship, and personal challenges parents faced in the first year. The analysis identifies several factors related to the dissolution of parents’ relationships before the birth of their child or in the first 12 months. It also considers obstacles to relationship stability among couples who stayed together and how these obstacles contributed to their reluctance to marry. Parents who ended their relationships often mentioned a specific event (such as a violent incident, vii drug relapse, or deportation) that triggered the breakup. Other parents mentioned general relationship problems such as conflicts about trust, infidelity, and commitment. Parents were more likely to point to specific events or relationship problems than to economic factors when discussing what prompted their breakups. However, their accounts indicate that financial instability, including not having a place to stay together, occurred simultaneously with these problems and probably contributed to them. Parents who broke up before the follow-up interview were typically experiencing multiple problems when their relationships ended. Parents who stayed together in the first year mentioned many of the same problems as those who did not. Although these couples may have experienced less-severe or less-numerous problems than those who dissolved their relationship, they were primarily differentiated from other parents by the fact that they were living together at the time of their child’s birth. Many of the parents who stayed together in the first year regarded their relationships as highly vulnerable to dissolution, and most said that they wanted to wait until the conditions were right for marriage. They were often hopeful that their economic and personal circumstances would improve, making marriage possible. Despite these hopes, these couples are likely to experience continuing economic problems. Couples in the general population share some of the problems reported by these unmarried couples, which may explain why a large number of men and women are delaying marriage; however, the material hardship experienced by most parents at the Oakland site exacerbated common relationship problems and introduced new ones. Factors Related to Early Expectations The report also tracks the relationships between issues identified in the interviews—economic problems, distrust, conflict, physical violence, and drug and alcohol abuse—and how parents assessed their chances of marrying each other. These assessments are important insofar as they reflect parents’ expectations and predict relationship outcomes at year one. This part of the report is based on statistical analyses of responses from 3,712 mothers and 2,775 fathers to the FFCW survey. Because the models control for significant interactions between parents’ responses viii and their residences, the 20-city FFCW results can be applied to the Oakland sample. The analysis includes information from qualitative interviews to illustrate important findings. Cohabitation has the strongest positive effect on marriage expectations. This pattern indicates that unmarried parents view cohabitation as a step toward marriage. Also, the experience of cohabitation may increase parents' expectations that they will marry. Fathers’ employment is positively related to mothers’ reporting high chances of marriage. Also, mothers with higher education—an indicator of employment potential—express more certainty about marriage than other mothers. The results show strong negative relationships between marriage expectations and the belief that the other gender cannot be trusted to be faithful. Likewise, reports of high levels of conflict, violence, and partners with drug or alcohol problems are negatively related to marriage expectations. The effects of these factors are more consistent than those associated with economic factors. Conclusion and Policy Implications These findings provide an early look at relationships between unmarried parents in the first year of their child’s life. About half of unmarried parents in Oakland were living together when their children were born, suggesting a relatively high level of commitment in their relationships. One year later, parents were divided between those living in formal or informal unions and those who were no longer involved in romantic relationships. Despite high expectations at the time of their child’s birth, only 7 percent of unmarried parents had married. Parents who were living together were more likely to break up than to marry, and over half of the parents who were romantically involved but not living together broke up within 12 months of their child’s birth. These results suggest that policies and programs to encourage marriage face serious challenges. Programs that focus on maintaining relationships, stabilizing families, and facilitating the transition to marriage are likely to have the highest chances of success among cohabiting parents. However, even these parents—who, on average, have relatively stable relationships and ix high marriage expectations-—will have difficulty making the transition to marriage because of personal and economic problems. More feasible, perhaps, are programs that help fathers remain emotionally and economically involved in their children’s lives. Since 1996, states have been given more flexibility to initiate or support such programs. Eligibility criteria for cash assistance (such as Temporary Assistance for Needy Families, or TANF) and in-kind assistance (such as food stamps, housing, and health coverage) are primarily geared to single custodial parents. New proposals to broaden eligibility criteria for in-kind assistance and to eliminate distinctions between single-parent and twoparent families in receiving cash assistance may help these fragile families remain intact. Conflict and distrust between mothers and fathers may be less amenable to policy intervention than economic problems. Some community-based fatherhood programs have addressed these issues directly, but such programs tend to target noncustodial fathers with child support obligations, many of whom are no longer cohabiting or romantically involved with the other parent. The state might consider implementing more programs to serve unmarried couples still involved in romantic relationships, particularly couples who are cohabiting. Programs for parents with drug or alcohol problems may allow them to stabilize their families and increase their capacity for employment. Drug treatment could also help some noncustodial fathers emotionally reengage with their children. However, encouraging marriage or increased family involvement for parents with drug or alcohol problems may be unrealistic or detrimental to other family members. In families characterized by domestic violence, this danger is even more pronounced. Knowing more about these parents’ circumstances will help policymakers identify cases for which encouraging marriage would be inappropriate or unfeasible. This preliminary evidence suggests that the personal and economic barriers to marriage and relationship stability among unmarried parents are numerous, significant, and complex. As the FFCW study continues, policymakers will have an even clearer idea of how these complex issues affect unmarried parents’ capacity to form and maintain stable relationships, including marriage. x Contents Foreword ..................................... Summary..................................... Figures ...................................... Tables ....................................... Acknowledgments ............................... iii v xiii xv xvii 1. INTRODUCTION ........................... 2. STUDY DATA AND METHODS ................. Qualitative Study ............................. Characteristics of the Oakland Sample ............... 3. OUTCOMES AND EXPECTATIONS IN THE FIRST YEAR .................................... Changes in Unmarried Parents’ Relationships Between Birth and 12-Month Follow-Up .................... Expectations About Marriage ..................... Relationship Expectations versus Outcomes ............ Conclusion ................................. 4. INSIGHTS FROM QUALITATIVE INTERVIEWS ..... Parents Not in Romantic Relationships at Follow-Up ...... Relationship Ended in the First Year ............... Relationship Ended Before the Birth ............... Parents in Romantic Relationships at Follow-Up ......... Married at Year One ......................... Involved in Other Romantic Relationships at Year One ... Fragile Families in the First Year .................. Conclusion ................................. 5. FACTORS UNDERLYING EARLY EXPECTATIONS ... Men’s Employment and Education ................. Women’s Employment, Education, Public Assistance, and Household Income ........................ Gender Distrust.............................. 1 7 9 10 19 19 26 30 33 35 35 37 40 41 41 42 47 49 53 55 57 60 xi Relationship Conflict .......................... Violence in Relationships ........................ Drug and Alcohol Problems ...................... Race, Ethnicity, and Immigration .................. Age and Other Children ........................ Cohabitation ............................... Conclusion ................................. 6. CONCLUSION AND POLICY IMPLICATIONS ...... Financial and Housing Stability ................... Distrust and Conflict in Relationships ............... Drugs, Alcohol, and Violence ..................... 61 63 64 65 66 67 69 71 72 74 75 Appendix A. The Fragile Families and Child Wellbeing Study ......... 77 B. Sample Means and Regression Results ................ 85 Bibliography .................................. 97 About the Author ............................... 105 Other Related PPIC Publications ..................... 107 xii Figures S.1. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers.................. 2.1. Relationships Between Unmarried Parents at Child’s Birth, As Reported by Mothers ................. 3.1. Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers .......................... 3.2. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers.................. 3.3. Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers, by Race .................... 3.4. Expectations About Marriage at Child’s Birth, As Reported by Oakland Mothers and Fathers.......... 3.5. Expectations About Marriage, As Reported by Oakland Mothers, by Relationship at Child’s Birth........... 3.6. Relationships at 12-Month Follow-Up, As Reported by Oakland Mothers, by Expectations About Marriage .... 5.1. Odds Ratios for Mothers’ Reports of an Almost Certain Chance of Marriage ........................ vii 16 20 23 24 27 29 31 60 xiii Tables 2.1. Demographic Characteristics of Unmarried Parents .... 2.2. Socioeconomic Characteristics of Unmarried Parents ... 2.3. Material Hardship Caused by a Lack of Money Among Unmarried Parents in Oakland Between Child’s Birth and 12-Month Follow-Up .................... 3.1. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers.................. 3.2. Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers, by Race ............ 3.3. Means for Marriage Expectations, by Type of Relationship at Child’s Birth, Oakland ............ 3.4. Odds Ratios from Logistic Regressions of Mothers’ Reports of Being Married, Married or Cohabiting, or Not Romantically Involved at 12-Month Follow-Up ... 5.1. Factors Related to Parents’ Expectations About Marriage................................ A.1. Comparison of Demographic and Socioeconomic Characteristics of Unmarried Parents in Oakland in the Sample at 12-Month Follow-Up and Parents Not in Sample at 12-Month Follow-Up ................ B.1. Means for Dependent Variables in Logistic Regressions .. B.2. Means for Independent Variables in Logistic Regressions .............................. B.3. Odds Ratios from Logistic Regressions of Mothers’ Reports of an Almost Certain Chance of Marriage ..... B.4. Odds Ratios from Logistic Regressions of Mothers’ Reports of Little or No Chance of Marriage ......... B.5. Odds Ratios from Logistic Regressions of Fathers’ Reports of an Almost Certain Chance of Marriage ..... 11 14 15 21 25 30 32 54 83 85 86 89 91 93 xv B.6. Odds Ratios from Logistic Regressions of Fathers’ Reports of Little or No Chance of Marriage ......... 95 xvi Acknowledgments This report would not have been possible without the contributions of many people--including the mothers and fathers who kindly agreed to participate in the study. I would like to thank PPIC for its generous support of the Fragile Families and Child Wellbeing Study’s first site in Oakland, California. In particular, David Lyon’s sustained enthusiasm, direction, and backing of the project were instrumental in launching what is now an important, national study of unmarried parents and their children. Michael Teitz provided guidance throughout various stages of the project and Debbie Reed offered important advice on this report. My thanks go to David, Mike, and Debbie for their support of this multiyear effort. I would also like to express my appreciation to the National Institute of Child Health and Development, the Robert Wood Johnson Foundation, and the William and Flora Hewlett Foundation for their contributions to the Oakland site. Thanks to Frank Furstenberg, Laura Hill, Wendell Primus, and Elaine Sorensen for their valuable suggestions for revising the manuscript. The report also benefited from Laura Moye’s diligence, care, and intellectual input into the analysis and production of the report. Renzo Tragsiel provided excellent assistance for the qualitative study and Suzanne Shoemaker expertly transcribed the qualitative interviews. Peter Richardson’s significant editorial skills helped to sharpen the report’s presentation. Nancy Reichman and Ofrira Schwartz at Princeton University made the use of the survey data possible and helped coordinate the research process. Finally, I would like to thank Sara McLanahan, Irv Garfinkel, and Ron Mincy for envisioning this innovative study of fragile families. As principal investigators of the study, Sara and Irv have also made this vision a reality. Despite these contributions, all errors in this report are the responsibility of the author alone. xvii 1. Introduction About 33 percent of all U.S. births occur outside marriage, up from about 4 percent in 1940 and 11 percent in 1970 (Ventura and Bachrach, 2000).1 In California, too, one-third of births occur to unmarried women (Curtin and Martin, 2000). Although increases in nonmarital births have leveled off in recent years, the prevalence of families headed by unmarried parents presents significant policy challenges.2 Children born outside marriage are more likely than other children to be poor and to receive welfare, and they are less likely to receive child support from their noncustodial parents. In light of these challenges, the language of welfare reform legislation and recent proposals to encourage “responsible fatherhood” emphasize marriage and strengthening two-parent families. According to the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, two of the four purposes for which Temporary Assistance for Needy Families (TANF) funds can be used include ____________ 1Daniel Patrick Moynihan’s controversial report, The Negro Family: The Case for National Action (1965), opened the contemporary policy debate about the role of public policy in promoting two parent families. Although its focus was on black families, the report drew public attention to the changing structure of families and called for a national response to this issue. At the time that report was released, however, less than 8 percent of all births in the United States occurred to unmarried women. The Moynihan report also focused on the fact that about one-quarter of black children were born to unmarried parents. This is about same proportion (22 percent) of children who are now born to unmarried white parents. Births to unmarried white women make up the largest percentage of births outside marriage in the United States (40 percent), followed by births to black women (32 percent) and Hispanic women (25 percent). However, the proportion of births outside marriage is higher among black and Hispanic women— about 69 percent and 42 percent, respectively (Ventura and Bachrach, 2000). 2Social scientists have reported that marriage has declined and that the proportion of nonmarital births has risen across social groups in the United States. Not only have recent cohorts of adults postponed marriage, but men and women are also more likely to divorce and less likely to remarry. Therefore, men and women have a higher risk of having a child outside marriage. Furthermore, fewer men and women today are deciding to marry in response to a nonmarital pregnancy (Ventura et al., 1995). 1 promoting marriage and encouraging the formation and maintenance of two-parent families.3 The President’s budget proposes $64 million in 2002 ($315 million over five years) for programs that strengthen the role of fathers in the lives of families. Competitive grants would be provided to faith-based and community organizations that promote marriage (U.S. Office of Budget and Management, 2001).4 Congress is also considering similar legislation primarily targeted at low-income parents.5 The Strengthening Working Families Act of 2001, for example, would provide funds “to promote the formation and maintenance of married, twoparent families, strengthen fragile families, and promote responsible fatherhood” (U.S. 107th Congress, 2001b). ____________ 3The 1996 welfare reform legislation that created TANF is part of PRWORA. PRWORA begins with the findings that “(1) marriage is the foundation of a successful society” and “(2) marriage is an essential institution of a successful society which promotes the interests of children” (U.S. 104th Congress, P. L. 104-193, Sec. 101). According to PRWORA, TANF funding can be used “to end the dependence of needy parents on government benefits by promoting job preparation, work, and marriage” and “to encourage the formation and maintenance of two-parent families” (U.S. Department of Health and Human Services, 1999, pp. 11–12). 4President Bush emphasized marriage during his campaign and since taking office (Government Executive Magazine Daily Briefing, 2001; Milbank, 2001). The nomination of Wade Horn as Assistant Secretary of Health and Human Services for Family Support also signals the importance of marriage in the current administration. Horn, the former president of the National Fatherhood Initiative, was quoted as saying that he viewed the administration as having a four-pronged strategy of “strengthening fatherhood, strengthening marriage, strengthening community organizations that help families, and seeking a role for religious organizations in building communities” (Milbank, 2001). 5The Strengthening Working Families Act of 2001 (S. 685), the Responsible Fatherhood Act of 2001 (H.R. 1300), and the Child Support Distribution Act of 2001 (H.R. 1471) (U.S. 107th Congress, 2001a,b,c) all contain provisions to encourage marriage and the formation and maintenance of two-parent families through fatherhood programs and media campaigns (Center on Fathers, Families, and Public Policy, 2001). The Child Support Distribution Act, for example, would make grants available to fatherhood programs “that promote marriage through such activities as (A) counseling, mentoring, disseminating information about the advantages of marriage, enhancing relationship skills, teaching how to control aggressive behavior, disseminating information on the causes and treatment of domestic violence and child abuse, and other methods; and (B) sustaining marriages through marriage preparation programs, premarital counseling, and marital inventories, and through divorce education and reduction programs, including mediation and counseling.” 2 For low-income unmarried parents, the barriers to forming stable relationships, including marriage, appear to be formidable and multifaceted (Furstenberg, forthcoming). Demographic data indicate that only about two out of five mothers who had a nonmarital birth married their child’s father or anyone else within five years of the birth (Ventura et al., 1995, pp. 58–59). Furthermore, relationships between unmarried parents, and between unmarried fathers and their children, decline significantly in these early years (Sorensen, Mincy, and Halpern, 2000; Lerman and Ooms, 1993). Although there is substantial and growing interest in relationship transitions between unmarried parents, researchers and policymakers have not had access to representative, longitudinal data on unmarried parents that examine where these changes begin and why they occur. This report draws on a unique survey of unmarried parents and their children to provide an early look at changes in relationships between unmarried parents. The report also looks at factors that support and discourage their efforts to form stable relationships, including marriage. Specifically, the report examines data on 250 families headed by unmarried parents living in Oakland, California, who participated in the Fragile Families and Child Wellbeing (FFCW) survey, a national, longitudinal study of unmarried parents and their children. Although Oakland was the first city to participate in the study, the total sample of unmarried parents includes 3,600 families living in 20 U.S. cities. Because most parents in the FFCW study have low incomes, examining a sample of unmarried parents in Oakland primarily means looking at economically disadvantaged families. The study refers to these unmarried parents and their children as fragile families because they have a higher risk of poverty and family dissolution than do families headed by married parents. Unmarried mothers and fathers participated in a baseline interview at the time of their child’s birth and a follow-up interview 12 months later. This report draws on both waves of interviews to provide new information on changes in the relationships between unmarried mothers and fathers in the first year of their child’s life and factors that underlie 3 these transitions. This report also uses information from qualitative interviews with Oakland families who participated in the study to enrich the analysis. Because the study focuses on men and women who have already become unmarried parents, it addresses only a piece of this complicated puzzle involving the changing nature of American families. However, the report uses newly available data to shed light on factors that may discourage union formation and stability. Most of the increase in nonmarital childbearing since the early 1980s has occurred to cohabiting parents, and a significant proportion of children born outside marriage are now living with both biological parents early in life (Bumpass and Lu, 2000).6 In recognition of these changes, the report pays particular attention to the effect of cohabitation on parents’ relationships and outlooks for the future. The report begins by examining early relationship outcomes (including marriage, cohabitation, and dissolution) in the Oakland sample. Chapter 2 describes the sample and research design, and Chapter 3 discusses relationship transitions that begin soon after unmarried couples have a child. In particular, this chapter looks at changes in relationships between parents who live together, parents who are involved in romantic relationships but do not live together, and parents who reported that they did not have a romantic relationship with the other parent at the time of their child’s birth. The report also examines whether marriage is something unmarried parents think will happen and how expectations match early relationship transitions. Chapter 4 draws on evidence from qualitative interviews conducted between the baseline survey and the 12-month follow-up survey to investigate the challenges to relationship stability that unmarried parents face in the first year of their child’s life. This analysis examines factors related to the dissolution of parents’ relationships before the birth of their child or in the first 12 months. It also looks at challenges experienced by couples who stayed together in the first year and explores how they may be related to couples’ reluctance to marry in the first year. ____________ 6Most of the increase in nonmarital births to cohabiting women occurred among white women (Ventura and Bachrach, 2000). 4 Chapter 5 tests whether the challenges identified in the qualitative interviews are related to expectations held by parents in the larger sample. Expectations reflect parents’ own sense of the potential for marriage, and the factors that influence expectations are likely to be the same as those that influence outcomes. Because the size of the follow-up sample in Oakland is not large enough to permit multivariate regression analysis, parents’ assessments of whether they will marry their child’s other parent are used as a proxy for relationship outcomes, such as marriage, cohabitation, and relationship dissolution. Examining parents’ expectations has the added benefit of allowing us to think about how relationships might continue to change beyond the first year. The analysis first considers how different dimensions of unmarried parents’ economic status and security may affect their relationships. A significant body of research shows a positive connection between male employment and marriage. Male employment is of particular concern in low-income communities, especially poor, black communities where, scholars argue, high levels of joblessness render men less “marriageable” in the eyes of women. The analysis next examines how general perceptions of trust in relationships influence unmarried mothers’ and fathers’ orientations toward marriage. Scholars suggest that trust issues may be especially important for couples living in economically unstable environments. The analysis then investigates how conflict affects unmarried parents’ relationships. Relationships with high levels of conflict are vulnerable to dissolution and may be less likely to result in marriage. Also, there is reason to believe that new, unmarried parents with limited economic resources may experience particularly high levels of stress and conflict in their relationships. Finally, the analysis considers the effects of domestic violence and drug or alcohol abuse—indicators of serious and potentially dangerous problems. Although many of these issues contribute to delayed marriage in the general population, they may be more severe or numerous among unmarried parents and present greater barriers to relationship stability.7 ____________ 7Social scientists have identified multiple changes in the economy, culture, and gender expectations that have discouraged marriage and encouraged increases in nonmarital childbearing. Many of these changes have occurred across social groups in the United States and in other industrialized countries. In recent years, men’s earnings have 5 This information and analysis can help policymakers, community leaders, and practitioners identify problems that unmarried parents face at the time of their child’s birth and how these problems affect their relationships during the first year. This information also provides a window into the kind of difficulties families may continue to experience in the coming years. ________________________________________________________ declined relative to women’s, changing the conventional “gender bargain” in marriage (McLanahan and Sandefur, 1994). Theories about culture and gender suggest that these factors had equally important roles in the shift toward delayed marriage (or nonmarriage) as economic factors. For example, young men and women now perceive fewer advantages to marriage than in the past (Thornton and Freedman, 1982) and seem to have higher expectations for realizing cultural ideals, such as compatibility in their relationships. Scholars suggest that changing expectations about gender have also meant that women are now less willing to tolerate extremely unequal or abusive marriages (Stacey, 1991; Luker, 1996). 6 2. Study Data and Methods This study of early relationship transitions focuses on unmarried parents living in Oakland, California—one of 20 cities across the United States in which parents are participating in the FFCW study.1 The FFCW study follows a new birth cohort of approximately 4,700 children to learn more about the circumstances and experiences of unmarried parents and their children in the first four years of their children’s lives.2 In particular, the study investigates such things as the relationships between unmarried parents, the involvement of unmarried fathers in their children’s lives, and the wellbeing of children born outside marriage (see Appendix A). The Oakland sample includes 250 families headed by unmarried parents and a comparison group of 75 families headed by married parents. These families were approached at the time of their children’s birth and will be followed for four years. New mothers were initially interviewed in person at the hospital within 24 hours of giving birth. Fathers were interviewed either at the hospital or elsewhere as soon as possible after the birth. The first interview with the mother lasted about 30 minutes and the interview with the father lasted 40 minutes. This report also includes information from the 12-month follow-up interview in Oakland. This interview was conducted over the telephone and lasted about 50 minutes.3 ____________ 1The other 19 cities, which were selected to represent different policy environments and labor markets, are Austin, Baltimore, Boston, Chicago, Corpus Christi, Detroit, Indianapolis, Jacksonville, Milwaukee, Nashville, New York, Newark, Norfolk, Philadelphia, Pittsburgh, Richmond, San Antonio, San Jose, and Toledo. 2The total sample includes 4,700 births (3,600 nonmarital and 1,100 marital). The data are representative of nonmarital births to parents residing in cities with populations over 200,000. The data are also representative of nonmarital births in each of the cities that make up our sample. See Reichman et al. (2001) for more information about the study’s methodology. 3Two more follow-up interviews will be conducted by phone, and the final follow-up interview will include an in-home assessment of the child’s school readiness at age four. 7 The data presented in this report are representative of nonmarital births to women in Oakland hospitals between February and June 1998.4 To generate a representative sample of births, all mothers who delivered during this time were approached in the hospitals and asked to participate in the study until both the nonmarital and marital quotas were reached. Approximately 90 percent of unmarried mothers agreed to participate in the study. Mothers were also asked to provide locating information on the fathers, and fathers were contacted either in the hospitals or as soon as possible after the child’s birth. About 75 percent of unmarried fathers participated in the study. The total baseline survey in Oakland includes interviews with 248 mothers and 189 fathers. Because most mothers we approached in the hospitals agreed to participate in the study, and because mothers in both the city’s birthing hospitals were interviewed, we are confident that the sample is representative of the population of unmarried women giving birth in Oakland hospitals.5 We are less certain that the sample of unmarried fathers is representative, as only 75 percent of these men were interviewed. Compared to average unmarried fathers, the men who were interviewed in this study are more strongly attached to the mothers of their children and differ in other ways as well.6 In the follow-up survey, 85 percent of unmarried mothers who were interviewed at baseline were also interviewed at 12 months. Seventy-one percent of unmarried fathers who were interviewed at baseline were interviewed again one year later. Therefore, the follow-up survey includes interviews with 212 unmarried mothers and 144 unmarried ____________ 4The research design assumes that families of children born in this four-month period are the same as families who have children at other times of the year. 5The sample includes unmarried mothers who gave birth in Oakland hospitals in early 1998, not the entire population of Oakland mothers who gave birth in this year. The sample does not include mothers living in Oakland who gave birth in hospitals outside the city. 6See Teitler, Reichman, and Sprachman (2001) for a description of how fathers who were interviewed differ from those who were not interviewed. 8 fathers.7 Appendix A describes how parents who participated in both surveys compare to those who dropped out of the survey at 12 months. Qualitative Study This report also includes information from qualitative interviews I conducted with a subsample of Oakland parents who participated in the survey. These interviews allowed me to explore survey themes in greater depth with the FFCW participants and to interpret their responses. All qualitative interviews were conducted between the two survey interviews (i.e., after the birth but before the child was one year old). The total qualitative sample includes 60 parents (37 unmarried parents and 23 married parents). The sample of unmarried parents had a response rate of about 86 percent for mothers and 88 percent for fathers.8 To select parents for qualitative interviews, I divided the survey sample into three groups: black mothers, Hispanic mothers of Mexican descent, and white mothers.9 I then drew a random sample of births to unmarried and married parents within race and ethnic groups.10 Although the percentages of black and Hispanic births selected for the qualitative interviews roughly reflect the proportions of these births in the larger survey, I oversampled white parents to allow for group ____________ 7Nine fathers who had not been contacted at baseline entered the sample at the 12month follow-up survey. These fathers were not included in the analysis. 8The qualitative sample is drawn from the survey sample. Of the 70 married and unmarried parents contacted for qualitative interviews, ten parents could not be located or refused to participate. I did not approach three families in cases where (1) the baby died shortly after birth, (2) the baby was put up for adoption, and (3) the parents told survey researchers that they did not want to participate in a follow-up interview. Furthermore, I did not attempt to locate fathers who did not participate in the baseline survey. 9Because only a small number of unmarried, Asian parents participated in the Oakland survey and because of the diversity of Asians within this group, I did not draw a qualitative sample of these parents. Unfortunately, Asians who did not speak English were not interviewed in the FFCW study because the expense of hiring multilingual interviewers for Asian parents was prohibitive. 10When mothers’ and fathers’ racial or ethnic characteristics differed, I selected cases on the basis of mothers’ characteristics. Five of these couples are interracial or interethnic. 9 comparisons.11 A final qualification is that the subsample selected only Mexican-American mothers who were born in the United States or had lived in this country since age 12.12 Because most of the parents of Mexican origin in my sample were born in the United States and most others had attended school in this country and had been exposed to its culture, I believed that this strategy would allow for better comparisons among parents. This analysis focuses on the 37 unmarried parents who participated in the qualitative interviews. Twenty-eight of the parents interviewed were part of a coupled sample of 14 families in which both the mother and father participated. For the remaining nine interviews, I was unable to obtain an interview with the other parent. In all, 23 families are represented in the qualitative sample of unmarried parents.13 The qualitative interviews were conducted with mothers and fathers separately (as they were in the survey) and lasted about one and one-half hours. Because questions asked in the interview were semi-structured and open-ended, the interview encouraged parents to elaborate on some of the questions and to introduce new issues for discussion. I conducted the majority of interviews in the parents’ homes. Finally, all interviews were recorded and transcribed for data analysis. Characteristics of the Oakland Sample The remainder of this chapter presents a profile of Oakland unmarried parents’ demographic, economic, and relationship ____________ 11The total qualitative sample is made up of 42 percent black mothers and 44 percent black fathers, 33 percent Hispanic mothers and 41 percent Hispanic fathers, and 24 percent white mothers and 15 percent white fathers. 12Two of these five mothers arrived at age one, two arrived at age nine, and one arrived at age 12 (she had been in the United States for eight years). Four of their partners, however, had not been in the United States since age 12. Three of them had moved to the United States within two years of the time they were interviewed; the fourth father had lived in the United States for eight years, but he moved to this country at age 26. Renzo Tragsiel (a FFCW survey interviewer) conducted these four interviews with fathers in Spanish. 13The sample of unmarried parents consists of 52 percent black mothers and 63 percent black fathers, 29 percent Hispanic mothers and 31 percent Hispanic fathers, and 19 percent white mothers and 6 percent white fathers. 10 characteristics.14 It also reports information from the combined survey sites to compare unmarried parents in Oakland with parents in the survey as a whole. Appendix A describes in more detail how the parents who were reinterviewed one year after their child’s birth compare to those parents who did not participate in the follow-up survey. Table 2.1 presents information about the demographic characteristics of unmarried parents in the Oakland sample and the total 20-city sample of the FFCW survey. Although this report focuses on parents in Oakland, it is useful to look at the two samples side by side to compare the characteristics of Oakland parents with a nationally representative sample of unmarried parents living in other large cities. Table 2.1 Demographic Characteristics of Unmarried Parents (in percent) Agea Under 20 21–25 26–30 Over 30 Race/ethnicitya White (non-Hispanic) Black (non-Hispanic) Hispanic Asian Native American Other/not classified Immigrant Have other children Total number of respondents Oakland Sample 20-City Sample Mothers Fathers Mothers Fathers 26 12 33 18 34 32 35 37 24 29 18 21 16 28 14 23 3 1 14 12 58 60 55 55 33 34 28 29 4 3 11 1 1 <1 1 1 2 12 31 33 15 16 69 57 61 56 248 189 3712 2775 NOTE: The table includes some parents who were unmarried at the time of their child’s birth but were married at year one. aPercentages may not sum to 100 because of rounding. ____________ 14See McLanahan, Garfinkel, and Waller (1999) for more information about sample characteristics at baseline. 11 As the table indicates, the largest group of unmarried mothers (34 percent) and fathers (32 percent) in the Oakland sample were between the ages of 21 and 25 at the time of their child’s birth. Slightly over onequarter of unmarried mothers were age 20 or under and slightly under a quarter were between the ages of 26 and 30. Fewer fathers than mothers were age 20 or under and more were over age 30. Unmarried parents in the Oakland sample were older than those in the total sample. The majority of parents in the Oakland and other sites had at least one biological child in addition to the child who was just born.15 The majority of parents in both samples are black or Hispanic. About six out of ten unmarried parents in Oakland are black and about one-third are Hispanic. A small percentage of unmarried parents in Oakland are white, Asian, or Native American.16 Less than 3 percent of parents identified their race as “other” or reported racial or ethnic ____________ 15On average, Oakland mothers have 2.5 children and fathers have 2.1 children (these means include the child just born). Information on the parent of previous children was not available at baseline. About 50 percent of mothers in both waves of the survey had additional children with their new baby’s father only or with the baby’s father and another partner. The other 50 percent had additional children with another partner only. For fathers in Oakland, 49 percent had additional children with the mother only or with the mother and another partner. About 51 percent had additional children with another partner only. 16Our racial and ethnic percentages are somewhat different from those reported by California Vital Statistics for the same hospitals, which show a higher proportion of births to blacks and a lower proportion of births to Hispanics and whites. The figures for unmarried births from California Vital Statistics indicate that 62 percent of children delivered in these hospitals in early 1998 were born to black mothers, 22 percent to Hispanic mothers, 10 percent to white mothers, and 7 percent to mothers of other races. The FFCW study excluded births from the sample if the parents planned to put their child up for adoption, if the father died before the birth of the child, and if the mother did not speak English or Spanish. Differential response rates by race or ethnicity were not recorded in Oakland. The differences between our sample and that of California Vital Statistics may result, in part, because our sample was likely to miss some mothers whose children were born with serious health problems and who were moved to other hospitals for care shortly after birth. We would also miss mothers whose children died immediately after birth. Although we do not know the racial composition of children with serious health problems in the two hospitals we sampled, we do know that unmarried, black mothers are more likely to have low-birth-weight children than are Hispanic and white mothers. Such children have much greater health risks, including infant mortality. See Spetz (2001, p. 82) for information about how birth weight varies by race and ethnicity in California. See Padilla and Reichman (2001) for an analysis of factors related to low birth weight in the FFCW sample. 12 information that could not be classified within one of the larger categories. The Oakland site includes a smaller percentage of whites and a somewhat larger percentage of Asians than the total sample. Compared to the larger sample, Oakland has more black and Hispanic parents. Most of the Hispanic parents in Oakland are of Mexican descent (72 percent of mothers and 84 percent of fathers). Furthermore, over 30 percent of Oakland parents are immigrants to the United States—about double the percentage in the total sample.17 Table 2.2 shows the socioeconomic characteristics of unmarried parents in the Oakland survey and total sample. The median household income of unmarried mothers and fathers in the Oakland site was close to $17,500 at the time of their child’s birth, similar to the incomes of mothers in the combined sample.18 Fathers in the larger sample, however, had higher household incomes than Oakland fathers. After household size is taken into account, over half of Oakland mothers and close to half of fathers had incomes at or below the poverty line. Again, more fathers in Oakland than fathers in the combined sample could be classified as poor. More than eight out of ten mothers and seven out of ten fathers in Oakland have incomes at or below 200 percent of the poverty line. Over 10 percent more mothers and fathers in Oakland than parents in the total sample fall into the poor or near-poor category. The results also show that the educational attainment of unmarried parents in Oakland and the total sample is generally low but that Oakland parents report lower levels of education than do parents in the whole survey. For example, about half of unmarried parents in Oakland did not have a high school diploma compared to about 40 percent in the total survey. Despite a strong economy at the time of the interview and an unemployment rate in Oakland of about 3.9 percent, only two-thirds ____________ 17The majority of unmarried, Hispanic immigrants (77 percent of mothers and 85 percent of fathers) in the Oakland study are also from Mexico. 18For cases in which household income was missing, it was imputed from the following variables: parents’ age, education, race or ethnicity, immigrant status, whether or not they worked in the previous year, number of hours worked in the previous year, earnings in the previous year, whether or not they received welfare in the previous year, and number of adults in the houshold. Alternative ways of imputing household income are currently being evaluated. 13 Table 2.2 Socioeconomic Characteristics of Unmarried Parents Education (%)a Less than high school High school only Some college College Oakland Sample 20-City Sample Mothers Fathers Mothers Fathers 52 47 41 40 31 33 34 36 15 19 23 21 2 23 3 Median household income at child’s birth ($) 17,300 17,500 17,500 22,500 Mothers who had earnings year before 35 — 67 — child’s birth (%) Mothers employed at 12-month follow-up (%) 39 — — — Poverty status (%) < 50 of line 50–100 of line 100–200 of line 20 18 23 17 31 28 20 16 31 26 28 26 Fathers employed week before child’s birth (%) — 67 — 76 Fathers employed at 12-month followup (%) — 71 — — Total number of respondents at baseline 248 189 3,712 2,775 aPercentages may not sum to 100 because of rounding. NOTE: The table includes some parents who were unmarried at the time of their child’s birth but were married at year one. of unmarried fathers in Oakland were employed in the week before their child’s birth compared to over three-quarters of unmarried fathers in the combined sites. The employment rate for unmarried fathers in Oakland who were reinterviewed was about 4 percent higher at the 12-month follow-up interview, when the general unemployment rate for Oakland had dropped to about 3.6 percent. Surprisingly, only about 35 percent of mothers in Oakland had earnings in the year prior to their child’s birth compared to 67 percent in other locations. Oakland mothers 14 continued to have low employment rates one year later (only about 39 percent were employed the week before the follow-up interview).19 Table 2.3 presents information about mothers’ and fathers’ reports of material hardship experienced between the time of their child’s birth and the 12-month follow-up interview. The most common types of hardship include having to borrow money from family and friends to pay bills, not paying the full amount of a gas or electricity bill, receiving free food or meals, not paying full amount of mortgage or rent, moving in with other people, and having phone service disconnected. Less than 10 percent of parents reported having their gas or electric service turned off, staying at a shelter, not seeing a doctor, or being evicted because there was not enough money. Although a small percentage of parents say their Table 2.3 Material Hardship Caused by a Lack of Money Among Unmarried Parents in Oakland Between Child’s Birth and 12-Month Follow-Up (in percent) Received free food or meals Children went hungry Parent went hungry Did not pay full amount of mortgage or rent Evicted from home for not paying rent/mortgage Did not pay full amount of a gas or electricity bill Service turned off by the gas or electricity company Service disconnected by the phone company Borrowed money from family/friends to help pay bills Moved in with other people Stayed at a shelter/some other place for at least a night Someone in the house did not see a doctor Mothers 17 5 11 16 2 25 9 13 32 13 6 6 Fathers 13 1 5 17 2 17 3 14 27 13 6 8 NOTE: The table includes parents who were unmarried at the time of their child’s birth but were married at year one. ____________ 19Statistical analyses show that Oakland mothers with lower levels of education were significantly less likely to have earnings the year before their baby was born or to be working the week before the 12-month follow-up interview. Other characteristics of mothers such as race/ethnicity, immigration status, and having another biological child were not associated with mothers’ employment status at either interview. Age was significant at the .10 level for employment at follow-up, indicating that older mothers were less likely than younger mothers to be employed. 15 children went hungry because of a lack of money, a larger percentage of parents reported going hungry. The majority of parents (about 58 percent) reported experiencing at least one type of material hardship in the past year. Of those experiencing hardship, most had experienced two or more types. For example, although the indicators of housing-related hardships individually range between 2 percent and 17 percent, about 26 percent of unmarried mothers and 29 percent of unmarried fathers reported some type of hardship related to housing.20 Figure 2.1 shows unmarried parents’ relationship characteristics at the time of their child’s birth in the Oakland site and combined sites of the FFCW survey. The most striking result is that the largest group of unmarried parents in Oakland were living with their baby’s other parent 60 50 49 48 40 30 38 34 Oakland 20-city Percentage 20 17 14 10 0 Cohabiting Involved in romantic Not romantically relationship not involved cohabiting NOTE: Percentages may not sum to 100 because of rounding. Figure 2.1—Relationships Between Unmarried Parents at Child’s Birth, As Reported by Mothers ____________ 20In addition, about 45 percent of unmarried parents said they had moved since the birth of their child. These parents moved 1.6 times on average between the birth and first year follow-up. For Oakland parents who moved in the first year, the number of moves ranged from one to seven times. 16 at the time of the birth. Almost half of Oakland parents were cohabiting, similar to the total sample. The next-largest group—parents who were in romantic relationships at the time of the birth but were not living together—consisted of about 38 percent of unmarried parents in Oakland and 34 percent in the total sample. A minority of parents said that they either “hardly ever” talked or did not talk to their child’s other parent; others described themselves as “just friends.” These parents were categorized as not being romantically involved at the time of birth. A somewhat smaller group fell into this category in Oakland (about 14 percent) than in the survey as a whole (about 17 percent). Therefore, we see that 87 percent of parents in Oakland were involved in romantic relationships at the time of their child’s birth compared to 82 percent in the total sample. The majority of these parents were living together. In sum, unmarried parents in Oakland with new babies had limited socioeconomic resources. About half reported incomes at or below the poverty line for their family size and about half did not have a high school diploma. Most were in their twenties and had another biological child. The majority were also black or Hispanic; over 30 percent were immigrants. Oakland parents tended to be in committed romantic relationships at the time of their child’s birth, with about half living together in informal unions. 17 3. Outcomes and Expectations in the First Year As we saw in the last chapter, unmarried parents’ involvement in relationships was high at the time of their child’s birth. This chapter examines how relationships between unmarried parents at birth compare to their relationships one year later. It then investigates how unmarried mothers and fathers assess their chances of marrying their child’s other parent and how these expectations about marriage correspond to relationship outcomes. Given the fragility of relationships between unmarried parents, it is important to understand where these changes emerge and how they are related to parents’ hopes for the future. Changes in Unmarried Parents’ Relationships Between Birth and 12-Month Follow-Up In the last chapter, we saw that most unmarried parents were involved in romantic relationships with their child’s other parent at the time of their child’s birth. About half were living together, about 35 percent were involved in romantic relationships with each other but were not living together, and only 15 percent were not romantically involved at the time of birth (Figure 3.1).1 Figure 3.1 also shows some important changes in the overall composition of Oakland parents’ relationships in the first 12 months. One year after the birth, parents are divided to a greater extent between those living in cohabiting or marital unions and those who do not have a romantic relationship with the other parent. About 7 percent of parents who were unmarried at the time of birth were married one year later (Figure 3.1). A similar percentage of parents ____________ 1The percentages differ slightly from those presented in Chapter 2 because they are shown only for families who participated in the baseline and 12-month follow-up surveys. 19 15% 50% 40% 7% 47% 35% 7% Married Cohabiting Involved in a romantic relationship but not cohabiting Not involved in a romantic relationship NOTE: Percentages may not sum to 100 because of rounding. Figure 3.1—Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers (about half) were living together at the baseline and at the 12-month interview. Cohabitation continues to be the most common relationship type at follow-up. A much larger percentage of parents reported not being in a romantic relationship with the other parent at the 12-month follow-up than at the child’s birth. Fully 40 percent of parents reported no romantic relationship with the other parent one year after their child’s birth.2 Another large change occurred in the category of unmarried parents who were not living together but had romantic relationships at birth. This group accounted for about 35 percent of parents at birth but only 7 percent of parents one year later. An important caveat is that the follow-up survey gave parents the option of reporting whether they were living together “some of the time” or “all of the time,” whereas the baseline survey simply asked whether ____________ 2About one-quarter of parents who reported no romantic relationship with the baby’s other parent at the follow-up said they had romantic relationships with other partners at that time (results not shown). 20 parents lived together.3 Therefore, we may not be comparing exactly the same type of cohabiting relationships at birth and at 12 months. Because the cohabitation variables differ, they are collapsed into one category (i.e., the baseline variable). We next look at how relationships between parents who were cohabiting, romantically involved but not cohabiting, and not romantically involved changed between birth and follow-up (Table 3.1). This analysis indicates which relationships changed the most in the first year and where these changes occurred. Among those who were living together when their child was born, about 85 percent maintained or increased their involvement with the other parent. Specifically, threequarters were still living together and an additional 10 percent were married one year later. However, about 15 percent of parents had decreased their involvement in relationships. Most of these changes occurred through Table 3.1 Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers (in percent) Married Relationship at 12-month follow-up Involved in a romantic Not relationship, romantically Cohabiting not cohabiting involved Relationship at birth Cohabiting 10 75 2 13 Involved in a romantic relationship, not cohabiting 4 27 15 55 Not romantically involved 0 0 3 97 NOTE: Percentages in rows may not sum to 100 because of rounding. ____________ 3At follow-up, about 11 percent of cohabitors were living together some of the time and 89 percent were living together all of the time. 21 relationship dissolution. About 13 percent of cohabitors reported no relationship at the 12-month follow-up interview. In comparison, about 4 percent of parents who were married at the time of birth were separated or divorced 12 months later (results not shown).4 Only 2 percent of former cohabitors were still involved in romantic relationships other than marriage or cohabitation. Since parents may be living apart from each other for a variety of reasons at 12 months, those parents did not necessarily decrease their commitment to the relationship. Unlike parents who were living together at the birth, those who were romantically involved but not living together experienced considerable changes in their relationships. In fact, only 15 percent of these parents reported being in the same type of relationship at both interviews. The greatest movement among non-cohabitors was toward relationship dissolution in the first year. Over half of these parents ended their relationship between the time of birth and the 12-month follow-up. Close to one-third of these parents appear to have increased their involvement in the relationship, with 4 percent marrying and another 27 percent moving in together. The overwhelming majority of parents who were not romantically involved at the time of birth remained uninvolved one year later.5 About 3 percent (i.e., one couple) seem to have re-initiated their relationships one year after their child’s birth, but these parents did not marry or move in together. Figure 3.2 depicts movement and non-movement in each type of relationship reported in Table 3.1. We notice that cohabiting relationships (represented in the first column) are generally stable in the first year. However, cohabitors were slightly more likely to decrease their involvement in relationships than to marry during this time. Although this group is more likely to marry than other groups of parents, marriage ____________ 4An additional parent who reported being married at the time of the birth characterized her relationship status as “cohabiting” at the follow-up interview. 5This group includes parents who said that they were just friends and those who said that they hardly or ever talked at the time of birth. The wording of the question at 12 months changed somewhat. Therefore, parents in the “no romantic relationship” category include those who said they were just friends and those who said they were not in any kind of a relationship with the father. 22 Percentage 100 13 90 2 80 70 55 60 50 75 97 40 15 30 20 27 10 10 0 43 Cohabiting Involved in Not involved at birth a romantic in a romantic relationship but relationship not cohabiting at birth at birth Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up NOTE: Percentages may not sum to 100 because of rounding. Figure 3.2—Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers rates among cohabitors are low. It will be important to watch the rate at which these relationships lead to marriage or dissolution in the coming years. Among parents who were not cohabiting at birth, the deterioration in many relationships seems to occur rapidly. Although some of these couples married and others may do so in the future, their relationships appear to be extremely fragile. For parents not in romantic relationships at birth, the likelihood of marriage, or even resuming a romantic relationship with the other parent, appears to be negligible. When we look at relationships at birth and at 12 months by race and ethnic subgroups, we see that a lower percentage of black parents (44 percent) were cohabiting at the time of the birth than Hispanic parents 23 Percentage 100 90 16 80 70 41 60 50 40 30 44 20 9 29 62 25 49 1 9 70 37 Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up 10 0 54 Black Hispanic Black Hispanic Relationship at birth Relationship at 12 months NOTE: Because only five unmarried white mothers were in Oakland sample at birth and 12 months, the percentages are not displayed. Percentages may not sum to 100 because of rounding. Figure 3.3—Relationships Between Unmarried Parents at Child’s Birth and at 12-Month Follow-Up, As Reported by Oakland Mothers, by Race (62 percent) (Figure 3.3).6 Conversely, more black parents (41 percent) than Hispanic parents (29 percent) were involved in romantic, noncohabiting relationships at this time. A small portion in each group were not involved in romantic relationships with the other parent. At the follow-up interview, a similar percentage of black and Hispanic parents were married. However, a larger proportion of Hispanic parents were living together at the follow-up interview than at birth, and a smaller proportion of black parents were living together. Almost half of black parents were not involved in romantic relationships one year after their child’s birth compared to one-quarter of Hispanic parents. When we look at relationship transitions, we see that a similar percentage of black parents (6 percent) and Hispanic parents (7 percent) ____________ 6The percentages are based on reports of mothers’ race. However, 12 percent of Oakland mothers who participated in both waves of the survey were in interracial or interethnic relationships. 24 who were cohabiting at baseline married in the first year after their child’s birth (Table 3.2). Although cohabiting relationships are stable for both groups, they appear to be particularly stable for Hispanic parents. Over 90 percent of Hispanic parents who were cohabiting at baseline were still living together at the 12-month interview compared to about two-thirds of black parents. Furthermore, about one-quarter of black cohabitors broke up in the first year, but few Hispanic cohabitors did so. About 20 percent of black parents moved in together in the year after their child was born; 45 percent of Hispanic parents did so. More non-cohabiting black parents retained this status in the first year than Table 3.2 Change in Relationships Between Unmarried Parents Between Child’s Birth and 12-Month Follow-Up, As Reported by Oakland Mothers, by Race (in percent) Relationship at birth Cohabiting Black Hispanic Relationship at 12-month follow-up Involved in a romantic Not relationship, romantically Married Cohabiting not cohabiting involved No. 6 66 7 91 4 0 25 53 2 43 Involved in a romantic relationship, not cohabiting Black Hispanic 6 — 20 45 18 5 55 50 Not romantically involved Black — — — 100 Hispanic — — — 100 NOTE: Percentages in rows may not sum to 100 because of rounding. 49 20 19 6 25 Hispanic parents. About half of non-cohabiting, romantic relationships in both groups ended in the first year. Finally, none of the black and Hispanic parents who did not have romantic relationships at baseline reunited in the first year. Expectations About Marriage In addition to looking at transitions in relationships between unmarried parents, we are also interested in how parents view the future of their relationships at the time of birth and how closely expectations match relationship transitions. The analysis considers how expectations about marriage are related not only to marriage but also to other relationship outcomes. Previous research suggests that unmarried parents in poor communities have low expectations about marriage. Reasons for low expectations include young men’s reluctance to break from their peer groups and women’s reluctance to commit to men with negative socioeconomic or personal attributes (e.g., Anderson, 1989; Wilson, 1996). It seems likely, therefore, that unmarried parents in Oakland would have low expectations about marriage. At the same time, research on cohabiting couples has shown that most couples who live together expect to marry each other (Bumpass, Sweet, and Cherlin, 1991, p. 922). Because about half of unmarried parents in Oakland were living together, parents’ relationship status may affect overall expectations in this sample. At the time of birth, new mothers and fathers in the FFCW survey were asked what they thought the chances were that they would marry their child’s other parent. The question originally had five response categories that allowed mothers and fathers to indicate whether there was an “almost certain,” “pretty good,” “50-50,” “little,” or “no chance” of marriage in the future. For the analysis, responses are recoded into high (almost certain), medium (good/50-50), and low (little/no) categories.7 ____________ 7Conceptually, an “almost certain” chance of marriage is distinguished from the other categories because this response indicates much more confidence that a marriage will occur. The two remaining categories, “good/50-50 chance” and “little/no chance,” were grouped to reflect the distribution of responses in Oakland. 26 Figure 3.4 displays parents’ response in three bar graphs. The first bar in each set presents the most comprehensive set of responses—those of all unmarried mothers in the survey. The second bar in each set shows the responses of the subsample of unmarried mothers whose partners were interviewed (this is the same size as the number of fathers interviewed).8 Again, we were able to interview only 75 percent of unmarried fathers compared to about 90 percent of unmarried mothers. The last bar in each set presents the responses of unmarried fathers. Bars one and three compare the entire sample of mothers (including mothers whose partners were not interviewed) to a smaller and more select sample of fathers.9 Bars two and three show responses from cases where both the mother and father completed the interview. Therefore, we would expect the percentages in bars two and three (i.e., the coupled interviews) 60 50 46 42 40 34 30 37 39 38 All mothers Mothers with father interviews Fathers 29 Percentage 20 20 16 10 0 Almost certain Good/50/50 chance Little/no chance chance of marriage of marriage of marriage NOTE: Percentages may not sum to 100 because of rounding. Figure 3.4—Expectations About Marriage at Child’s Birth, As Reported by Oakland Mothers and Fathers ____________ 8One father who was in the survey did not answer this particular question, but the mother of his child was left in the analysis. 9See Teitler, Reichman, and Sprachman (2001) for more description of how fathers who were interviewed differ from those who were not interviewed. 27 to more closely resemble each other than the percentages in bars one and three. These findings show that unmarried parents in Oakland reported fairly high expectations about marriage at the time of their child’s birth. For example, about 34 percent of all unmarried mothers, 42 percent of the subsample of mothers with father interviews, and 46 percent of unmarried fathers in Oakland thought there was an almost certain chance that they would marry the other parent in the future (Figure 3.4). The fact that mothers reported lower expectations about marriage than fathers may reflect the selectivity of the fathers’ sample. These fathers seem to be in more highly involved relationships with their child’s mother than fathers not interviewed. We would expect this more select group of fathers in the survey to have higher expectations than the full sample of mothers. However, even in matched pairs, fathers appear to be more optimistic than their partners. The analysis has so far indicated that cohabiting relationships show more stability than non-cohabiting relationships in the first year. Cohabiting parents also have higher expectations about marriage (Figure 3.5). Among mothers who were living with the father when their child was born, 53 percent reported an almost certain chance of marriage (again, this group represents over half of all families in the survey). About 38 percent said there was a good or 50-50 chance of marriage. Only 8 percent of cohabitors indicated that there was little or no chance they would marry the other parent. Parents involved in romantic relationships with each other but who were not living together assessed their chances of marriage lower than cohabitors. The responses of almost half of mothers in these noncohabiting relationships appear in the middle category, indicating a good or 50-50 chance of marriage. Only 22 percent of these mothers reported an almost certain chance of marriage and about 32 percent reported little or no chance of marriage. Consistent with parents’ relationship transitions, those mothers who were not involved with the father at the time of the interview had very low expectations about marriage. Almost all mothers in this group indicated that there was little or no chance that they would marry the father. 28 Percentage 100 8 90 80 38 70 32 60 97 50 47 40 Little/no chance of marriage Good/50/50 chance of marriage Almost certain chance of marriage 30 53 20 10 22 03 Cohabiting Involved in Not involved at birth a romantic in a romantic relationship but relationship not cohabiting at birth at birth NOTE: Percentages may not sum to 100 because of rounding. Figure 3.5—Expectations About Marriage, As Reported by Oakland Mothers, by Relationship at Child’s Birth The responses to the question asking parents to rank their chances of marriage (with 1 indicating the lowest chance of marriage and 5 the highest chance) show that cohabiting parents’ mean chances of marriage are somewhat higher than 4, whereas the means of parents not romantically involved are close to 1 for mothers or 2 for fathers (Table 3.3). Means tests indicate that responses for cohabitors are significantly higher than for non-cohabitors. In this table, we are particularly interested in whether both samples of mothers give different responses than fathers within each relationship group, to see whether fathers are significantly more optimistic than mothers. The columns in Table 3.3 display parents’ expectations in three groups—the sample of all unmarried mothers, the sample of unmarried mothers whose partners were also interviewed, and the sample of unmarried fathers (as in Figure 3.4). Reading across the columns, we can again see that the responses of mothers whose partners were interviewed more closely resemble those of their partners. Although fathers’ expectations appeared to be higher than 29 Table 3.3 Means for Marriage Expectations, by Type of Relationship at Child’s Birth, Oakland All unmarried parents Cohabiting All Mothers 3.40 (1.49) Mothers with Father Interviews 3.73 (1.34) Fathers 3.88 (1.33) 4.13 (1.10) 4.19 (1.05) 4.39 (.92) Involved in a romantic relationship, not cohabiting 3.20 (1.39) 3.41 (1.29) 3.32 (1.43) Not involved in a romantic relationship 1.30 (.53) 1.31 (.48) 2.08* (1.26) No. 245 188 187 NOTES: Scale: 1 = no chance, 2 = little chance, 3 = 50/50 chance, 4 = good chance, 5 = almost certain chance. Standard errors in parentheses. *p ≤ .05. mothers’, results from means tests indicate that only fathers who were not involved with the child’s mother at the time of birth (but who agreed to be interviewed) had significantly higher expectations about marriage than the mothers of their children. Again, the fathers who were interviewed are a somewhat more select group than those who were not. Relationship Expectations versus Outcomes Given the fact that parents generally believed their chances of marriage to be high but that only 7 percent of unmarried parents were actually married one year later, it appears that unmarried parents do not anticipate well what happens in their relationships. However, parents were asked whether they expect to marry in the future not in the coming year. Although we would expect many of these changes to occur relatively soon after the birth, presumably more relationship transitions will take place after the first year. Parents’ expectations may also be an indicator of overall relationship stability, not just of the transition to marriage. In fact, parents’ expectations are closely related to whether they will marry and or live in cohabiting unions and to whether they will have no 30 romantic relationship with the other parent one year after the birth (Figure 3.6). About 82 percent of unmarried parents who reported an almost certain chance of marriage were in marital or cohabiting unions at the follow-up interview. Specifically, about 14 percent married in the first year and 68 percent were cohabiting. Only 14 percent did not have a relationship at the 12-month interview. In comparison, about 5 percent of parents who reported a good or 50-50 chance of marriage were married and slightly over half were cohabiting. About one-third did not have romantic relationship 12 months after the birth. Of those parents who reported little or no chance of marriage, about 83 percent did not have a romantic relationship with the other parent at the 12month follow-up. These results point to a clear relationship between expectations and early outcomes. These correlations are also statistically significant (Table 3.4). Results from a simple logistic regression show that the odds of Percentage 100 90 14 4 80 32 70 60 11 83 50 68 40 52 30 20 4 10 14 14 5 0 Almost certain Good/50/50 Little/no chance of chance of chance of marriage marriage marriage Not involved in a romantic relationship at follow-up Involved in a romantic relationship but not cohabiting at follow-up Cohabiting at follow-up Married at follow-up NOTE: Percentages may not sum to 100 because of rounding. Figure 3.6—Relationships at 12-Month Follow-Up, As Reported by Oakland Mothers, by Expectations About Marriage 31 Table 3.4 Odds Ratios from Logistic Regressions of Mothers’ Reports of Being Married, Married or Cohabiting, or Not Romantically Involved at 12-Month Follow-Up Married at 12-month follow-up Almost Certain Chance of Marriage 7.44** (5.47) Little or No Chance of Marriage — Married or cohabiting at 12-month follow-up 5.63*** (2.12) .07*** (0.03) Not romantically involved at 12month follow-up .18*** (.07) 19.14*** (8.64) No. 201 201 NOTES: Reporting a little or no chance predicts failure to marry perfectly. These models control for age, education, race/ethnicity, and other biological children. Standard errors in parentheses. ** p ≤ .01. ***p ≤ .001. being married among mothers who report an almost certain chance of marriage increase over sevenfold. 10 The odds of reporting little or no chance of marriage are perfectly correlated with not marrying (i.e., no one with low expectations married in the first year). The odds of being ____________ 10These results and results in Chapter 5 are presented in terms of odds ratios. Odds are the ratio of favorable to unfavorable outcomes. Odds are expressed in terms of the probability of an outcome divided by one minus the probability of the outcome. For example, 40 percent of unmarried mothers are not involved in a romantic relationship at the 12-month follow-up. Therefore, we could say that the odds of not being involved are .40/(1-.40) = .67, or .67 to one. This means that there are .67 mothers who are not in a romantic relationship for each mother in a romantic relationship of any kind (i.e., marriage, cohabiting, romantic non-cohabiting). Odds ratios allow us to compare odds between groups we are interested in analyzing. For example, we can take the ratio of the odds for mothers who report an almost certain chance of marriage and compare it to those who do not by dividing the odds of the former group by the odds of the latter group. In the logistic regressions models examined in this report, odds ratios are interpreted net of other variables controlled for in the models. In Table 3.4 we see, for example, that the odds of being married or cohabiting at the follow-up among mothers who report an almost certain chance of marriage are 5.6 times the odds of mothers who do not report an almost certain chance of marriage. 32 married or cohabiting at the 12-month follow-up were about 5.6 times greater for mothers who reported an almost certain chance of marriage than for other mothers. However, the odds of marriage or cohabitation were about 93 percent lower for mothers who reported little or no chance of marriage. Similarly, reporting little or no chance of marriage increased the odds of not being in romantic relationship at follow-up by about 19-fold. The odds of not being in a relationship at the 12-month interview are reduced by about 82 percent for mothers who reported an almost certain chance of marriage at the time of birth. 11 In other words, higher expectations are negatively related to being uninvolved with the other parent one year after the birth. Because parents’ expectations not only show us how they assess the future of their relationships but also predict early relationship transitions, this variable serves as a proxy for relationship outcomes in Chapter 5. Conclusion Data from the FFCW survey indicate that the majority of unmarried parents in Oakland were romantically involved with each other when their children were born. In fact, close to half of parents were living together at this time. About 7 percent of parents married in the first year after their child’s birth. About half were cohabiting, similar to the proportion in the previous year. Although only 15 percent of parents ____________ 11Brown (2000) shows that cohabiting couples in which both partners (or the male partner only) reported high expectations for marriage were significantly more likely to marry than couples who did not hold these expectations. I also analyzed the outcomes of couples who held different or similar expectations about marriage. Information on transitions in the first year is available only for two cities in the FFCW survey (i.e., Oakland and Austin). This analysis is also limited to couples so that we can include both mothers’ and fathers’ reports. Logistic regression analysis shows that when mothers report a certain chance of marriage and fathers do not, when fathers report a certain chance of marriage and mothers do not, and when both parents report a certain chance of marriage (compared to neither reporting a certain chance of marriage), the odds of being married or being married or cohabiting at 12 months are increased. The effect of both parents’ reporting an almost certain chance of marriage are statistically significant. Similarly, the odds of not being in a relationship with the father are decreased when mothers are more optimistic than fathers, fathers are more optimistic than mothers, and both are optimistic about marriage. The effects of fathers’ having higher expectations than mothers and both having high expectations are statistically significant in this model. 33 said that they were not involved in a romantic relationship at the time of birth, 40 percent had no romantic relationship one year one later. Parents living in cohabiting relationships at the time of birth tended either to remain in stable relationships or to marry, whereas most parents in non-cohabiting romantic relationships dissolved their relationships. About 85 percent of parents living together at the time of birth had either maintained this relationship or had married; but cohabitors were slightly more likely to end their romantic relationship than to marry. In contrast, over half of those who were involved in romantic relationships but not living with the other parent reported no romantic relationship one year later. The number of non-cohabiting, romantic relationships that dissolved in the first year indicates the fragility of these relationships. Only 3 percent of the parents (i.e., one couple) who reported no relationship at year one seem to have reunited. As a result of these changes, parents’ relationships were more likely to be divided between those that showed a high level of involvement and those with no romantic involvement one year after the birth than they were at baseline. Cohabitation at birth seems to be the key indicator of which way unmarried parents’ relationships will move in the first year. Given parents’ high expectations for marriage at birth, the fact that only 7 percent of parents married in the first year indicates that parents’ expectations were inflated—perhaps because parents were especially optimistic at the time of their child’s birth. At the same time, parents were asked about their expectations for marriage in the future, not simply in the first year, and more transitions will likely occur. Expectations show us how parents view the future of their relationships. They are significantly related to whether parents will be living in married or cohabiting union or will not have a romantic relationship with the other parent one year after the birth. This suggests that expectations are a meaningful proxy for relationship outcomes, including, but not limited to, marriage. 34 4. Insights from Qualitative Interviews As we saw in Chapter 3, parents were divided between those in nonromantic relationships and those involved with each other one year after the birth of their child. This chapter turns to the qualitative data to examine the experiences of these two groups. For the purposes of this chapter, the category of intact romantic relationships includes parents who married before the follow-up interview, those living together at follow-up, and those involved in non-cohabiting, romantic relationships one year after the birth. As discussed in Chapter 2, I conducted openended, qualitative interviews with 37 unmarried parents who participated in the FFCW survey at the time of their child’s birth and at the 12month follow-up interview. Most of these interviews were completed for both the mother and father (this group includes 28 parents, or 14 families). In other cases, I was able to interview only one parent (this group includes nine parents, or nine families). All interviews with unmarried mothers and fathers were conducted separately. This chapter draws on qualitative interviews to illustrate some of challenges to relationship stability or increased commitment parents faced in the first year of their child’s life. The first part of this chapter asks what factors contributed to the dissolution of unmarried parents’ relationships by the time of the 12-month interview, as reported by parents themselves. The second part of the chapter investigates factors that influenced parents’ plans for the future of their relationships. In particular, it explores why couples romantically involved may be reluctant to increase their involvement through marriage. Parents Not in Romantic Relationships at Follow-Up About two out of five parents were not romantically involved with the other parent at the time of the follow-up interview. Thirty-five 35 percent of these relationships ended before the birth (or were never established), and 65 percent ended in the first year. I conducted qualitative interviews with parents in each group to find out more about what happened in their relationships to precipitate the breakup. Before turning to parents’ qualitative responses, I first look at how parents in the qualitative sample responded to a survey question about relationship dissolution. In the FFCW survey, parents were asked why their relationship with the other parent ended.1 Among parents who ended their relationships in the first year, mothers from the qualitative sample reported in the survey that their relationships dissolved because of violence/abuse, “relationship reasons,” alcohol or drugs problems, and the deportation of the father.2 In the survey, all fathers in my qualitative sample reported that their involvement with the mother ended for “relationship reasons.”3 In addition, I interviewed another couple who had already broken up when I talked to them but who did not participate in the follow-up interview. In the qualitative interview, both parents agreed that they did ____________ 1Respondents were coded in up to seven categories for why their relationship ended, including (1) financial reasons, (2) distance, (3) incarceration, (4) relationship reasons, (5) drug/alcohol problems, (6) violence/abuse, and (7) other. Parents were allowed to give more than one reason why the relationship ended. Of the 37 Oakland mothers who ended their relationship between their child’s birth and the follow-up and who answered this question in the FFCW survey, 11 percent cited distance, 73 percent cited relationship reasons, 5 percent cited drug or alcohol problems, and 11 percent cited violence or abuse as reasons why they were no longer involved with the father. Of those 21 Oakland fathers who responded to this question in the FFCW survey, 5 percent cited financial reasons, 5 percent cited the mother’s incarceration, 95 percent cited relationship reasons, and 5 percent cited abuse or violence as reasons why the relationship ended. 2 One mother I interviewed was not asked why the relationship ended because she said she was uncertain about who the father was. “Deportation” was not a response category but was written in for two mothers. 3One father was skipped from this question because the couple was still together at the time of the interview, but he was deported shortly after. Two additional fathers were not asked why their relationship ended because they revised their relationship status at birth when asked about it in the follow-up. That is, they reported that they were romantically involved with the mother in the baseline interview. However, when asked one year later about their relationship status at birth, they said they did not have a romantic relationship with the mother at that time. Their former partners, however, did classify the initial relationship as romantic in both interviews. One of these mothers said the relationship ended because of violence/abuse and the other mother attributed the breakup to “relationship issues.” 36 not have a serious romantic relationship at the time of birth and that the father reunited with the mother of his older children in the first year. Among the two parents in the qualitative sample who had broken up before the birth of their child, the first said their involvement ended for “relationship reasons” and the second parent’s response was coded “other.”4 Relationship Ended in the First Year The reasons parents gave in the qualitative interviews about why their relationships ended were basically consistent with the survey responses. The qualitative data provide more in-depth information about why couples broke up and illustrate the multiple challenges parents faced as their relationships were dissolving. These open-ended interviews are particularly helpful in allowing us to unpack the “relationship reasons” response most frequently reported in the follow-up survey. According to qualitative reports, “relationship reasons” point to high conflict in the relationship around issues such as trust, fidelity, and commitment. These interviews also help put other responses about what triggered the breakup, such as drug use, violence, and deportation, into a larger context. Furthermore, the qualitative data point to additional problems with financial and housing instability, which appear to have contributed to the dissolution of parents’ relationships. Unlike most parents who ended their relationship in the first year, Marian and Daniel were living together at the time of their child’s birth.5 In the qualitative interview, Marian attributed the couple’s breakup to an argument in which Daniel hit her. Although this serious event motivated the final breakup, Marian explains that it was preceded by prior breakups, the last of which resulted from Daniel’s “cheating.” ____________ 4Of the 24 Oakland mothers whose relationship ended before the birth and who responded to this question in the FFCW survey, about 13 percent cited financial reasons, 17 percent cited distance, 20 percent cited relationship problems, 13 percent cited drug or alcohol problems, 13 percent cited abuse or violence, and 30 percent cited other reasons why the relationship ended in the FFCW survey. Of the 11 Oakland fathers who responded to this question, 36 percent of fathers cited financial reasons, 18 percent cited distance, 9 percent cited the mother’s incarceration, 36 percent cited drug or alcohol problems, and 27 percent cited other reasons the relationship ended. 5All names used in this report are pseudonyms. 37 Marian recounted what was happening in their relationship at the time she decided to end the relationship for good: Once I got this place and was pregnant with my daughter, [the problems] started. He stayed here and it was like he would start leaving for weeks . . . and we got into an argument one night, and he hit me. . . . Before he cheated on me, so we broke up because of that. Then we were talking about getting back together and we got into an argument and he hit me, and that’s when I decided [to end things]. Marian also found it difficult to trust the father because he had been in and out of jail for selling drugs. Although she says that he promised he would not sell drugs again, he began leaving for “weeks at a time” to do so. Katherine reported that her relationship with the father ended because of problems related to drugs and alcohol. Both she and the father were sentenced to jail at the same time on drug-related charges. In the qualitative interview, Katherine explained that she did not find out she was pregnant until she was already incarcerated. The couple corresponded during the pregnancy and planned to marry when they were released from jail. Having spent half her sentence in drug rehabilitation, Katherine hoped to start a new, drug-free life when they got out. However, she broke up with the father (and broke off their engagement) two weeks after he was released, explaining that he clearly “wanted to continue in the drug scene” and she couldn’t risk going to prison again, for her own and her children’s sakes. In the survey, Leticia and Ricardo both reported that they broke up for relationship reasons. When I spoke to these parents, they had recently had a serious argument and said that their relationship was “pending.” By the follow-up interview, the relationship had dissolved. Leticia’s two major concerns about the relationship revolved around trust issues and Ricardo’s reluctance to commit to a more serious relationship. The couple had not been dating long before Leticia became pregnant, and concerns about trust went both ways. Leticia said they argued frequently about Ricardo’s relationship with another woman, whom, she believes, he was romantically involved with. During the pregnancy, she said they also broke up for a while when Ricardo expressed some doubt about whether he was the baby’s father. 38 At the time of the interview, Leticia had given Ricardo an ultimatum that she would leave him unless he agreed to settle down with her. According to Leticia, Ricardo did not act committed to the relationship and would not make plans for them to live together as a family. Leticia was struggling to raise another child alone and did not want to go through this experience again with another father. She explains, “It’s stressful because you don’t know what he really wants. . . . I don’t know if he just expects me to wait around.” Although Ricardo considered their relationship serious, he thought they argued too frequently about insignificant issues. More important, he thought it would be pointless to reconcile because they were incompatible as partners and their arguments would likely continue. Other parents ended their relationships by the time of the 12-month interview but after I interviewed them. When I spoke to these parents, however, they were experiencing obstacles to keeping their relationships together. In the survey, Norma attributes her breakup with Miguel to his recent deportation. Miguel was interviewed for the follow-up survey shortly before Norma and before he had been deported. Norma did not anticipate this event, but she discussed how his immigration status contributed to other issues that made relationship stability difficult. When I spoke to Norma and Miguel, she was working part time and Miguel was working under the table. The couple met when they were staying in the same building (as neighbors), but they were forced to move out before their child’s birth. Norma and Miguel were currently living with their families. She explained that the couple could not get their own place because she did not make enough money to get an apartment and could not put Miguel on the lease because he was not in the country legally. She applied for a Section 8 housing voucher but said she is on a very long list with other low-income families. Because the couple could not find a way to live together, Norma did not feel as if they had a foundation upon which to build a marriage or family. She explained: “We were supposed to be getting married last month. We postponed it cause we had no money. How can we be married if he lives there and I live over here?” Norma stressed that the couple argued frequently and that many of their arguments arose because Miguel was not living with her and the baby. Because they lived apart, 39 she felt as if the baby was not getting to know him and that Miguel was not helping her take care of the baby. Furthermore, she believed they had more problems with trust and communication because they lived in separate households: “He [does] not know what I’m doing, and I don’t know what he’s doing. We don’t know how to trust each other. And we got a lot of he said/ she said interactions going between us. A lot of conflicts.” Although Miguel also thought living apart was an issue in their relationship, he did not seem as dissatisfied with their living arrangements as Norma. Miguel agreed that the couple could not move in together because of financial difficulties. However, Miguel did not think that the couple argued frequently. Because Miguel’s deportation forced an end to their relationship, we do not know whether this couple could have overcome these problems and established a stable relationship. This example reminds us that many unmarried parents in Oakland are foreign-born and may face distinctive challenges in their relationships. Relationship Ended Before the Birth Lashae, who was living in a drug rehabilitation center when I interviewed her, said her relationship with her baby’s father ended two months before the baby was born. Lashae explained that they had seen each other for a few years before the pregnancy and “it got serious for a minute, but we kind of drifted away, and he was cheating.”6 Lashae described the father’s response when she found out she was pregnant: He was like—he wasn’t ready. I wasn’t really ready either, but I didn’t really agree on abortion. So I just went on and had her whether he was gonna do something or not cause I’m already a single parent, so I figured I can do it— I’m already doing it. But he said he wasn’t ready, wasn’t financially stable. But from what I’ve heard, he has had jobs before. Lashae says that the father has seen the baby only once or twice since the birth and has not offered any financial support. Although she does not want a relationship with the father, Lashae would like him to be ____________ 6In the survey, Lashae’s response about why the relationship ended did not fit into one of the response categories in the survey. Therefore, it was coded “other.” 40 involved with their child.7 Lashae did not have a job and was not receiving financial support from the baby’s father. Therefore, she was primarily concerned about how she would support her family when she lost her welfare benefits in 6 months.8 Parents in Romantic Relationships at Follow-Up As we saw in the previous chapter, about 60 percent of parents who were unmarried at birth were romantically involved with the other parent at the time of the follow-up interview. This section explores the obstacles parents experienced in their relationships during the first year. Given that many parents were in stable relationships, such as cohabitation, in the first year, it also investigates why parents postpone marriage or do not anticipate marriage. Married at Year One Only one of the unmarried couples I spoke to in the qualitative interviews married between the baseline and 12-month follow-up. Monica and Jack had lived together for about nine years before marrying and had three children together. Initially, Jack moved in with Monica and her father because he did not have a permanent place to stay (Monica was a teenager at this time). For much of their relationship, Monica and Jack had been struggling with problems such as methamphetamine addiction and financial instability. Monica and Jack each described their marriage as the result of a particular set of circumstances that converged when they “bottomed out” on drugs. Intervention from his family, Child Protective Services (CPS), their church, and their drug rehabilitation program prompted them to stop using drugs and “get themselves straight.” Because the couple already considered themselves in a common law marriage, they said it was not difficult for them to formalize their relationship. According to Jack, ____________ 7Unmarried fathers’ involvement with children is an important issue and will be the focus of a future report on the Oakland FFCW study. 8Lashae was facing time limits on welfare. 41 We’ve always wanted to get married, you know. From day one, you know. But it took us this long to finally get married, you know, because of financial reasons. . . . We’ve never been really financially stable. You know. Bank account, things like that. We’ve never had things like that . . . and then we had another child, and then another child. And we kind of put that off. We knew we were together as a couple. Common law marriage, and that was fine. . . . We were probably going through this crisis or that crisis in our life . . . every time we’d say, “Hey, let’s get married next year.” . . . Something was always going on in our life at that time, and it just never happened. I think when we finally got our acts straight last year, then we decided to set things right. But it was never, you know, like I told people, it was never a thing of I didn’t want to marry her or she didn’t want to marry me. We always felt like there were other things to do at the time. Although it is hard to generalize from Monica’s and Jack’s experience, their transition to marriage was probably greatly facilitated by the fact that they had been living together in an informal union for many years. When they stopped using drugs, Monica and Jack were also able to become more financially secure. In this case, overcoming their drug addiction seems to have had a larger effect on their lives than marrying. However, their reasons for delaying marriage may shed light on why other parents in committed relationships do not marry in the first year. Involved in Other Romantic Relationships at Year One Tammy and Jeff were living together when their child was born. When I contacted them for the qualitative interview, they were staying at a shelter with their baby until they could find affordable housing (the apartment building where they had been living was condemned). During the qualitative interview, the couple expressed high hopes for marriage. When I asked Jeff if he considered his relationship with Tammy to be serious, he replied, “I’m gonna marry her.” Jeff added that he would like to be married by next year but wanted to have more financial and residential stability before taking this step. At the same time, Jeff was optimistic that he would get a good job, similar to the ones he held in the past: I want to have things right before we marry. You know. I want to have our own place, and [be] financially established. . . . I’d love to be able to do what I want to do now. But, I look at it like this: I’ve had all these things before . . . and I’ll have it again. It’s just a matter of time. You know. And I’m a patient 42 man. Patience is a virtue. It’s nothing that can hurt you. Everything can make you stronger. Huh buddy? [to baby]. Although Tammy had no reservations about staying with Jeff, she also wanted to be sure she was “ready” and that the circumstances were “right” before they married.9 My main thing is just making sure everything’s right, and we’re ready for it. Not just one of us, but both of us. . . . Jeff makes me feel like he’s the right person. And I know he’s ready for it. I believe in time I will be ready for it. I’m just still a little young, and I got my baby. And I don’t know, saying “I do” is a big thing . . . because it’s like saying that you’re going to be with him for the rest of your life. I would love to be financially ready, you know. If I had my own place, you know. My own car instead of the bus. You know. Things like that. Despite this hesitation, Tammy felt that Jeff was the right person because he is a good father to their child and has stood by her through difficult times. For example, the couple met after Tammy had been kicked out of her own family’s home. Soon after, Jeff offered to take her in. Jeff also stayed committed to Tammy, despite his estranged wife’s attempts to reconcile with him. Susan and Timothy were also living together at both interviews in public housing. When I spoke to them, Susan had gone back to work full-time and Timothy occasionally worked under the table. Timothy also took care of the baby and Susan’s other children while she worked. Susan and Timothy talked about how financial problems and conflict in their relationship affected their outlooks on marriage. Recounting how they went back and forth about this subject, Susan says: ____________ 9In qualitative interviews, I asked parents what they believed were the most important things to consider before getting married and what makes a marriage or relationship work. In addition to love, parents emphasized the importance of financial stability (including housing stability), communication, compatibility (including the ability to get along), and trust (including fidelity). For parents with children by other partners, an additional concern was whether their baby’s other parent would accept the extra emotional and economic responsibilities they bring to the relationship. These general considerations came into play when parents talked about their own plans for the future. 43 We just talked about it [marriage]. . . . He had give me a ring, and proposed to me. And I got mad, and I broke it, and I gave it back to him, and he threw it away somewhere on the creek or somewhere. . . . And he’s talking about it again. He wants us to get married. So, I was like, okay, I want to. In the interview, Susan expressed a great deal of love for Timothy and thought they were highly compatible, referring to Timothy as her “other part.”10 However, she admits that the relationship has been rocky at times. At the end of the pregnancy, she said, “our relationship just started falling apart” and had just recently improved. Both parents attribute their interpersonal conflict to things like the other person’s “attitude,” lack of patience, sensitivity, temper, and jealousy. However, when asked about their plans to marry, Susan emphasized the couple’s lack of economic resources. Susan is optimistic about their relationship and chances that things will improve for them: “hopefully next year, something will change. . . . Maybe next year we won’t have to struggle as much.” Timothy expressed concerns about their relationship following some recent conflicts. However, when I later asked him whether they were planning to marry, he agreed with Susan that they would be married by next year. Although he was nervous about marrying someone with other children to support, he explained: “we going to probably [get married], when we get some more money. That’s the only thing keeping us from getting married.” And when asked if he could imagine changing anything to make it easier to have a relationship, Timothy responded simply: “Her a better paying job, and me a job.” Similarly, Timothy thought the most important thing to consider when thinking about marriage is: “Income. . . . The whole household income. Period.” Other parents also described how problems in their relationships, personal problems, and the stress of being new parents occurred simultaneously with economic problems. Marta and Roberto were living together at the time of their child’s birth and 12 months later. Like Tammy and Jeff, Marta took Roberto in to live with her when he had no place to stay. Roberto was hopeful about their future together, saying, “I ____________ 10Susan says “my other part” to interpret a common Spanish phrase “mi media naranja.” The English translation of this phrase is “my half orange.” 44 already feel married.” However, he acknowledged: “We got our little knick-knacks, we gotta straighten up like any relationship. But we’re gonna straighten those up.” Roberto was unemployed and Marta was receiving welfare when I spoke to them. Roberto explained the decision to delay marriage in terms of these financial difficulties saying, “Just like the position we’re at now. It’s just hard to really budget and have a really good marriage.” Like other parents, he wanted to wait for the right time to marry. However, he acknowledged that this decision could be put off indefinitely if they tried to wait for an ideal economic opportunity: We want to have a real good marriage. . . . It’s never going to be good enough once you start having kids. It doesn’t matter. They’re in school, they’re out of school. . . . It’s never the right time. Sometimes, maybe it is a bad time. There’s a lot of bills . . . all the time. You still got to budget and get through it, you know. . . . I mean, some people, they got it made. They just got it made, you know what I’m saying? Some people got it bad, real bad. . . . You just don’t leave somebody because life’s bad. Marta—who said she had been depressed since the baby was born— expressed more uncertainty about their relationship than did Roberto. Marta identified specific problems in the relationship, such as frequent arguments and Roberto’s prior problem with drugs, which concerned her. She also discussed the history of their relationship with some ambivalence: I don’t know why, but we’re still here. Still together. Roberto has a good heart. He has a lot of bad habits. A lot of things that I can’t even under . . . explain, but the main thing that I do know for a fact is that he does have a good heart. Marta characterized their relationship as highly volatile and says that Roberto is like “two different people” sometimes. According to Marta, his emotional unpredictability affected their plans for the future: At first he was like, yeah, he wanted to marry me. Then it’s like—when he turns into that other person that I don’t know, he’s like “who would want to marry you?” and then when he’s back to [being himself], he’s like “I want to marry you” again. It’s like a yo-yo. 45 Because she wanted to have a marriage that lasted “forever,” Marta said that she would consider how they are getting along, whether or not they were employed, and whether Roberto was using drugs before making any decisions about marriage. We’d have to make sure that the baby’s in a good environment as far as arguing. And, jobs - we’re working. No drugs at all. . . . If I’m going to get married, I want to stay married. I don’t want to go through the divorce. I want to be sure that it’s going to be forever, you know. Although Marta wanted things to improve for them financially and emotionally, she was unsure about this happening. Finally, Gina was romantically involved with her child’s father but was not living with him one year after the birth. Although they lived together before, they could not afford to keep the apartment. When I asked her about whether they had thought about marriage, Gina said she would prefer to be cautious rather than jump into a marriage that could easily fall apart because of financial or personal problems. She had some doubts about the father’s character and temperament and wanted to make sure that he was responsible, stable, and mature before considering marriage. She also wanted to have more confidence in her feelings for him. Gina talked about her own considerations about marriage in reference to the high incidence of divorce: Me speaking personally, I’m seeing all these divorces, and I don’t want to end up like that. I don’t want to have to do any of that whole thing. I don’t want to have to go through drag[ging] my kids to court, or whatever. And it’s just a scary thing. I want to be sure, you know. . . . People are taking more time to see if this is the person that they really want to spend the rest of their life with. And too, financially, people aren’t financially set like they want to be. . . . Not to say I couldn’t gain that after I get married, I just don’t want to go into it with a whole bunch of problems. Although couples in intact relationships were often considering marriage, and some were even planning to marry in the next year, parents suggested that they wanted to move into marriage cautiously. Because parents believed that their relationships could easily fail, they wanted to make sure their partner and the conditions were “right” for marriage and they were emotionally and financially “ready” to marry. In time, parents 46 were often hopeful that their economic situations, relationships, and personal circumstances would improve, making marriage possible. Although waiting appears to be a rational decision for couples experiencing multiple challenges to their relationship, as Roberto acknowledged, low-income, unmarried parents may have trouble ever reaching a point where they feel economically prepared for marriage. And more doubts may arise as they wait. These findings are consistent with previous research on unmarried parents’ orientations toward marriage (Luker, 1996; Wilson, 1996; Waller, 1999; Edin, 2000; Furstenberg, forthcoming). Fragile Families in the First Year This chapter concludes with a more complete description of a couple in the study who, like the majority of parents in Oakland, were living together at the time of their child’s birth and one year later. Cynthia and James had been together for close to four years and were trying hard to succeed as a family. Despite their economic difficulties, Cynthia and James talked enthusiastically about their relationship and their future together. James recounts the story of how the couple met and why their relationship is special: The first day, my momma enrolled me [in school], she was in with her momma, being enrolled, too. And I saw her that day, and we looked, and like something clicked. . . . We saw a movie, chilled one time. . . . Took her back to her house. . . . I gave her a little kiss on the cheek, gave her my phone number to call. We talked till about four in the morning. The next thing I know, I wake up with the phone still on my ear, but it’s going “uh uh uh.” I was like, man. So the next morning, I went and picked her up for school, we walked to school together. So from that day on, everything we did is how we do it now—with each other. So there ain’t no flaws to me in the relationship. It’s cool ever since that day. Christmas, it’ll be four years. All my friends look at me like “I don’t see how you do it.” You just gotta have patience. If it ain’t real love, you ain’t gonna do it. . . . All the time, my friends going “Man, you leavin’ us—you ain’t been with us no more.” “Hey man, I just trying to make this work right here. You don’t want to try to make it work?” Cause my friends, none of their relationships gonna be successful. Although the couple has been living together for a number of years, they have had difficulty maintaining a stable residence. James moved out 47 of his family’s home and into Cynthia’s home in high school, explaining that it was too crowded. When it became difficult for the couple to stay with Cynthia’s mother after it became too crowded there, the couple’s families helped them move into their own apartment. At this time, Cynthia had a one year-old son from a previous relationship and they were both attending school. However, they had to move out of one apartment because Cynthia’s welfare check was cut and they could not afford the rent. They were then evicted from another apartment after losing a lawsuit filed with other tenants against the landlord. Cynthia’s mother allowed them to use her house as a fallback, but it was difficult for them to stay indefinitely, given that this added four members (Cynthia, James, their baby, and Cynthia’s older child) to the household. Her mother’s rent was also subsidized by a Section 8 voucher and required her to report the other tenants living in the home. Cynthia explains their current living situation: He lives here, then sometimes he goes back to his mom’s, but then he comes over here every day cause he like to be around little James. . . . [My mom] say we gotta be married in order to stay together. . . . She say it alright for him to stay here . . . only if he helping me with the baby. . . . He gotta sleep in here, and I got to sleep in there with the baby. Both Cynthia and James want to keep their family together. James already acts as a step-father towards Cynthia’s older son and they think of themselves as being in a marriage-like relationship. According to James, We’re planning on getting married on her birthday. We planned it like a year ago. I really don’t think it’s gonna be too different, cause I feel like we married right now. Just by . . . how everything operate—how we work. It feel like that now. It wouldn’t really be no different. It would be a good moment, that’s probably about it. Like other parents, however, Cynthia and James said they wanted to be ready to take this step. Cynthia is serious about the relationship and thinks James’ actions show that he is too: It seems like James want to be with me and be willing to help me out. He’s doing a lot to try and get a job and make sure we get us a place to stay. But I ain’t seen no doubt [from him] cause he be around me. He want us to be together, and he make sure everyone’s happy. 48 Before marrying, Cynthia would ideally like them both to have jobs, child care, and a place to live on their own. Cynthia suggests that their educational limitations prevented them from finding steady work in the past.11 Cynthia described what she wanted for the future: My mom say, “I think you guys can make it. You all be a happy family.” . . . I think we might make it. . . . We talk about that [getting married] a lot. . . . First I want to get in school, get me a nice job, find me a nice apartment where I can [be] set and not worry about nothing. . . . The best thing for us would be [if] James would find a good paying job, and they would be willing to help us out so everybody has the benefits—dentist or whatever—and for us to have a nice house. [And] for little James to have child care. . . . There’s a lot of jobs out there, but it’s like you gotta have a lot of things in order to get that job. Like a high school diploma or GED. Or you gotta have some experience in that job. I think that would be best for us. James and Cynthia have a strong commitment to each other and want to raise their children together. However, we see that they have both positive and negative forces affecting their relationship. On the positive side, they have a close relationship, a shared commitment to raising their children, and support from their families. On the negative side, they are young, do not have education beyond high school, have difficulty finding jobs, and cannot afford to live independently as a couple. Their future may depend upon whether these positive or negative factors tip the balance of their relationship. Conclusion Unmarried parents who ended their relationships before the 12month follow-up interview sometimes mentioned a specific event, such as an incidence of violence, a drug relapse, or deportation, that triggered the breakup. Other parents talked about more general problems in their relationships, such as conflicts around issues such as trust, infidelity, and commitment. Although parents were more likely to point to a particular event or problem that prompted the breakup than to talk about economic problems, their accounts indicate that financial instability, including not having a place to stay together, occurred simultaneously ____________ 11James called me a month after the interview to report that he had found a job. 49 with these problems and most likely contributed to them. In most cases, parents were facing multiple challenges in their relationship during the time that it ended. Parents who stayed together in the first year mentioned experiencing many of the same kinds of economic, relationship, and personal problems (such as financial and housing instability, frequent conflict, trust issues, and drug problems) as parents who had did not have intact relationships one year after their child’s birth. Financial problems were mentioned more often by parents in intact relationships (when they described their plans for the future) than by those in nonintact relationship (when they identified why the relationship ended). This is perhaps because the financial problems are endemic to both groups and are not seen as prompting a breakup. Because parents who stayed together in the first year believed that their relationships could easily fail, they wanted to make sure that their partners and the conditions were “right” for marriage and they were “ready” to do so. These parents—particularly unmarried mothers who may have to assume primary responsibility for parenting—also wanted reassurance that their partners are reliable and will be there through difficult times. Often they were hopeful that their economic and personal circumstances would improve, making marriage possible. Despite these hopes, unmarried couples are likely to experience continuing economic problems. In addition, more doubts may arise as they wait. Of course, couples in the general population share some of the problems reported by these unmarried couples, which may explain why a large number of men and women are delaying marriage. However, the material hardship experienced by most parents at the Oakland site exacerbated common relationship problems and introduced new ones. It is possible that couples who stayed together had less severe or immediate problems in their relationships than couples who broke up. It is also possible that they were better able to negotiate these problems during trying times in the first year. We do know that most of the couples who remained intact during the first year were living together. It 50 is likely that these parents established closer emotional or financial ties before moving in together or after they set up a household together.12 The next chapter returns to the survey data to examine how expectations about marriage of parents in the larger sample are related to issues parents say contributed to relationship dissolution or influenced their decisions to become more involved with the other parent. ____________ 12This increased interdependence may also inhibit breakups in the future. Thanks to Laura Moye for observing that cohabiting couples are more interdependent than noncohabiting couples. 51 5. Factors Underlying Early Expectations This chapter investigates factors that underlie parents’ expectations about marriage. In particular, the analysis considers how issues identified in the qualitative interviews, such as parents’ economic circumstances, distrust, conflict and physical violence in their relationships, and problems with drugs and alcohol influence how they assess their chances of marrying each other. Expectations reflect parents’ own sense of the potential for marriage, and the factors that influence expectations are likely to be the same as those that influence outcomes. Because the question about parents’ expectations asks them to assess their chances of marriage to the other parent in the future, these results may also point to changes in parents’ relationships beyond year one. Ideally, I would also model transitions as measured by relationship status several years after the birth. These data will ultimately be part of the FFCW study, but at this point, only the 12-month follow-up data have been collected and these are the only data available for two cities.1 Because the size of the follow-up sample in Oakland is not large enough to permit multivariate regression analysis, parents’ assessments of whether they will marry their child’s other parent in the future serve as a proxy for relationship outcomes.2 The following analysis uses data from the first wave of the FFCW survey conducted at the time of birth. It also brings in information from qualitative interviews to discuss important findings. This analysis includes responses of 3,712 unmarried mothers and 2,775 unmarried fathers living in Oakland and the 19 other cities in the FFCW survey. ____________ 1Follow-up data are available in Oakland and Austin. 2When I regressed the right-hand-side variables from the expectations models on relationship outcomes at year one for the two cities for which data are available, none of the results were invalidated. 53 Because the models control for significant interactions between parents’ responses and their residence in Oakland, the 20-city results can be applied to the Oakland sample.3 The analysis compares the responses of parents who report an almost certain chance of marriage to those of parents with lower expectations, using separate logistic regression models for mothers and fathers. Regressions were also run to examine the responses of parents who report little or no chance of marriage; differences between the findings in these models are indicated at points in the text and in the footnotes. Regression models were run with and without parents’ cohabitation status. Living with the other parent is expected not only to raise marriage expectations but also to mediate the effects of other variables because parents with certain characteristics are more likely to live together. The discussion primarily focuses on the model that does not include cohabitation (summarized in Table 5.1). However, the effect of parents’ living together on their marriage expectations is addressed in the final section of this chapter and in footnotes. The means for variables Table 5.1 Factors Related to Parents’ Expectations About Marriage Father’s employment Mother has high school diploma Mother has more than high school diploma Women/men cannot be trusted to be faithful Frequent arguments (scale) Mother/father hit or slapped Mother/father has drug/alcohol problem Mothers: Fathers: Certain Chance Certain Chance + n/s + + –– –– –– –– NOTE: These results are from the models without cohabitation. + = significant, positive. – = significant, negative. n/s = not significant. ____________ 3The models were run with full interactions of parents living in Oakland and the independent variables. Significant interactions were left in the models. See Appendix Table B.2 for the means of these interaction variables. Appendix Tables B.3–B.6 indicate which interaction variables are used in each model. 54 examined in the models and the odds ratios from the logistic regressions are reported in Appendix B. 4 Men’s Employment and Education A large body of research shows a positive connection between male employment and marriage. Researchers also argue that high levels of joblessness in poor communities have rendered men less “marriageable” in the eyes of women (Wilson, 1996, 1987).5 Because most unmarried parents in the FFCW Oakland sample are poor or near-poor, we might expect women with partners in the labor force to view them as having greater potential for marriage. Fathers with jobs might also feel like they are more economically prepared for marriage. Education is an indicator of future earnings, and fathers with higher levels of education should also have higher expectations about marriage. In Oakland, about 57 percent of unmarried mothers said their child’s father was employed the week before the child’s birth, whereas about 67 percent of unmarried fathers reported being employed. 6 Fathers who were not working but were in school most of the time during the week before the interview (about 5 percent of unmarried fathers in Oakland) were examined separately because they could be considered to be better marriage candidates than those neither employed nor in school. Because almost one out of ten mothers did not report whether the father was working, a missing variable was created for this response. About 90 percent of unmarried mothers who did not report employment information on the father responded that they did not know how he spent most of his time the week before the interview. These mothers are likely to be out of contact with the father. ____________ 4See Chapter 3, footnote 11 for a description of odds and odds ratios. 5Exacerbating the shortage of employed men is an unequal sex ratio between lowincome black men and women resulting, in part, from high male incarceration and mortality rates (Tucker and Mitchell-Kernan, 1995). 6Although the models for unmarried fathers include fathers’ own reports of employment, models for unmarried mothers use women’s reports of men’s employment. This allows all unmarried mothers to be included, rather than just those with a father interview. Mothers’ perception of whether the father was employed may also influence their expectations about marriage more strongly than men’s actual employment status, to the extent these differ. 55 The model for fathers also includes information about their educational status. Fathers without a high school diploma are compared to parents with a high school diploma or GED, fathers with education beyond high school but not a college degree, and fathers with a college degree. As we saw in Chapter 2, about 47 percent of Oakland fathers had less than a high school diploma, about 33 percent had a high school diploma or GED, about 19 percent had more than a high school diploma (but not a college degree), and about 2 percent had a college degree. The analysis shows that unmarried mothers who said that their baby’s father was employed in the week before the interview have 38 percent higher odds of reporting an almost certain chance of marriage than those whose partners were not working (see Table 5.1 for a summary of regression results and Appendix B, Tables B.3–B.6, for all odds ratios).7 Among mothers, not reporting employment information for the father is negatively related to marriage expectations. The odds of indicating an almost certain chance of marriage among mothers who did not report father’s employment information are about 76 percent lower. Again, most of these mothers say that they do not know the father’s employment status. Fathers’ being in school the week before the interview does not significantly affect mothers’ expectations. Unlike the case for mothers, employment is not related to fathers’ having high expectations. However, fathers’ employment status is significantly related to fathers’ reporting little or no chance of marriage, indicating that fathers with jobs do not have low expectations about marriage. Fathers’ education does not have a significant effect in this model (which excludes cohabitation).8 Qualitative interviews shed light on how fathers’ economic characteristics influence unmarried parents’ assessments of marriage. When I asked parents about their considerations for marriage, they often emphasized the need for financial security. Whether the father has a stable job seems to represent an important part of parents’ feeling ____________ 7Fathers’ employment status is not significantly related to whether mothers report little or no chance of marriage. 8Fathers’ education is significant in the model with cohabitation. 56 financially secure. For example, Leslie lived with her baby’s father before their baby was born. He did not have a job at that time, and she explains their decision to stop living together “was more so [for] financial [reasons] . . . because he was hurt, and he wasn’t working, and I didn’t want to take on the strain of trying to keep the apartment.” She explains that she would like to make sure this job lasts before they make a decision about marriage: He has a good job, now . . . but I want him to be on the job for a while. Not to say, “Okay, I have a good job” or then when I’m tired, or “I quit, I’m gonna get another job.” I just want him to be stable, you know. Somewhere stable. And I think a lot of people are starting to think like me. . . . Once you’re married, you take on their problems . . . as far as financial and otherwise. And now I’m looking at the total picture, and long term the things you have to [think about]. . . . And it’s hard for me. I just can’t say “yes”. . . since I have kids, something’s telling me to wait. I don’t know what it is. It’s just telling me, wait. Leslie’s response is consistent with those of parents in the previous chapter: She would like to be financially ready for marriage. However, Leslie expresses doubts about the father’s financial stability despite the fact that he recently found a job, suggesting that she has additional reservations about marriage. Women’s Employment, Education, Public Assistance, and Household Income The analysis also considers how various dimensions of mother’s economic status—their employment, education, welfare use, housing assistance, and household income—affect parents’ expectations about marriage. Mothers with earnings and higher earnings potential may believe that their chances of marrying the father are better because they could contribute to the household income. Because women’s employment allows them be more independent, however, working may make marriage less desirable to them. A related hypothesis is that welfare and housing assistance acts as a disincentive to marriage, if mothers believe that they can keep benefits only by remaining single (e.g., Murray, 1984). Because women in the baseline survey had just given birth, they were not asked whether they were working during the previous week. 57 Therefore, the variable measuring women’s participation in the workforce is based on whether they had earnings in the last year. Similarly, the variable for welfare receipt indicates whether women received income from public assistance, welfare, or food stamps in the year before the birth and the household income variable is reported for the prior year. The variable for housing assistance asks if the mother is currently living in a housing project or receiving help from the federal, state, or local government to pay rent. As we saw in Chapter 2, about 35 percent of mothers in Oakland had earnings from work the year before the birth.9 About 52 percent had less than a high school diploma, 31 percent had a high school diploma or GED, 15 percent had some education beyond high school, and 2 percent had a college degree. In addition, about 53 percent of Oakland mothers received welfare or food stamps and about 24 percent lived in public or subsidized housing. Mothers’ average household income was about $19,600.10 The regression results indicate that mothers’ receipt of earnings in the previous year was not significantly related to how they assessed their own chances of marriage. Because mothers were pregnant in the prior year, their educational level may be a better measure of how they view their earnings potential. The findings show that mothers who have education beyond high school believe that their chances of marriage are higher than other mothers. The odds of reporting an almost certain chance of marriage are about 25 percent higher for mothers with a high school diploma and about 39 percent higher for mothers with more than a high school diploma but not a college degree (see Table 5.1). Results from the regression models indicate that mothers’ receipt of welfare or housing assistance is not significantly related to how parents assess their chances of marriage. The household income variable was included in the regression primarily as a control to see if there was an additional effect of receiving welfare or housing assistance after income ____________ 9In the total sample, about 67 percent of unmarried mothers had earnings the prior year. 10Household income is a continuous variable and is divided by $10,000 in this analysis. 58 was taken into account. Although mothers with higher household incomes are significantly more likely to report an almost certain chance of marriage, there is only a modest change in the odds of mothers’ giving either response. For example, the odds of reporting an almost certain chance of marriage increase about 6 percent for every $10,000 increase in mothers’ household income. Consistent with the findings for mothers, receipt of earnings, welfare, and housing assistance is not significantly related to fathers’ marriage expectations.11 Furthermore, mothers’ household income has no significant effect. Because mothers were pregnant in the last year, education may be the best indicator of future economic stability. In particular, education that leads to good paying jobs may also increase the chance that couples will be more financially prepared to marry. Before making definite plans to marry the father, Rene says she would like to receive a high school diploma: I’m going to go back to school and finish. . . . I’d rather take my high school diploma than my GED cause most of these jobs tell me to go get my diploma, then they give me a call back. . . . [By next year] I want to be out of school and I want to be able to work. In addition to looking at how economic factors influence marriage, we are also interested in the effect of trust, conflict, violence, and drug/alcohol problems on parents’ expectations. Figure 5.1 presents the odds ratios of some of the significant variables in the model for mothers who report an almost certain chance of marriage. It is important to keep in mind that positive odds ratios may look larger than negative odds ratios in this figure because odds cannot go below 100 percent in the negative direction. We can see that the effects of noneconomic factors ____________ 11Mothers’ earnings are significantly related to whether fathers report little or no chance of marriage. This finding could indicate that fathers consider mothers with earnings to be less-desirable spouses (perhaps because they are more independent). At the same time, fathers who report a little or no chance of marriage are less likely to be in romantic relationships with or living with the mother. Therefore, mothers who had children with these men may be working more often because they are not sharing resources with the father. 59 60 40 38 39 20 Percentage 0 –20 –40 –60 –80 –100 –55 –67 –88 Father employed Mother has more than high school diploma Men cannot Frequent be trusted arguments to be faithful Father hit/ slapped mother –69 Father has drug/ alcohol problem Figure 5.1—Odds Ratios for Mothers’ Reports of an Almost Certain Chance of Marriage are large. In this model, the variable indicating frequent arguments has the largest effect. Gender Distrust Parents’ general beliefs about how men and women are likely to act in relationships should also be associated with marriage expectations. In particular, scholars suggest that feelings of distrust between men and women may be heightened in economically unstable and unpredictable environments (Furstenberg, forthcoming).12 Although academic discussions of “gender distrust” have primarily focused on women’s trust issues, this analysis examines whether mothers and fathers who do not trust members of the opposite sex to be faithful—an important ____________ 12According to Furstenberg (forthcoming), a culture of gender distrust encourages women to “monitor relationships closely and detect the early warning signs of distress.” He observes that men are aware that they are being watched closely for signs of failure, and they, in turn, have a lack of confidence about their own abilities and low expectations for success in their relationships. 60 component of gender distrust—have lower expectations about marriage. Given the salience of trust in the qualitative interviews, it is likely that this issue goes beyond individual relationships. Results from the FFCW survey indicate that 36 percent of unmarried mothers in Oakland do not believe that men can be trusted to be faithful compared to about 14 percent of fathers. The regression results also show that parents who report distrust have significantly lower expectations about marriage: The odds of mothers’ reporting an almost certain chance of marriage are about 55 percent lower (see Figure 5.1). Distrust is also an important issue for fathers in the FFCW survey. Among fathers who do not trust women to be faithful, the odds of believing there is a certain chance of marriage are about 42 percent lower. In qualitative interviews, parents identified trust as one of the most important things to consider before marrying. Mothers and fathers who did not believe potential partners could be trusted were also more pessimistic about marriage. Jamila generalizes from her experiences with her children’s fathers to explain why she believes men cannot be trusted to be faithful. In part, because of these experiences, she does not want to get married. I ain’t getting married. Cause if you can’t commit, if you gonna cheat on me, if you cheated on your woman, you’ll cheat on me. . . . I’m not just saying with him [the father] . . . just in general. That’s what I’m saying, just in general. Fathers also emphasized the importance of trusting the other partner in a marriage. Like other parents, John views trust as the most important thing to consider before deciding to marry: You gotta be able to trust, you know, the person, you have to trust who you love. That’s the main thing, the trust . . . not to cheat. Not to lie about anything, you know, be open, and you know, I think that’s just the main thing right there. Relationship Conflict Couples that argue often are likely to have lower expectations about marriage, in part because they believe that their relationships are likely to dissolve and that they are incompatible as a couple. Married couples often experience a great deal of stress in their relationships after the birth 61 of a child (Cowan and Cowan, 1997). It is likely that new unmarried parents with limited economic resources would experience a particularly high level of stress and conflict, because they tend to be in less-secure interpersonal, social, and economic situations. In the regression equations, a scale is used to measure whether parents report a high level of conflict in their relationship. This scale is constructed from a series of six questions asking if couples argue often about money, spending time together, sex, the pregnancy, drinking or drug use, and being faithful—issues often cited as sources of conflict by parents in qualitative interviews.13 Oakland mothers have a mean score of .11 and fathers have a mean score of .07 on a scale from 0 to 1.14 The regression results indicate that the level of conflict in parents’ relationships does influence their expectations a great deal. Among unmarried mothers, the odds of reporting an almost certain chance of marriage decrease by about 88 percent for every unit increase in the conflict scale (see Figure 5.1). This means that as mothers’ scores on the scale increase, mothers’ expectations decrease. Among fathers, the odds ____________ 13Couples who said that they argue often about these issues were coded 1 on each variable; those who argue sometimes or never were coded 0. A scale was then created from these variables. A scale is the sum of the scores of the individual variables. In generating the scale, a score was created for every observation for which there was a response to at least one variable. The summative score was divided by the six variables over which the sum was calculated (StataCorp, 1997). The new scale variables that were generated have scores between 0 and 1. For this scale, the alpha is .60 for all mothers and .57 for all fathers in the FFCW survey. Cronbach’s alpha measures the reliability of a scale. Specifically, it measures how well the variables measure a single, unidimensional latent construct. As the average interitem correlation increases, the alpha will also increase. The alpha is also affected by the number of variables in the scale. (See the UCLA Academic Technology Services website at www.ats.ucla.edu/stat/spss/faq/alpha. html). 14Although this scale is intended to indicate an overall level of conflict, parents reported the most frequent arguments about money, spending time together, and being faithful to each other (results not shown). Examining arguments about money is another way to look at how parents’ economic circumstances affect their expectations. When frequent arguments about money were used as an independent variable (rather than as part of a scale), logistic regression results indicate that these arguments significantly reduce the odds of mothers’ reporting an almost certain chance of marriage. Among fathers, arguments about money are significantly related to diminished expectations about marriage, decreasing the odds of reporting an almost certain chance and increasing the odds of reporting little or no chance of marriage. 62 of reporting an almost certain chance of marriage decrease by about 87 percent for every unit change in the conflict scale. In the qualitative interviews, we saw that parents also expressed hesitancy about marriage when they argued frequently. According to Patricia, the frequency with which she and the father argued, and their inability to resolve these conflicts, caused her to re-think the possibility of marriage. We talked about getting married, maybe a year ago. A year ago marrying him—I probably would have. Now —I’m—I’m more mm. No. Not now. . . . Cause it’s like—we can’t get through little things, and if we’re married—we’re not ready to be married. . . . It’s just the little arguments we have—stuff that we can’t even get through. Like I said, sometimes we don’t even settle arguments. I wouldn’t want to go into a marriage like that. I would want to see things change between us first, and then maybe. You know, we could think about it, or talk about it, or maybe even do it. Just basically little stuff. Just little stuff that add up to be a lot. Violence in Relationships Qualitative research suggests that physical violence in relationships may be associated with the decision not to marry, particularly among women (Waller, 1999; Edin, 2000). To examine relationship-specific experiences, a variable was used to indicate whether the mother or father reported that the other parent has hit or slapped them in an argument. In Oakland, about 8 percent of unmarried mothers and 17 percent of unmarried fathers said that this had happened. Because this variable asks about any single incidence of violence, it does not indicate prevalence. Therefore, this measure may represent varying types of conflict—ranging from a single event to more serious patterns of abuse. The variable can, however, indicate heightened conflict and actual, or potential, danger in parents’ relationships. The results show that reporting that the other parent has hit or slapped them in an argument reduces both unmarried mothers’ and fathers’ expectations about marriage. The odds of reporting an almost certain chance of marriage are almost two-thirds lower for mothers who report an incidence of violence (see Figure 5.1). Among fathers, the odds of reporting an almost certain chance of marriage are about 39 percent 63 lower for fathers who say that the mother has hit or slapped them.15 Jackie describes her relationship with her baby’s father before they had a violent argument and how their relationship has changed since this time: We had been together for so many years, and then we had talked about getting married, talked about starting a family—well, we had already started a family. It meant a lot to me, [to be] together. . . . When he first hit me, he write me a letter saying he was sorry, he never meant for it to happen. But when he called me, he said he did it on purpose. He know what he did. He’d do it again if he has to. He’ll come up to my job to make sure I lose my job. . . . He threatened to take my kids . . . and make sure I never see them again. So that’s exactly why I just keep my distance. I had a restraining order on him . . . but I dropped it so he could see his kids. And his mother said she would come over with him to make sure nothing happened. Drug and Alcohol Problems Qualitative results in Chapter 4 suggest that having a partner with drug or alcohol problems affects the other parents’ expectations for the future. These problems may, for example, prevent the other parent from entering the workforce or sustaining relationships. Therefore, a variable was included in these models to measure whether the other parent has a problem with alcohol or drugs that interfered with work or personal relationships. A missing variable was also created for parents who did not respond to this question. About 6 percent of unmarried mothers and fathers in Oakland said that their baby’s other parent had drug or alcohol problems. Another 3 percent of Oakland mothers did not respond to this question. Most of these mothers said that they did not know whether the father had a problem. Like mothers who did not report employment information for fathers, some of these mothers are likely to be out of contact with the fathers; others may not have wanted to report this sensitive information.16 For mothers whose partners have drug or alcohol ____________ 15Unlike measures of conflict and distrust which predict high and low expectations among mothers and fathers, this measure has a more consistent effect in the models for mothers than fathers. Specifically, experiencing violence is significantly associated with mothers’ but not fathers’ reporting little or no chance of marriage. 16The correlation between not reporting employment information and not reporting drug/alcohol problems is about .26 for unmarried mothers in Oakland. 64 problem, the odds of reporting an almost certain chance of marriage are about 69 percent lower (see Figure 5.1).17 Among unmarried fathers, the odds of reporting an almost certain chance of marriage are about 38 percent lower if their child’s mother has a drug or alcohol problem.18 In the qualitative interviews, Kelly illustrates the connection between marriage expectations and fathers’ drug use. Kelly explains, “I know that’s who I’m going to marry. I just, I want him to get off drugs, and get a more stable job.” But when I asked Kelly where she saw their relationship going, she replied: If it keeps going the way it is going? Absolutely nowhere. If . . . I actually get the strength to do what I know I need to do and am supposed to do, I see our relationship as going into marriage, and being a long-term relationship. Her plan is to get her own apartment and prevent her boyfriend from living there if he continues to use drugs. She continues: His Reverend wants to marry us and whatnot. We have to get a license. But things are really rocky at this point. . . . The drugs and whatnot. I just, I totally want them out of my life and all those people out of my life. Race, Ethnicity, and Immigration In the general population, we know that blacks are less likely to marry than whites or Hispanics (see Tucker and Mitchell-Kernan, 1995). In this analysis, the expectations of unmarried white and Hispanic parents are compared to those of unmarried black parents. Whether parents were born in the United States is also included in the analysis, as the majority of Hispanics in the Oakland sample are immigrants.19 As we saw in Chapter 2, about 58 percent of unmarried mothers and 60 percent of unmarried fathers in the Oakland sample are black, about 3 ____________ 17Not reporting fathers’ drug or alcohol use is significantly associated with mothers’ indicating little or no chance of marriage, and less strongly to reducing certainty about marriage. 18Mothers’ having drug or alcohol problems is not significantly related to fathers’ reporting little or no chance of marriage. Like reporting the mother has hit or slapped them, the effect of this variable is also less consistent for fathers than mothers. 19Qualitative findings in Chapter 4 indicated that some couples’ relationships broke up in the first year because the father was deported. Although we would like to look at the effect of immigrants’ being documented, this question was not asked in the survey. 65 percent of mothers and 1 percent of fathers are white, and about 33 percent of mothers and 34 percent of fathers are Hispanic. Immigrants represent about 31 percent of mothers and 33 percent of fathers in the Oakland sample of unmarried parents. The findings indicate that the odds of reporting a high chance of marriage are about two times higher among white mothers and 1.5 times higher among Hispanic mothers than among black mothers. Being born in another country does not have a significant effect on mothers’ expectations. In models for fathers, we see some similar patterns. White and Hispanic fathers are significantly more likely than black fathers to report an almost certain chance of marriage. Unlike Hispanic parents, however, white mothers and fathers are not significantly less likely than black parents to report little or no chance of marriage.20 Again, fathers’ immigration status has no effect. Because parents did not discuss their race or ethnic status in relation to their expectations in the interviews, I do not draw on qualitative data to interpret these findings. Age and Other Children Demographic variables are included for parents’ age and having other biological children (in addition to the child that was just born). Mothers and fathers over age 20 are compared to parents age 20 and under. As we saw in Chapter 2, about 26 percent of Oakland mothers and 12 percent of Oakland fathers are age 20 or younger, about 34 percent of mothers and 32 percent of fathers are age 21–25, about 24 percent of mothers and 29 percent of fathers are age 26–30, and about 16 percent of mothers and 28 percent of fathers are over age 30. About 69 percent of Oakland mothers and 57 percent of fathers have another biological child. We might expect parents with additional children to have more motivation to form a stable partnership with the other parent for themselves and their children.21 However, the results show that age ____________ 20In fact, white mothers report significantly lower expectations about marriage in the little or no chance model after cohabitation is controlled for. White parents’ expectations appear to be divided, perhaps because more white parents report high expectations and fewer report mid-range expectations. 21Parents who have additional children with the baby’s other parent may already have a marriage-like relationship and, therefore, have higher expectations for marriage. 66 and having other children are not associated with not reporting an almost certain chance of marriage among mothers or fathers.22 Cohabitation In Oakland, about 49 percent of unmarried mothers and 60 percent of unmarried fathers were living with the other parent at the time of their child’s birth. Earlier chapters show the importance of cohabitation to the stability of unmarried parents’ relationships in the year after their child was born. Previous research suggests that cohabiting couples in the general population typically expect to marry each other and that couples who hold these expectations are more likely to marry than others (Bumpass, Sweet, and Cherlin 1991; Brown, 2000).23 Although the results of this chapter are presented for a model without cohabitation, this variable was added to a second model because it was expected to strongly influence parents’ expectations. For example, cohabitation may indicate a relatively high commitment to the relationship or the ability to set up an independent household. Perhaps not surprisingly, the second model shows that cohabitation is the strongest predictor of parents’ reporting high expectations about ________________________________________________________ Although it is important to understand the effect of having additional children with the baby’s parent rather than another parent, this question was asked only in the 12-month interview. To date, information is available only for the two cities that have completed follow-up surveys (i.e., Oakland and Austin). The follow-up data for Oakland indicate that about 50 percent of mothers had additional children with only the baby’s father or with the baby’s father and another partner, and 50 percent had additional children with another father only. About 49 percent of fathers had additional children only with the baby’s mother or with the mother and another mother, and about 51 percent had children with another mother only. Regression results from the Oakland and Austin sample indicate that having additional children with the baby’s other parent is positively related to parents’ marriage expectations. 22Mothers age 26–30 report significantly lower chances of marriage and those who have other children report significantly higher chances of marriage in the little or no chance model. 23Researchers have found that cohabiting black couples are just as likely as cohabiting white couples to expect to marry. About 70 percent of white and black cohabitors and about 64 percent of Mexican-American cohabitors expect to marry (Bumpass, Sweet, and Cherlin, 1991). However, white cohabitors are more likely than black cohabitors to actually make the transition (Manning and Smock, 1995; Brown, 2000). 67 marriage (see Appendix Tables B.3–B.6). Cohabiting increases mothers’ odds of reporting an almost certain chance of marriage over fourfold and fathers’ odds almost threefold. 24 However, cohabitation is itself an outcome and should be determined by many of the same factors that affect marriage expectations. Therefore, adding cohabitation to the model dilutes the effect of factors shown to be important determinants of parents’ marriage expectations. For example, Hispanic parents are more likely to be cohabiting, and cohabitors have higher expectations about marriage than other parents. As a result, the effect of being Hispanic is no longer significant after controlling for cohabitation status.25 Similarly, employed fathers are significantly less likely to report low expectations for marriage before but not after cohabitation is included in the model. This suggests a selection into cohabitation among employed fathers. Unmarried parents may decide to live together to test whether their relationship would be successful as a marriage. Sandra explains this decision: I think we wanted to move in together kind of to test it out, because we knew we wanted to be together, you understand what I’m saying? But at the same time, I don’t think we were ready for marriage. . . . It [moving in together] was serious. It’s still serious, but as a test. Other parents did not move in together with the intention of marrying. However, their expectations for marriage were heightened after they established a home together. One mother said that she and the father did not make an explicit decision to live together but that this happened as he began spending the night on a regular basis. ____________ 24Because information about length of cohabitation before the birth was asked only in the follow-up survey, these data are available in only two cities. The follow-up data for Oakland indicate that parents who were cohabiting at birth had been living together on average for about 1.4 years. Regression results from the Oakland and Austin sample indicate that parents who had lived together longer had higher expectations about marriage. 25Looking at both models for high and low expectations, we see that the effects of fathers’ having missing information about a drug or alcohol problem, mothers’ household income, and mothers’ having another biological child also lose significance after controlling for cohabitation. In addition, the effect of mothers’ having a drug or alcohol problem loses significance in models for fathers. 68 Like a couple nights he had spent the night, and I was getting ready for work, and I was like “You don’t have to leave, you can stay here or whatever.” . . . And then after a couple of months . . . I said, “You can spend the night, go get you a pair of clothes and you can spend the night.” His clothes started building up, building up, building up. And I said “Shoot, you might as well move in.” Kind of joking, you know, joking about it. And then he did move in. But it was like, I don’t even know how it came about. But I know he just started spending the night, spending the night. . . . It was like he gradually moved in. It wasn’t like: “Okay, go get your stuff, I want you to move in.” It wasn’t like that. Since living together, she says they have begun to think of their relationship as a marriage: “It’s more like husband and wife instead of boyfriend and girlfriend. That’s how it feels.” Several parents in Oakland noted the difficulty of marrying and raising children together when they could not afford an apartment together. About 45 percent of unmarried parents in the FFCW survey had moved since the birth of their child. On average, parents moved 1.6 times between the birth and first-year follow-up, with the number of moves ranging from one to seven. For some parents, being able to afford a place together was a prerequisite for marriage.26 The results suggest that effect of fathers’ employment may work in part through allowing parents to move in together. Mark explains that he could not afford an apartment for him and his baby’s mother after losing his job. He feels it is critical to have a place to live together before marrying: “I feel like if you take a woman’s hand in marriage, you should at least have a place for that woman to go stay. . . . We got a kid involved in this. I want to be able to take care of them.” Conclusion Consistent with the qualitative findings, the results from the logistic regression models show that unmarried parents face multiple barriers in their relationships. Socioeconomic factors, distrust between men and ____________ 26We would expect the number of times parents moved in the previous year—a measure of housing stability—to be negatively related to parents’ outlooks on marriage. However, this question was asked at the 12-month interview rather than at the time of the birth. Therefore, the regression analysis cannot look at the effect of this variable on expectations at birth. 69 women, relationship conflict and violence, and drug and alcohol problems all play a role in explaining parents’ expectations about marriage. The results show that fathers’ employment is positively related to mothers’ reporting high chances of marriage. Mothers with higher education—an indicator of employment potential—express more certainty about marriage than other mothers. Compared to the qualitative findings, the effects of economic variables in the statistical analysis are not as pronounced. This difference could result because unmarried parents in this survey show little variation in their economic and educational characteristics—that is, most are poor and have not attained high levels of education. The statistical analysis, however, is intended to examine how variations in parents’ economic status affect outcomes. As in the qualitative findings, these results show strong negative relationships between parents’ expecting to marry and certain other factors. Parents who believe that the other gender cannot be trusted to be faithful, who report a high level of conflict in their relationship, who report that the other partner has hit them, and have partners with drug or alcohol problems have lower expectations about marriage. In the regression models, these effects are more consistent than the effects of economic factors. Taken together, these economic and non-economic factors explain more of the variation in mothers’ expectations than fathers’ expectations about marriage. When cohabitation is added to the model, it is the strongest predictor of parents’ expectations about marriage. Living with the other parent consistently increases the odds that mothers and fathers will report higher chances of marrying the other parent. Because some of the same determinants of cohabitation also underlie parents’ expectations about marriage, adding cohabitation to the model also reduces the effects of variables. These findings suggest that unmarried parents perceive cohabitation to be a step toward marriage, whether or not this transition eventually occurs. Men and women who expect to marry or want to test whether their relationship is right for marriage may move in together before the birth of their child. The experience of living with the other parent may also increase parents’ expectations that they will marry. 70 6. Conclusion and Policy Implications These findings provide an early look at the stability of relationships between unmarried parents in the first year of their child’s life. They show that about half of unmarried parents in Oakland were living together when their children were born, suggesting a relatively high level of commitment in their relationships. One year later, we observed the following outcomes: • Parents were divided between those living in formal or informal unions and those who were no longer involved in romantic relationships. • Only 7 percent of unmarried parents had married, despite high expectations about marriage at the time of their child’s birth. • Relationships between parents who were cohabiting were relatively stable. About 75 percent of these parents were still living together, 15 percent had broken up or stopped living together, and 10 percent had married. • Relationships between parents who were romantically involved but not cohabiting at the time of their child’s birth experienced considerable dissolution. Over half of these relationships had ended, and although over one-quarter of these parents moved in together, few married. • With one exception, couples who were not romantically involved at the time of their child’s birth did not reunite. Taken together, these results suggest at least two conclusions. First, policies and programs to encourage marriage face serious challenges. Second, efforts to strengthen two-parent families should probably target cohabiting parents. Although these couples were not inclined to marry in the first year, their relationships seem to have the best chance of 71 providing stable families and making the transition to marriage. Programs that help couples still involved in romantic relationships overcome common obstacles may be effective in promoting relationship stability. Policymakers should also be especially careful not to implement programs that unintentionally interfere with family stability or discourage paternal involvement (National Conference of State Legislators, 2000a; Sorensen, Mincy, and Halpern, 2000).1 There seems to be little chance that parents who are not in romantic relationships with one another will marry or even reunite. However, programs could help noncustodial parents (typically fathers) remain emotionally and economically involved in their children’s lives and help unmarried parents raise their children cooperatively in the absence of a romantic relationship. These efforts are particularly important because fathers’ involvement with their children often declines sharply after the romantic relationship with the mother ends (Mincy and Pouncy, 1999; Primus and Daugirdas, 2000; Sorensen, Mincy, and Halpern, 2000).2 Fatherhood programs could also help unmarried fathers who are involved in romantic relationships with their child’s mother gain access to jobs. Under TANF, states have been given more flexibility to support programs for fathers that promote paternal support and involvement (National Conference of State Legislators, 2000b). Financial and Housing Stability Compared to the effects of personal and relationship issues in the statistical analysis, the effects of economic variables were moderate. However, the qualitative interviews show that unmarried parents were experiencing serious financial problems that made them feel less secure about the future of their relationships and more hesitant about marriage. There is also some evidence from the study that these economic effects ____________ 1See Waller and Plotnick (1999) for a discussion of the mismatch between child support policies and the situations of low-income parents. See Sorensen, Mincy, and Halpern (2000) for an explanation of how some cash and in-kind assistance programs may be disadvantageous to two-parent families. 2The issue of paternal support and involvement will be the subject of a future report on the FFCW study in Oakland. 72 work, in part, by impeding cohabitation, which has the strongest positive effect on expectations. Some state and community programs provide coordinated services to promote employment and increase earnings among low-income parents, which may assist fragile families in overcoming economic insecurity.3 Although primarily directed toward low-income custodial mothers, some county- and community-based programs in California and elsewhere provide employment services to noncustodial parents of low-income children (California Department of Social Services, 1999).4 Cash assistance programs (such as TANF) and in-kind assistance (food stamps, housing, and health coverage) for families headed by unmarried parents has also been directed primarily toward single custodial parents. To support two-parent families, new proposals call for further eliminating distinctions between single-parent and two-parent families in receiving cash assistance and broadening eligibility criteria for in-kind assistance to ____________ 3For example, states offer educational and training opportunities to allow parents to obtain better-paying jobs and to provide transitional jobs to parents who lack job skills. Child care, health care, and transportation are important supportive services that allow parents to enter the workforce and retain jobs. Improving access to Medi-Cal, Healthy Families (i.e., the state Child Health Insurance Program), food stamps, and child care for non-TANF families are regarded as important strategies to help low-income families cover their expenses while working at low-paying jobs. In addition, programs such as a state Earned Income Tax Credit and increases in the state minimum wage are intended to boost the incomes of low-income families and reward work. Finally, state efforts to improve education (especially through lowering dropout rates and increasing access to education beyond high school) could also assist fragile families in bettering their economic situations over the long run. Many of these state strategies are described at greater length in Lazere, Fremstad, and Goldberg (2000). 4The goal of helping noncustodial parents obtain employment is consistent with the purpose of promoting marriage in PRWORA, according to the Department of Health and Human Service’s guide to states: “Under this purpose, a state could help any needy parent, including a non-custodial parent or a working parent, by providing employment, job preparation, or training services. Examples of potential services include job or career advancement activities, marriage counseling, refundable earned income tax credits, child care services, and employment services designed to increase the non-custodial parent’s ability to pay child support. Activities that promote any one of the three objectives—job preparation, work, and marriage—would be consistent with this purpose” (U.S. Department of Health and Human Services, 1999, p. 11). 73 include all poor parents and their children, including noncustodial parents (Sorensen, Mincy, and Halpern, 2000).5 In the qualitative interviews, parents emphasized housing problems. These problems may be more pronounced in the Bay Area, where rents are particularly high and housing is limited, than they are in other locations. Programs that increase parents’ employment and earnings would also allow them to better afford housing. Although the statistical analysis did not show a connection between mothers’ receiving housing subsidies and parents’ expectations about marriage, there is also some evidence from welfare reform experiments that government housing subsidies have contributed to positive employment and earnings outcomes in California and elsewhere.6 In particular, housing subsides may help families stabilize their lives, enabling them to find and retain employment (Sard and Lubell, 2000). However, if parents do not cohabit or do not report that they are cohabiting because they fear losing eligibility for housing, these subsidies are unlikely to stabilize fragile families. Distrust and Conflict in Relationships Parents reported conflicts around trust, infidelity, and commitment that contributed to the dissolution of their relationships in the first year. They also expressed concerns about conflict and trust when discussing reasons for delaying marriage. In the statistical analysis, the negative effects of these variables were more consistent than those associated with economic factors, but the qualitative analysis suggests that these issues may be especially pronounced when parents are living in conditions of poverty. Compared to economic problems, conflict and distrust between parents may be less amenable to policy intervention. However, some ____________ 5See Knox, Miller, and Gennetian (2000) for a discussion of findings from the Minnesota Family Investment Program evaluation. Among other things, the Minnesota Family Investment Program had less-restrictive eligibility rules for two-parent families. The evaluation of the overall program shows small but positive effects on increasing marriage rates and larger effects on increasing marital stability and reducing domestic violence among recipients of Aid to Families with Dependent Children. 6Most housing assistance, including public housing, publicly assisted housing, and tenant-based assistance (such as Section 8 vouchers) is funded at the federal level and allocated to localities (Kingsley, 1997). 74 community-based fatherhood programs (and programs for noncustodial parents) have addressed these issues directly. Although these programs are designed primarily to increase parents’ ability to secure a job and meet their child support obligations, they also provide services to increase parents’ skills in building or maintaining family relationships. For example, some programs help unmarried mothers and fathers develop a “team parenting” plan for raising their children cooperatively. Programs also offer peer support sessions for fathers that help them better communicate with the mother and address issues of distrust between men and women more generally. TANF funds can be used for relationship skills training for noncustodial parents as well as for employment services. California’s plan for providing assistance under PRWORA states its intention to do so.7 Although fatherhood programs offer promising ways to help parents develop and maintain family relationships, they are limited in number. Furthermore, they tend to serve noncustodial fathers with child support obligations, who are more likely than other fathers to have ended their romantic relationship with their child’s mother. To be effective on a larger scale, more programs may be needed to serve unmarried parents who are still in romantic relationships. Drugs, Alcohol, and Violence Reports of drug and alcohol abuse were negatively related to relationship outcomes and marriage expectations. These effects were ____________ 7The Department of Health and Human Services’ guide on funding TANF services suggests that states could address PRWORA’s goal of encouraging two-parent family formation and maintenance by providing services such as parenting skills training, premarital and marriage counseling, mediation services, activities to promote parental access and visitation, job placement and training services for noncustodial parents, initiatives to promote responsible fatherhood and increase the capacity of fathers to provide emotional and financial support for their children, and crisis or intervention services (U.S. Department of Health and Human Services, 1999). According to the California state plan (California Department of Social Services, 1999), “Services will be provided to unemployed or underemployed noncustodial parents whose children are receiving public assistance such as CalWORKs, Food Stamps, Medi-Cal, Healthy Families or Child Support Assurance benefits. Services may include employment training, job finding skills, family mediation services, parental skill building, supportive services and applicable child support program services.” 75 more consistent for mothers than for fathers, suggesting that these issues may be a particular concern for women. As with conflict and distrust, these problems tend to coincide with economic problems. Currently, some state programs are available to help welfare recipients deal with substance abuse issues that may interfere with their ability to work or maintain employment. TANF money can also be used to fund programs for fathers that provide substance abuse services (National Conference of State Legislators, 2000b). Helping parents overcome these problems may allow them to stabilize their relationships and families. However, it would be important to proceed cautiously in encouraging parents with serious drug and alcohol problems to marry or become more involved with their families. Similarly, some community and state programs are available to help victims of violence repair their lives and support their families. Because domestic violence is considered a barrier to workforce participation, some TANF recipients now have access to domestic violence services. Other programs (such as child support enforcement) can also help ensure the safety of custodial parents and their children (National Women’s Law Center and the Center on Fathers, Families, and Public Policy, 2000). Moreover, TANF funding can also be used for fatherhood programs that address issues of domestic violence (National Conference of State Legislators, 2000b).8 However, encouraging marriage or increased involvement in relationships characterized by domestic violence would be dangerous. By knowing more about the circumstances of unmarried parents, policymakers can better discern when and how to support these fragile families. This report has focused on a critical time in family formation, but future results will deepen our understanding of the challenges these families face and guide program development and implementation. ____________ 8Proposed legislation would also encourage fatherhood programs and campaigns to address these issues. For example, the Child Support Distribution Act of 2001 (U.S. 107th Congress, 2001a) would require “a written commitment by the entity that the entity will make available to each individual participating in the project education about the causes of domestic violence and child abuse and local programs to prevent and treat abuse, education about alcohol, tobacco, and other drugs and the effects of abusing such substances.” 76 Appendix A The Fragile Families and Child Wellbeing Study This report presents information from Oakland, California—the first city surveyed in the FFCW study. The study follows families living in 20 cities from the birth of their child through age four. New mothers were interviewed at the hospital after giving birth. Fathers were interviewed either at the hospital or elsewhere as soon as possible after the birth. Follow-up interviews are being conducted 12, 30, and 48 months later; the final follow-up interview will include an in-home assessment of the child’s school readiness at age four. The data are representative of nonmarital births to parents residing in cities with populations over 200,000. The data are also representative of nonmarital births in each of the cities in the study. The total sample of 4,700 families includes 3,600 births to unmarried parents and 1,100 births to married parents. The study addresses four sets of related questions: • What are the conditions and capabilities of new unmarried parents, especially fathers? How many of these men hold steady jobs? How many want to be involved in raising their children? • What is the nature of the relationships between unmarried parents? How many of these couples are involved in stable relationships? What proportion expects to marry? What proportion is exposed to high levels of conflict or domestic violence? • What factors push new unmarried parents together? What factors pull them apart? In particular, how do public policies affect parents’ behaviors and living arrangements? • What are the long-term consequences for parents, children, and society of new welfare regulations, stronger paternity establishment and stricter child support enforcement, and changes in health care and child care financing and delivery? 77 This report focuses on a subsection of questions in the baseline and follow-up survey. These questions measure parents’ relationship status at baseline and at year one as well as their expectations about marriage and the factors that predict expectations. The text of these survey questions appears below. Survey Questions Marriage expectations M1b14 and F1b14. What do you think the chances are that you will marry [baby’s father]/[baby’s mother] in the future? No chance ........................................1 A little chance .................................. 2 A 50-50 chance ................................3 A pretty good chance ........................4 An almost certain chance .................. 5 Relationship (at birth) M1b3. Which of the following statements best describe your current relationship with [baby’s father]? We are romantically involved on a steady basis .......................1 We are involved in an on-again and off-again relationship......2 We are just friends..................................................................3 We hardly ever talk to each other ...........................................4 We never talk to each other ....................................................5 M1b8. Are you and the [baby’s father] living together now? Yes .....................................................1 No .....................................................2 Relationship (12-month follow-up) M2a7. What is your relationship with the father now? Married .............................................1 Romantically involved .......................2 Separated/divorced ............................3 Just friends ........................................4 Not in any kind of relationship .........5 M2a7a. Are you and the father currently living together? All or most of the time ......................1 Some of the time ...............................2 Rarely or never ................................. .0 78 Employment (at birth) M1j1a. In the last year did you have income from earnings? Yes..................................................... 1 No..................................................... 2 M1i6. What was [baby’s father] doing most of last week (working, going to school, else)? Working at a regular job .................... 1 Looking for work ............................... 2 Going to school ................................. 3 Unable to work .................................. 4 Other ................................................ 5 F1j1. Last week, did you do any regular work for pay? Yes..................................................... 1 No..................................................... 2 Drugs and alcohol M1g5. In the past year, has drinking or using drugs ever interfered with your work on a job or with your personal relationships? Yes..................................................... 1 No..................................................... 2 M1i9. Does [baby’s father] have problems such as keeping a job or getting along with family and friends because of alcohol or drug use? Yes..................................................... 1 No..................................................... 2 Don’t know ....................................... 3 Physical violence M1b7b & M1b13b and F1b7b & F1b13b. Thinking about your past relationship with [baby’s father]/[baby’s mother], how often would you say that: Often Sometimes Never He/she hit or slapped you when he/she was angry?......................................1............2.............3 Conflict in the relationship M1b6a-f & M1b12a-f and F1b6a-f & F1b12a-f. The following is a list of subjects on which couples often have disagreements. How often, if at all, in the last month have you and [baby’s father] had open disagreements about each of the following: 79 Often Sometimes Never Money? Would you say often, sometimes, never?...............................1..........2...............3 Spending time together?........................1..........2...............3 Sex?.......................................................1..........2...............3 The pregnancy?.....................................1..........2...............3 Drinking or drug use?............................1..........2...............3 Being unfaithful?...................................1..........2...............3 Gender distrust M1d3e and F1d3e. Please tell me whether you strongly agree, disagree, or strongly disagree with the following statement: Strongly Strongly Don’t Agree Agree Disagree Disagree Know Men/Women cannot be trusted to be faithful..............................4.............3..........2..............1...........98 Education M1i1 and F1i1. Now I’d like to ask some questions about your education and work experience. What is the highest grade or year of regular school that you have completed? No formal schooling.................................................... 1 8th grade or less........................................................... 2 Some high school (grades 9, 10, 11, and 12) ............... 3 High school diploma (completed 12th grade) .............. 4 GED ........................................................................... 5 Some college or two-year degree .................................. 6 Technical or trade school ............................................ 7 Bachelor’s degree ......................................................... 8 Graduate professional school ....................................... 9 Welfare Use M1j1. In the last 12 months, did you have income from public assistance, welfare, or food stamps? Yes .....................................................1 No .....................................................2 80 Public/Subsidized Housing M1f3. Do you live in a public housing project? Yes.................................................... 1 No.................................................... 2 M1f4. Is the federal, state, or local government helping to pay for your rent? Yes.................................................... 1 No.................................................... 2 Household Income M1j3. Thinking about your income and the income of everyone else who lives with you, what was your total household income before taxes in the past 12 months? Under $5,000 ....................................1 $5,000 to $9,999 ...............................2 $10,000 to $14,999 ...........................3 $15,000 to $19,999 ...........................4 $20,000 to $24,999 ...........................5 $25,000 to $34,999 ...........................6 $35,000 to $49,999 ...........................7 $50,000 to $74,999 ...........................8 Greater than $75,000 ........................9 Refused ............................................97 Don’t know ..................................... 98 Age M1h1 and F1h1. What is your date of birth? (month, day, year) Race/ethnicity M1h3 and F1h3. Which of these categories best describes your race. White ...............................................1 Black, African-American ...................2 Asian or Pacific Islander ...................3 Native American, Eskimo, Aleut.......4 Other ...............................................5 Don’t know ....................................98 M1h3a and F1h3a. Are you of Hispanic or Latino origin or descent? Yes.................................................... 1 No.................................................... 2 Don’t know .................................... 98 81 Country of origin M1h2 and F1h2. Were you born in the United States? Yes .....................................................1 No .....................................................2 Other children M1a12 and F1a6. Do you have any other biological children? Yes .....................................................1 No .....................................................2 Between the interview at birth and at the 12-month follow-up, the Oakland sample lost 15 percent of unmarried mothers and 29 percent of unmarried fathers. Table A.1 shows the percentage of parents retained and lost by selected demographic and socioeconomic characteristics. Reviewed below are some notable changes in characteristics of the sample between the two interview periods. About 16 percent of mothers age 21–25 and 19 percent of mothers age 26–30 who participated at baseline were not interviewed at the 12month follow-up, compared to only 12 percent of mothers age 20 and under and 8 percent of mothers age 30 and over. In addition, the study lost more mothers with less than a high school diploma (16 percent) and mothers with a high school diploma or GED (19 percent) than mothers at higher educational levels. It is important to note that 19 percent of mothers who were romantically involved with the father at the time of birth but not living with him did not participate in the follow-up interview. Among unmarried fathers in the study who were interviewed at baseline but not at 12 months, more fathers age 20 or under (30 percent) and age 26–30 (35 percent) did not participate in the follow-up than fathers in other age groups. Between the baseline and the follow-up, 32 percent of black fathers had dropped out of the study compared to 22 percent of Hispanic fathers. Another change to note is that the median household income was higher for fathers who remained in the study than for those who dropped out: $17,500 versus $12,700, respectively. Finally, 40 percent of fathers who were romantically involved with the mother at birth but not cohabiting dropped out of the study, and 31 82 Table A.1 Comparison of Demographic and Socioeconomic Characteristics of Unmarried Parents in Oakland in the Sample at 12-Month Follow-Up and Parents Not in Sample at 12-Month Follow-Up 83 20 and under 21–25 26–30 Over 30 White Hispanic Black Immigrant Another biological child Less than high school diploma High school diploma/GED More than high school diploma College degree Mother had earnings Father employed Mother receives welfare/food stamps Cohabiting Romantic relationship Not in a romantic relationship Median household income, $ Total Mothers Fathers Baseline 12-Month Sample Not in Sample Baseline 12-Month Sample Not in Sample No. of % of No. of % of No. of No. of % of No. of % of No. of Obs. Baseline Obs. Baseline Obs. Obs. Baseline Obs. Baseline Obs. 65 88 57 12 8 23 70 16 30 7 85 84 71 16 14 60 73 44 27 16 59 81 48 19 11 54 65 35 35 19 39 92 36 8 3 52 77 40 23 12 7 71 5 29 2 2 100 2 0 80 86 69 14 11 63 78 49 22 14 140 86 121 14 19 112 68 76 32 36 75 89 67 11 8 62 76 47 24 15 168 85 143 15 25 105 71 75 29 30 127 84 107 16 20 88 72 63 28 25 77 81 62 19 15 63 68 43 32 20 38 97 37 3 1 35 77 27 23 8 4 100 4 — 0 3 67 2 33 1 85 85 72 15 13 — — — —— — — — — — 125 75 94 25 31 130 87 113 13 17 — — — —— 121 88 106 12 15 113 78 88 22 25 93 81 75 19 18 63 60 38 40 25 34 91 31 9 3 13 69 9 31 4 17,300 17,300 212 16,400 36 17,500 17,500 135 12,700 54 248 85 212 15 36 189 71 135 29 54 NOTES: All parents were in the baseline sample. Nine unmarried fathers participated in the 12-month follow-up survey but not the baseline survey. These fathers were taken out of the 12-month total when calculating the percentages for characteristics of fathers in the survey at 12-month. percent who reported no romantic relationship with the mother at birth were not interviewed at follow-up. As a result, the follow-up sample of fathers is more selective of fathers in committed relationships than the baseline sample. 84 Appendix B Sample Means and Regression Results This appendix presents the means for the dependent and independent variables used in the logistic regression models (Tables B.1 and B.2). It also shows the results from these logistic regressions as odds ratios. Table B.3 reports the odds that mothers will indicate an almost certain chance of marriage and Table B.4 reports the odds that mothers will indicate little or no chance of marriage. Similarly, Table B.5 shows the odds that fathers will indicate an almost certain chance of marriage and Table B.6 shows the odds that fathers will indicate little or no chance of marriage. In Tables B.3–B.6, Model 1 is run without cohabitation and Model 2 adds cohabitation to the regressions. Table B.1 Means for Dependent Variables in Logistic Regressions Mothers Mother reported an almost certain chance of marriage Oakland .34 20-City .33 Fathers Father reported an almost certain chance of marriage Oakland 20-City .46 .47 Mother reported a Father reported a good or 50/50 good or 50/50 chance of marriage .37 .40 chance of marriage .38 .41 Mother reported a little or no chance of marriage .29 Father reported a little or no chance of .27 marriage .16 .12 85 Table B.2 Means for Independent Variables in Logistic Regressions Mothers Father employed (mother’s report) Oakland .57 Father’s employment missing (mother’s report) .09 Father in school, not employed (mother’s report) .05 Mother had earnings last year .35 Mother has less than high school diploma .52 Mother has high school diploma or GED .31 Mother has more than high school diploma .15 Mother has college degree .02 Mother receives welfare/food stamps .53 Mother lives in public/subsidized housing .24 Mother’s household income (divided by 10,000) 1.96 20-City .67 .11 .02 .67 .41 .34 .23 .03 .44 .21 2.42 Fathers Father employed Father has less than high school diploma Father has high school diploma or GED Father has more than high school diploma Father has college degree Mother had earnings last year Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Women cannot be trusted to be faithful Frequent arguments (scale) Mother hit/slapped father Oakland .67 .47 .33 .19 .02 .35 .53 .26 1.91 .14 .07 .17 20-City .76 .40 .36 .21 .03 .68 .44 .21 2.50 .14 .11 .15 Men cannot be trusted to be faithful .36 Mother has drug/ .24 alcohol problem .06 .03 86 Table B.2 (continued) Mothers Frequent arguments (scale) Father hit/slapped mother Oakland .11 .08 20-City .12 .05 Fathers Black White Hispanic Father has drug/ alcohol problem Father’s drug/ alcohol missing Immigrant .06 .06 Age 20 and under .03 .04 Age 21–25 Black .58 .55 Age 26–30 Oakland 20-City .60 .55 .01 .12 .34 .29 .33 .16 .12 .18 .32 .37 .29 .21 White Hispanic Immigrant Age 20 and under Age 21–25 Age 26–30 Age over 30 Mother has another biological child .03 .33 .31 .26 .34 .24 .16 .69 .14 Age over 30 .28 .28 Father has another biological child .57 .15 Cohabiting with .33 mother (mother’s report) .60 .35 Lives in Oakland — .18 Interaction between .16 living in Oakland and father having more than high .14 school diploma — .23 .56 .58 .07 .01 Cohabiting with father Lives in Oakland Interaction between .49 .48 living in Oakland and father having — .07 another child — .04 Interaction between living in Oakland and father having drug problem — Interaction between living in Oakland .004 and mother receiving welfare — .04 Interaction between living in Oakland and mother having earnings — Interaction between living in Oakland and mother having .02 earnings — .02 87 Table B.2 (continued) Mothers Interaction between living in Oakland and mother’s household income Oakland — 20-City .13 Fathers Interaction between living in Oakland and mother living in public/subsidized housing Oakland — 20-City .02 88 Table B.3 Odds Ratios from Logistic Regressions of Mothers’ Reports of an Almost Certain Chance of Marriage Father employed (mother's report) Father’s employment missing (mother’s report) Father in school, not employed (mother’s report) Mother had earnings last year High school diploma More than high school diloma College degree Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s Household Income (divided by 10,000) Men cannot be trusted to be faithful Frequent arguments (scale) Father hit/slapped mother Father has drug/alcohol problem Father’s drug/alcohol missing White Hispanic 1.379** (.146) .236*** (.058) 1.056 (.288) .941 (.086) 1.248* (.121) 1.389** (.156) 1.432 (.336) .942 (.082) .860 (.092) 1.058** (.019) .454*** (.049) .118*** (.033) .331*** (.098) .311*** (.085) .393* (.156) 2.031*** (.233) 1.449*** (.145) 1.318** (.146) .395*** (.099) 1.289 (.373) .910 (.087) 1.301** (.132) 1.508*** (.179) 1.703* (.422) .985 (.090) .960 (.107) 1.043* (.020) .493*** (.056) .149*** (.043) .329*** (.101) .372*** (.103) .506 (.205) 1.599*** (.194) 1.171 (.123) 89 Table B.3 (continued) Immigrant Model 1 1.210 (.156) Model 2 1.075 (.146) Age 21-25 1.147 (.117) 1.088 (.115) Age 26-30 1.012 (.126) .982 (.129) Age more than 30 .832 (.120) .787 (.119) Mother has another biological child 1.188 (.110) 1.010 (.098) Cohabiting with father — 4.253*** (.377) Lives in Oakland 1.196 (.201) 1.129 (.199) Interaction between living in Oakland and father having drug or alcohol problem 9.554** (7.072) 10.627** (8.269) No. 3,487 3,487 Pseudo R2 0.127 0.191 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of mother living in Oakland and father having drug/alcohol problem. Model 2 controls for the interaction of mother living in Oakland and father having drug/alcohol problem. *p < = .05. **p < =. 01. ***p < = .001. 90 Table B.4 Odds Ratios from Logistic Regressions of Mothers’ Reports of Little or No Chance of Marriage Father employed (mother's report) Father’s employment missing (mother’s report) Father in school, not employed (mother’s report) Mother had earnings last year High school diploma More than high school diploma College degree Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Men cannot be trusted to be faithful Frequent arguments (scale) Father hit/slapped mother Father has drug/alcohol problem Father’s drug/alcohol missing White Model 1 .893 (.100) 7.287*** (1.175) 1.181 (.333) 1.199 (.124) .831 (.088) .862 (.110) 1.171 (.319) 1.131 (.110) 1.121 (.122) .929** (.022) 2.210*** (.219) 4.043*** (.935) 2.976*** (.563) 2.548*** (.437) 3.310*** (.893) 1.010 (.137) Model 2 .971 (.117) 4.637*** (.796) .935 (.277) 1.132 (.121) .788* (.090) .772 (.106) .984 (.294) 1.077 (.113) .981 (.114) .960 (.024) 2.009*** (.216) 2.926*** (.730) 3.258*** (.698) 2.366*** (.439) 2.691*** (.761) 1.555** (.238) 91 Table B.4 (continued) Hispanic Immigrant Age 21–25 Age 26–30 Age more than 30 Mother has another biological child Cohabiting with father Lives in Oakland Interaction between living in Oakland and mother having earnings Interaction between living in Oakland and mother’s household income No. Model 1 .782* (.091) .969 (.143) 1.004 (.115) 1.674*** (.229) 1.289 (.203) .707*** (.074) — .689 (.200) .407* (.165) 1.302** (.130) 3,487 Model 2 1.026 (.131) 1.201 (.198) 1.084 (.134) 1.753*** (.263) 1.334 (.228) .890 (.102) .100*** (.012) .618 (.181) — 1.208 (.124) 3,487 Pseudo R2 0.176 0.294 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of mother living in Oakland and having earnings and mother living in Oakland and household income. Model 2 controls for the interaction of mother living in Oakland and household income. *p < = .05. **p < = .01. ***p < = .001. 92 Table B.5 Odds Ratios from Logistic Regressions of Fathers’ Reports of an Almost Certain Chance of Marriage Father employed Father has high school diploma or GED Father has more than high school diploma Father has college degree Mother had earnings Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Women cannot be trusted to be faithful Frequent arguments (scale) Mother hit/slapped father Mother has drug/alcohol problem White Hispanic Immigrant Age 21–25 Age 26–30 Age over 30 Model 1 1.161 (.122) 1.041 (.105) 1.195 (.147) .831 (.204) 1.050 (.100) .842 (.076) .837 (.093) 1.022 (.020) .581*** (.075) .125*** (.035) .610*** (.076) .615* (.155) 2.448*** (.345) 1.438*** (.154) .959 (.127) 1.210 (.150) 1.241 (.178) 1.276 (.186) Model 2 1.121 (.122) 1.071 (.111) 1.272* (.158) .924 (.235) 1.024 (.100) .877 (.084) .905 (.104) 1.013 (.020) .590*** (.078) .161*** (.046) .601*** (.076) .684 (.176) 1.914*** (.280) 1.202 (.133) .874 (.119) 1.124 (.143) 1.129 (.166) 1.146 (.171) 93 Table B.5 (continued) Father has another biological child Model 1 .982 (.094) Model 2 .898 (.088) Cohabiting with mother (mother’s report) — 2.781*** (.254) Lives in Oakland .729 (.202) 1.074 (.339) Interaction between living in Oakland and having more than high school diploma 2.374 (1.161) — Interaction between living in Oakland and having another child 1.692 (.592) 1.939 (.714) Interaction between living in Oakland and mother — .495* receiving welfare (.181) No. 2,543 2,543 Pseudo R2 0.069 0.106 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of father living in Oakland and having more than a high school diploma and father living in Oakland and having another child. Model 2 controls for the interaction of father living in Oakland and having another child and father living in Oakland and mother receiving welfare. *p < = .05. ***p < = .001. 94 Table B.6 Odds Ratios from Logistic Regressions of Fathers’ Reports of Little or No Chance of Marriage Father employed Father has high school diploma or GED Father has more than high school diploma Father has college degree Mother had earnings Mother receives welfare/food stamps Mother lives in public/subsidized housing Mother’s household income (divided by 10,000) Women cannot be trusted to be faithful Frequent arguments (scale) Mother hit/slapped father Mother has drug/alcohol problem White Hispanic Immigrant Age 21–25 Age 26–30 Age over 30 Model 1 .721* (.106) .747 (.116) .972 (.177) 1.549 (.517) 1.369* (.212) .965 (.133) .987 (.171) .981 (.031) 1.735*** (.284) 13.077*** (4.167) 1.212 (.198) .846 (.319) .713 (.163) .706* (.122) 1.262 (.261) .831 (.155) .927 (.198) 1.011 (.217) Model 2 .785 (.120) .740 (.119) .996 (.187) 1.280 (.448) 1.358* (.216) .933 (.133) .999 (.168) .995 (.033) 1.601** (.273) 9.842*** (3.273) 1.259 (.212) .642 (.252) 1.109 (.270) .954 (.172) 1.476 (.320) .930 (.179) 1.058 (.236) 1.200 (.270) 95 Table B.6 (continued) Father has another biological child Model 1 .920 (.130) Model 2 1.040 (.154) Cohabiting with mother (mother’s report) — .182*** (.028) Lives in Oakland 1.574 (.578) 2.391** (.737) Interaction between living in Oakland and mother having earnings .222** (.136) .259* (.162) Interaction between living in Oakland and mother living in public/subsidized housing 3.010* (1.554) — No. 2,543 2,543 Pseudo R2 0.071 0.146 NOTES: Standard errors are in parentheses. Model 1 controls for the interaction of father living in Oakland and mother having earnings and father living in Oakland and mother living in public/subsidized housing. Model 2 controls for the interaction of father living in Oakland and mother having earnings. *p < = .05. **p < = .01. ***p < = .001. 96 Bibliography Anderson, Elijah, “Sex Codes and Family Life Among Poor Inner-City Youths,” Annals of the American Academy of Political and Social Science, Vol. 501, 1989, pp. 59–78. Brown, Susan L., “Union Transitions Among Cohabitors: The Significance of Relationship Assessments and Expectations,” Journal of Marriage and the Family, Vol. 62, No. 3, 2000, pp. 833–846. Bumpass, Larry, and Hsien-Hen Lu, “Trends in Cohabitation and Implications for Children’s Family Contests in the U.S.,” Population Studies, Vol. 54, 2000, pp. 29–41. Bumpass, Larry L., James A. Sweet, and Andrew Cherlin, “The Role of Cohabitation in Declining Rates of Marriage,” Journal of Marriage and the Family, Vol. 53, 1991, pp. 913–927. California Department of Social Services, State Plan for the Provision of Public Assistance Under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, October 10, 1999, available at www.dss.cahwnet.gov/whatnew/default.htm (September 12, 2000). Center on Fathers, Families, and Public Policy, Policy Briefing, Vol. 3, No. 3, April 2000, available at www.cffpp.org/faxbriefings/ apriltwoonefaxbriefing.htm (May 9, 2001). Cowan, Carolyn Pape, and Philip A. Cowan, “Working with Couples During Stressful Transitions,” in Soolly Dreman (ed.), The Family on 97 the Threshold of the 21st Century: Trends and Implications, Lawrence Erlbaum Associates, Mahwah, New Jersey, 1997. Curtin S. C., and J. A. Martin, Births: Preliminary Data for 1999, National Vital Statistics Reports, Vol. 24, No. 14, National Center for Health Statistics, Hyattsville, Maryland, 2000. Edin, Kathryn, “What Do Low-Income Single Mothers Say About Marriage?” Social Forces, Vol. 47, No. 1, 2000, pp. 112–133. Furstenberg, Frank F., “The Fading Dream: Prospects for Marriage in the Inner-City,” in Elijah Anderson and Dave Massey (eds.), Problem of the Century: Racial Stratification in the United States at Century’s End, Russell Sage, New York, Forthcoming. Government Executive Magazine Daily Briefing, “Beefing Up the Bureaucracy to Promote Marriage,” February 5, 2001, available at www.govexec.com (February 7, 2001). Kingsley, G. Thomas, Federal Housing Assistance and Welfare Reform: Uncharted Territory, The Urban Institute, Washington, D.C., 1997, available at www.urban.org. Knox, Virginia, Cynthia Miller, and Lisa A. Gennetian, Reforming Welfare and Rewarding Work: Final Report on the Minnesota Family Investment Program, Manpower Demonstration Research Corporation, 2000, available at www.mdrc.org/RecentPublications. htm. Lazere, Ed, Shawn Fremstad, and Heidi Goldberg, States and Counties Are Taking Steps to Help Low-Income Working Families Make Ends Meet and Move Up the Economic Ladder, Center on Budget and Policy Priorities, November 27, 2000, available at www.cbpp.org/ 11-27sfp.htm (January 1, 2001). 98 Lerman, Robert, and Theodora J. Ooms (eds.), Young Unwed Fathers: Changing Roles and Emerging Policies, Temple University Press, Philadelphia, Pennsylvania, 1993. Luker, Kristin, Dubious Conceptions: The Politics Of Teenage Pregnancy, Harvard University Press, Cambridge, Massachusetts, 1996. Manning, Wendy D., and Pamela J. Smock, “Why Marry? Race and the Transition to Marriage among Cohabitors,” Demography, Vol. 32, No. 4, 1995, pp. 509–520. McLanahan, Sara, Irv Garfinkel, and Maureen Waller, “The Fragile Families and Child Wellbeing Study: Oakland, California,” 1999, available at www.ppic.org/publications/occasional.html. McLanahan, Sara, and Gary Sandefur, Growing Up with a Single Parent: What Hurts, What Helps, Harvard University Press, Cambridge, Massachusetts, 1994. Milbank, Dana, “Fixing Families Is a Priority,” The Washington Post, Weekly Addition, April 23–29, 2001. Mincy, Ronald B., and Hillard Pouncy, “There Must Be 50 Ways to Start a Family,” in Wade F. Horn, David Blankenhorn, and Mitchell B. Pearlstein (eds.), The Fatherhood Movement: A Call To Action, Lexington Books, Lanham, Maryland, 1999. Moynihan, Daniel Patrick, The Negro Family: The Case for National Action, U.S. Department of Labor, Office of Policy Planning and Research, Washington, D.C., 1965. Murray, Charles, Losing Ground: American Social Policy, 1950–1980, Basic Books, New York, 1984. 99 National Women’s Law Center and the Center on Fathers, Families, and Public Policy, Improving Paternity Establishment Practices and Procedures for Low-Income Mothers, Fathers, and Children, Washington, D.C., and Madison, Wisconsin, 2000. Padilla, Yolanda C., and Nancy E. Reichman, “Low Birthweight: Do Unwed Fathers Help?” Children and Youth Services Review, Vol. 23, No. 4, 2001, pp. 505–530. Primus, Wendell, and Kristin Daugirdas, Improving Child Well-Being by Focusing on Low-Income Non-Custodial Parents in Maryland, The Abel Foundation, Baltimore, Maryland, 2000. Reichert, Dana, “Why Focus on Fathers?” in Connecting Low Income Fathers and Families: A Guide to Practical Policies, National Conference of State Legislators, 2000a, available at www.calib.com/ peerta/policies/pdf/whyfocus.pdf (April 19, 2001). Reichert, Dana, “Financing Fatherhood Programs,” in Connecting Low Income Fathers and Families: A guide to Practical Policies, National Conference of State Legislators, 2000b, available at www.calib.com/ peerta/policies/pdf/financin.pdf (April 19, 2001). Reichman, Nancy, Julien Teitler, Irwin Garfinkel, and Sara McLanahan, “The Fragile Families and Child Wellbeing Study: Background, Research Design, and Sampling Issues,” Children and Youth Services Review, Vol. 23, No. 4, 2001, pp. 303–326. Sard, Barbara, and Jeff Lubell, The Value of Housing Subsidies to Welfare Reform Efforts, Center on Budget and Policy Priorities, February 24, 2000, available at www.cbpp.org/2-24-00hous.htm (January 22, 2001). 100 Sorensen, Elaine, Ronald Mincy, and Ariel Halpern, Redirecting Welfare Policy Toward Building Strong Families, The Urban Institute, Washington, D.C., 2000, available at http://www.urban.org. Spetz, Joanne, “Health Outcomes,” in Belinda Reyes (ed.), A Portrait of Race and Ethnicity in California, Public Policy Institute of California, San Francisco, California, 2001. Stacey, Judith, “Backward Toward the Postmodern Family: Reflections on Gender, Kinship, and Class in the Silicon Valley,” in A. Wolfe, (ed.), America at Century’s End, University of California Press, Berkeley, California, 1991. StataCorp, Stata Statistical Software: Release 5.0, Stata Corporation, College Station, Texas, 1997. Teitler, Julien O., Nancy E. Reichman, and Susan Sprachman, “CostBenefit Analysis of Improving Response Rates for a Hard-to-Reach Population,” presentation at the American Association of Public Opinion Research Annual Meeting, Montreal, May 17–20, 2001. Thornton, Arland, and Deborah Freedman, “Changing Attitudes Toward Marriage and Single Life,” Family Planning Perspectives, Vol. 14, 1982, pp. 297–303. Tucker, Belinda M., and Claudia Mitchell-Kernan (eds.), The Decline in Marriage Among African-Americans, Russell Sage, New York, 1995. UCLA Academic Technology Services, SPSS Frequently Asked Questions: What Does Cronbach’s Alpha Mean? available at www.ats.ucla.edu/stat/ spss/faq/alpha.html (March 26, 2001). 101 U.S. 104th Congress, The Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Public Law 104-193, U.S. Government Printing Office, August 22, 1996, available at www.thomas.loc.gov (August 29, 2000). U.S. 107th Congress, The Child Support Distribution Act of 2001, H.R. 1471, April 4, 2001a, available at www.thomas.loc.gov (May 9, 2001). U.S. 107th Congress, The Strengthening Working Families Act of 2001, S. 685, April 3, 2001b, available at www.thomas.loc.gov (May 9, 2001). U.S. 107th Congress, The Responsible Fatherhood Act of 2001, H.R. 1300, March 29, 2001c, available at www.thomas.loc.gov (May 9, 2001). U.S. Department of Health and Human Services, Helping Families Achieve Self-Sufficiency: A Guide on Funding Services for Children and Family Through the TANF Program, Washington, D.C., December 21, 1999, available at www.acf.dhhs.gov/programs/ofa/funds2.htm (September 11, 2000). U.S. Office of Budget and Management, Budget of the United States, Fiscal Year 2000, available at www.whitehouse.gov/omb/budge/ index.html (April 12, 2001). Ventura, S. J., and C. Bachrach, Nonmarital Childbearing in the United States, 1940–99, National Vital Statistics Reports, Vol. 48, No. 16, National Center for Health Statistics, Hyattsville, Maryland, 2000. Ventura, S., C. Bachrach, L. Hill, K. Kaye, P. Halcomb, and E. Koff, “The Demography of Out-of-Wedlock Childbearing,” in Report to Congress on Out-of-Wedlock Childbearing, U.S. Department of Health and Human Services, (PHS) 95-1257, Hyattsville, Maryland, 1995. 102 Waller, Maureen R., “Meanings and Motives in New Family Stories: The Separation of Reproduction and Marriage Among Low-Income, Black and White Parents,” in M. Lamont (ed.), The Cultural Territories of Race: Black and White Boundaries, University of Chicago Press and the Russell Sage Foundation, Chicago and New York, 1999. Waller, Maureen, and Robert Plotnick, Child Support and Low-Income Families: Perceptions, Practices, and Policy, Public Policy Institute of California, San Francisco, California, 1999. Wilson, William Julius, The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy, University of Chicago Press, Chicago, Illinois, 1987. Wilson, William Julius, When Work Disappears: The World of the New Urban Poor, Alfred A. Knopf, New York, 1996. 103 About the Author MAUREEN WALLER Maureen Waller is a sociologist with research interests in poverty, the family, culture, and social policy. Her previous work uses qualitative methods to examine how low-income families reconcile informal support with the requirements of the welfare and child support systems. She holds a B.A. in political science from the University of Dayton and an M.A. and Ph.D. in sociology from Princeton University. 105 Other Related PPIC Publications Child Support and Low-Income Families: Perceptions, Practices, and Policy Maureen Waller, Robert Plotnick Expensive Children in Poor Families: The Intersection of Childhood Disabilities and Welfare Marcia K. Meyers, Henry E. Brady, Eva Y. Seto The Fragile Families and Child Wellbeing Study: Oakland, California Sara McLanahan, Irwin Garfinkel, Maureen Waller PPIC publications may be ordered by phone or from our website (800) 232-5343 [mainland U.S.] (415) 291-4400 [Canada, Hawaii, overseas] www.ppic.org 107" ["post_date_gmt"]=> string(19) "2017-05-20 09:35:13" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_701mwr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:35:13" ["post_modified_gmt"]=> string(19) "2017-05-20 09:35:13" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_701MWR.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }