Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Transmission and transitions: young adults' beliefs, values and life course transitions in familial context
(USC Thesis Other)
Transmission and transitions: young adults' beliefs, values and life course transitions in familial context
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
TRANSMISSION AND TRANSITIONS: YOUNG ADULTS’ BELIEFS, VALUES AND LIFE COURSE TRANSITIONS IN FAMILIAL CONTEXT by Casey E. Copen ___________________________________________________________________ A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (SOCIOLOGY) August 2008 Copyright 2008 Casey E. Copen ii Dedication To my parents, Peggy and Steve Copen, whose unwavering support and love have guided me through the long road of graduate school. They have always believed in my abilities to achieve my goals. Without them, this dissertation would not have been possible. My sister, Erin Copen, for her pragmatic advice during stressful times and her continued and tireless devotion to helping me navigate the professional world. Thank you, Erin. My grandparents Dorothy and Reuben Copen and Berta and Herbert Wilkinson. They would have been proud to witness my accomplishments. I work toward the next step in my career with their memories always close to my heart. My family is what makes me whole. To them, I dedicate this dissertation and my career in family sociology and demography. Thank you for inspiring me, for giving me strength when I’ve needed it most, and most of all, for taking pleasure in watching me grow. iii Acknowledgements I would like to thank the members of my dissertation committee for their thoughtful guidance, valuable insight and continued support: Dr. Lynne Casper, Dr. Merril Silverstein, Dr. Timothy Biblarz and Dr. Eileen Crimmins. I am grateful for the time each member has spent in helping me develop my own research questions. I am also thankful for their understanding and patience during the time it took for me to “discover” the kind of research that makes me tick: family sociology and demography, particularly concerning young adulthood. My research interests today diverge quite dramatically from the origin of my graduate career in gerontology but I would not trade the journey! Special thanks to Dr. Lynne Casper, my advisor, who fuels in me a growing love for family demography and who pushed me to fully realize my true potential as a scholar. Thanks also go to Dr. Merril Silverstein. He inspired me to give great attention to detail in statistical modeling and a desire to explore research questions from many different angles. A gracious thank you also goes to Dr. Timothy Biblarz, whose thoughtful advice throughout my graduate career has been truly heartfelt. Special thanks also to Dr. Vern Bengtson whose wisdom, insight and expertise in family research go unmatched. Thank you all for helping me succeed. Last, thanks to the Longitudinal Study of Generations (LSOG) and the Transmission of Religion Across Generations (TRAG) projects for funding my research. Dr. Norella Putney was particularly helpful in this regard. Working with these projects has been invaluable in teaching me about longitudinal design and methodology. iv Table of Contents Dedication ii Acknowledgements iii List of Tables v List of Figures vii Abstract viii Chapter One: Introduction, Background and Conceptual Framework 1 Chapter Two: Data Description 32 Chapter Three: The Transmission of Religion Across Generations: Do Grandparents Matter? 36 Chapter Four: The Effects of Age and Life Course Transitions on Changes in Values Toward Individualism and Collectivism Among Young Adults 56 Chapter Five: The Effects of Parents’ Education on the Union Formation Process of Young Adults: Are There Differences by Gender and Union Type? 85 Chapter Six: Discussion and Conclusions 114 Bibliography 121 v List of Tables Table 3.1: Characteristics of Grandmothers, Parents, and Grandchildren, LSOG 1971-2000 46 Table 3.2: Standardized Regression Estimates Predicting Religiosity in Grandchildren (N = 529) 48 Table 3.3: Standardized Regression Estimates for Interactions Predicting Religiosity in Grandchildren (N = 529) 50 Table 4.1: Correlation Matrix for Items Included in the Individualism and Collectivism Scales Among Young Adults Ages 16-35, LSOG 2000 (N=451) 65 Table 4.2: Means, Standard Deviations and Descriptions of Variables Used In Fixed and Random Effects Models Among Young Adults Ages 16-35, LSOG (N=479) 69 Table 4.3: Descriptive Statistics of Married, Cohabiting and Young Adults with Children, LSOG (N=451) 70 Table 4.4: Random Effects Regression Model Relating Age and Gender to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG 74 Table 4.5: Fixed Effects Regression Model Relating Age to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG 74 Table 4.6: Random Effects Regression Model Relating Age, Gender and Parental Characteristics to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG 75 Table 4.7: Random Effects Regression Models Relating Age, Gender, Parental Characteristics and Young Adults’ Life Course Transitions to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG 77 Table 4.8: Fixed Effects Regression Models Relating Age and Young Adults’ Life Course Transitions to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG 81 vi List of Tables (continued) Table 5.1: Unweighted Descriptive Statistics for Variables Used in Analysis of Union Formation: National Survey of Families and Households (N=2,388) 99 Table 5.2: Odds Ratios of the Effects of Parents’ Education and Men’s Educational Experiences on Union Formation, NSFH 101 Table 5.3: Odds Ratios of the Effect of Parents’ Education and Women’s Educational Experiences on Union Formation, NSFH 104 Table 5.4: Odds Ratios of the Effects of Parents’ Education and Women and Men’s Educational Experiences on Marriage as a First Union, NSFH 107 Table 5.5: Odds Ratios of the Effects of Parents’ Education and Women and Men’s Educational Experiences on Cohabitation as a First Union, NSFH 109 vii List of Figures Figure 1.1: The Transmission of Religious Beliefs Across Generations: Do Grandparents Matter? 8 Figure 1.2: The Effects of Age and Life Course Transitions on Changes In Values Toward Individualism and Collectivism Among Young Adults 9 Figure 1.3: The Effects of The Effects of Parents’ Education on the Union Formation Process of Young Adults: Are There Differences by Gender and Union Type? 10 Figure 4.1: A Comparison of G3 Parents and G4 Children’s Individualism Values at Age 20 in 1971 and 2000 71 Figure 4.2: A Comparison of G3 Parents and G4 Children’s Collectivism Values and Age 20 in 1971 and 2000 72 viii Abstract This dissertation aimed to explain how young adults’ social origins influence their beliefs, values and behaviors in adulthood. More specifically, the objectives were: 1) to measure the extent to which there is intergenerational religious continuity in the family; 2) to identify the ways in which parental divorce and gender condition the religious transmission process in families; 3) to examine how the acquisition of adult roles, such as marriage and parenthood, influence young adults’ values toward individualistic and collectivism; 4) to assess how status attainment processes within the family influence the union formation process and 5) to determine how the effect of parents’ education on their children’s union formation varies by gender and union type. Analyses were performed using two data sources: 1) the Longitudinal Study of Generations (LSOG)—a multi-wave four-generation study, and 2) the National Survey of Families and Households (NSFH)— a nationally representative study of parents and their children. Multivariate models from the third chapter using the LSOG (N=565) showed strong resemblance of grandmothers, parents and children on religious beliefs and values. This chapter also demonstrated that parental divorce weakened the intergenerational transmission of religion within families. Using the LSOG, my fourth chapter uses fixed and random effects to study value changes in young adulthood (N=451). These models revealed a strong shift in values toward collectivism with the occurrence of life events such as marriage and parenthood. In contrast, young adults who cohabited with a partner became more individualistic over time. These findings suggest values are shaped by social roles adopted in young adulthood. Event-history models from the fifth chapter (N=2,388) using the NSFH indicated parents’ college education reduces the monthly ix odds of young adults’ union formation. This finding suggests the transmission of educational attainment from parents to children influences the timing of young adults’ first unions. As a whole, my dissertation provides evidence parents—and grandparents— influence young adults’ beliefs, values and behaviors. My dissertation also demonstrates young adults’ values toward individualism and collectivism are malleable over the life course. 1 Chapter One: Introduction, Background and Conceptual Framework A. Introduction The stage of life from adolescence to young adulthood has been defined as ‘emerging adulthood’, a time period characterized by a high degree of change, diversity and experimentation (Arnett, 2000; Rindfuss, 1991). High school graduation, completion of college and entrance into the labor force are examples of life course transitions which occur during young adulthood. Romantic relationships also develop and many young adults marry and have children during this period. Although these transitions today may appear to be similar to the age-structured life course of a half-century ago (Kertzer & Schaie, 1989), today they differ widely in their timing, context and meaning. Life course transitions today bear little resemblance to those described as part of the family life cycle theory, proposed by Glick (1947). The family life cycle theory was defined as a succession of critical stages which altered the relationships between family members. Such turning points included marriage, birth of first child, marriage of first child and death of spouse. The family life cycle theory was later revised once U.S Census Bureau trends revealed an increase in the marital age among women, as well as a decline in marital fertility (Glick, 1977). The rise in cohabitation was also documented in the late 1970’s (Casper & Cohen, 2000; Glick & Norton, 1977). However, the early reports of demographic changes could not have predicted the rising incidence of cohabitation over the past thirty years, or the shifts in attitudes and values which resulted from and spurred on these behavioral changes. Since the 1970’s, cohabitation has become a widely accepted living arrangement in the United States. The National Survey of Family Growth (NSFG) shows over half of 2 women in their late thirties have cohabited with a romantic partner (Bumpass & Lu, 2000). Revised estimates have shown about 60 percent of women ages 24 and younger have ever cohabited and many young adults expect to cohabit before marriage (Manning, Longmore & Giordano, 2007; Schoen, Landale & Daniels, 2007). Today, cohabitation is viewed as an acceptable alternative to being single and as a stage in the marriage process (Heuveline & Timberlake, 2004; Smock & Gupta, 2002). Younger cohorts with new ideas concerning the meaning of love, relationships and sexuality have replaced older cohorts with more traditional views regarding family formation (Ryder, 1965). Values toward cohabitation and marriage are shaped by young adults’ life course transitions that accompany union formation, such as residential independence, educational experiences and entry into the labor force (Guzzo, 2006). The rise in cohabitation is one example of a macro-level social change that contributes to changes in values over time. Another example of macro-level change is increasing rates of college graduation in the United States, particularly among women (DiPrete & Buchmann, 2006; Sayer, Cohen & Casper, 2004). Women who are college graduates are more likely to marry than women with less than a college degree (Goldstein & Kinney, 2001). College-educated women also have less favorable attitudes toward divorce than women with less than college (Martin & Parashar, 2006). In addition to changing values at the macro-level, the increase in educational attainment has changed the context of intergenerational transmission of values and behaviors within the family. For example, highly educated parents who instill similar values toward education in their children are likely to have children who marry at later ages (Bramlett & Mosher, 2002; Teachman & Paasch, 1998). Each of the three analytical chapters of my dissertation 3 measures intergenerational transmission of values and behaviors within the family and studies the mechanisms by which they are transmitted. Young adults’ union formation decisions originate from experiences in childhood and early adolescence. The family of origin influences young adults’ union transitions in three important ways. Most frequently studied is the relationship between parents’ own marital transitions and children’s union formation behaviors. The majority of this research has centered on the intergenerational transmission of divorce. This research suggests the propensity to divorce is strongly transmitted between parents and children (Amato, 1996; Amato & Keith, 1991). Parents’ marital history also influences children’s marital timing and whether they cohabit (South, 2001; Teachman, 2003). For example, children from divorced families are often more likely to enter cohabitating unions, compared to children from intact families (Axinn & Thornton, 1993; Cherlin, Kiernan & Chase-Lansdale, 1995; Thornton, 1991; Teachman, 2003). These studies suggest childhood family structure is an important “breeding ground” for attitudes about marriage and the development of interpersonal skills that sustain healthy relationships (Amato & DeBoer, 2001). A second avenue of intergenerational transmission is parents’ socioeconomic status. Parents’ education and income influence their offspring’s union formation decisions through the provision of social and economic resources that facilitate academic success (Axinn & Thornton 1992). Highly educated parents are more likely to be involved in children’s education, such as helping with homework and paying for college (Astone & McLanahan, 1991; Steelman & Powell, 1991). Parents’ educational attainment also has a direct effect on their offspring’s educational attainment through 4 their educational expectations for children (Sewell & Shah, 1968; Teachman & Paasch, 1998). Young adults who have high educational aspirations are likely to enroll in higher education and/or graduate college, thereby delaying union formation (Thornton, Axinn & Teachman, 1995). The last route through which parents’ influence their children’s union formation behaviors are the values shared between parents and children (Axinn & Thornton, 1993). Intergenerational influence is defined in this chapter as family socialization, or the extent to which family members resemble one another’s values and behaviors. The family is an important conduit through which values are shared across generations. Maternal preferences for the timing of marriage and parenthood decisions influence the rate at which their offspring, particularly daughters, enter marriage and/or have children (Barber, 2000). Family socialization, particularly concerning religious beliefs, also influences young adults’ attitudes toward school and work, thereby influencing the rate at which unions are formed (Glass & Jacobs, 2005). Indeed, family socialization goes beyond two generations to include grandparents as important socializing agents (see Chapter 3). B. Importance of this Study My dissertation builds on previous research examining parents’ influence on union formation in three important ways. First, my research builds on two-generation models to include grandparents as important agents of family socialization. The few, but growing, number of three-generation studies suggest grandparents influence grandchildren’s emotional well-being in adulthood (Amato & Cheadle, 2005; Goodman & Silverstein, 2001). Thus grandparents are important in the lives of their grandchildren 5 (Uhlenberg, 2005). The third chapter of my dissertation expands on these studies and demonstrates that grandparents are an important, yet overlooked, source of religious influence within families. The second contribution of my dissertation is its focus on value changes and life course transitions in young adulthood. Educational and work experiences influence young adults’ values toward gender roles, family life and career success (Cunningham, Beutel, Barber & Thornton, 2005; Clarkberg, Stolzenberg & Waite, 1995; Johnson, 2002). However, most studies on the relationship between values and union formation examine how values influence the likelihood of an event occurring within a particular time period, such as the timing of marriage. These studies treat values as explanatory variables (y) and event occurrence as the dependent variable (x). In contrast, the fourth chapter of my dissertation investigates how life events influence values. Thus my approach examines the change in values over time as the dependent variable. Understanding how life events influence values is important because longitudinal research suggests values change in response to life experiences (Axinn & Barber, 1997; Johnson, 2002). The fourth chapter offers a detailed look at five life course transitions— college graduation, college enrollment, marriage, cohabitation and parenthood—to asses how values change in relation to these life events. Models of status attainment and social mobility are distinctly intergenerational in scope (Sewell & Hauser, 1980, Sewell & Shah, 1968). Likewise, decisions to cohabit and/or marry a partner are closely linked to the experiences of parents (Axinn & Thornton, 1996; Barber, 2000). However, few studies have examined the ways in which status attainment within families influences the union formation process (see Axinn & 6 Thornton, 1992; South, 2001 for notable exceptions). The reevaluation of the relationship between status attainment and union formation is important because it allows for a more complete explanation of the ways socioeconomic origins influence life course decisions in adulthood. The fifth chapter explores the ways in which status attainment, such as parents’ education and income, influences the union formation process. The study of social origins in childhood and their effects on adult life chances has been one of the main areas of life course research (O’Rand, 2006). Status attainment research has focused on how parental transfers and socialization confer advantages to children in later life (Haveman & Wolfe, 1995). Studies on status attainment within families have most often focused on how parents’ socioeconomic status influences their children’s educational attainment and occupational position in adulthood. My dissertation research takes on an important piece of the status attainment model that has not been studied to examine how parents’ education influences their children’s union formation decisions. The fifth chapter of my dissertation explores whether the transmission of status attainment from parents and children influences the type of union formed by young adults, as well as whether these effects vary by gender. C. Conceptual Framework There are several theories used to account for the ways in which young adults make decisions to cohabit and/or marry a partner. First, I discuss the conceptual models for each of my research studies which illustrate the relationship between variables used in the three chapters of my dissertation. Second, I describe the life course perspective, which is the overarching framework for my dissertation research. Third, I bring in theories of family socialization to provide a basis for understanding how union formation 7 process among young adults is influenced by one’s family of origin. Fourth, the theory of individuation is discussed to put into context the family’s influence on young adults’ union formation in a world of increasing diversity and complex life choices. Lastly, theories used to explain attitudinal/values changes over time are discussed in order to better understand how life events, such as college completion, marriage, cohabitation and parenthood influence values over the life course. Conceptual Models Figures 1.1, 1.2 and 1.3 show the conceptual models for each of the three chapters of my dissertation. There are two themes which show the relationship between these three chapters. First, each of the three chapters of my dissertation examines the effects of family background on young adults’ beliefs, values and behaviors. By family background, I am referring to variables which capture important characteristics of the family of origin which influence the beliefs and values of young adults. These variables include parents’ (and grandparents’) beliefs and values, educational attainment, household income and parents’ marital transitions. Chapter Three includes the effects of parents’ (and grandmothers’) education and parents’ divorce on young adults’ religious beliefs. Chapter Four measures the effect of parents’ education and parents’ divorce on their children’s individualistic/collectivistic values. These studies also include the independent effect of parents’ (and grandmothers’) beliefs and values on their children. Parents’ own beliefs and values are included as a measure of family socialization, which is defined as the resemblance of beliefs and values within families. Figures 1.1 and 1.2 illustrate the variables measured in Chapter Three and Chapter Four, respectively. 8 Figure 1.1 displays the relationship between grandmothers’ and parents’ religiosity on young adults’ own religiosity, while controlling for parents’ education (at least some college), parental divorce and family lineage. Young adults’ attainment of at least some college, gender and age are also included in the model to ascertain how these influence the transmission of religious beliefs and values across generations. Figure 1.1: The Transmission of Religious Beliefs Across Generations: Do Grandparents Matter? Figure 1.2 shows how the effects of family background, measured in this chapter as parents’ education (one parent has college or more) and parental divorce influence young adults’ individualistic and collectivistic values. Similar to parents’ (and grandmothers’) religiosity included in Chapter Three, parents’ values toward Family Background Grandmothers’/ Parents’ Religious Values Grandmothers’/Parents’ Education (At least some college) Parental Divorce Family Lineage (Maternal/Paternal) Young Adults’ Religious Values Controls: Parents’/Young Adults’ Gender Young Adults’ Age Young Adults’ Life Course Transitions: 1) Marriage 2) Parenthood 3) Educational Attainment (At least some college) 9 individualism/collectivism are included in Chapter Four to gauge the effects of family socialization on young adults’ values. Figure 1.2: The Effects of Age and Life Course Transitions on Changes in Values Toward Individualism and Collectivism Among Young Adults The inclusion of family background variables in Chapter Five are also in line with the first theme of my dissertation research. Figure 1.3 illustrates the relationship between parents’ education, household income and parents’ marital transitions on young adults’ entry into a first union. These variables are included to assess the independent effects on the family of origin on the monthly odds of young adults’ marriage and cohabitation. Family Background Parents’ Individualistic/Collectivistic Values Parents’ Education (one parent has college degree or more) Parental Divorce Young Adults’ Life Course Transitions: 1) Marriage 2) Cohabitation 3) Parenthood 4) College degree attainment 5) College enrollment Young Adults’ Individualistic/Collectivistic Values Controls: Parents’/Young Adults’ Gender Young Adults’ Age 10 Figure 1.3: The Effect of Parents’ Education on the Union Formation Process of Young Adults: Are There Differences by Gender and Union Type? The second theme of my dissertation is each chapter includes one or more measures of young adults’ life transitions and how they are linked to young adults’ beliefs, values and behaviors. As shown in Figure 1.1, young adults’ marital and parenthood status are included in the model to assess these effects on young adults’ religiosity. Chapter Four includes five life transitions among young adults to measure the effects of these life events on young adults’ values toward individualism/collectivism over time. These life transitions are enumerated in Figure 1.2. Young adults’ life transitions are also included in Chapter Five. As shown in Figure 1.3, college enrollment and college degree attainment are measured for their influence on the timing of young Family Background Mother/Father has college degree Mother/Father has at least some college Household income Parents’ marital transitions (separated, divorced, remarried, cohabited, widowed) Young Adults’ Life Course Transitions: 1) College degree attainment 2) College enrollment Controls: Parents’/Young Adults’ Gender Young Adults’ Age Young Adults’ Entry into a First Romantic Union (Marriage or Cohabitation) 11 adults’ union and whether choose to marry or cohabit. The study of life course transitions in young adulthood is a focal point of my dissertation research. As described in the next section, the life course perspective is a useful framework to explain how the timing and context of these transitions have changed over time. Life Course Perspective My dissertation focuses on the interrelationships between three types of life course transitions: educational enrollment, college degree attainment and union formation. The life course perspective allows for the study of these transitions both separately and in conjunction with each other. The three principles of the life course perspective that frame my dissertation research include: 1) human agency, 2) timing of life transitions, and 3) linked lives. The principle of human agency states that people construct their own life courses based on a set of choices and constraints (Elder, Johnson & Crosnoe, 2003). Decisions to further schooling, enter a job, cohabit, marry and/or have children are made in response to the set of options available to the individual. These sets of option may be wider, or more selective, depending on the social constraints which are often beyond the purview of one’s control. Individual choices are made in the context of these structural constraints (Sewell, 1992). Theories which focus on human agency within a set of constraints, such as rational choice theory, are increasingly used in family demography to explain family-building behaviors (Casper & Bianchi, 2002; Hecter & Kanazawa, 1997). Second, the timing of life events is also important for these analyses, particularly because transitions into school, work, marriage and childbearing today are less likely follow a sequential order compared with fifty years ago. The meaning of these life events 12 changes according to the age at which they are experienced (George, 1993). Given the delay in marriage, the increasing variation in the timing of education and frequent job transitions, the age-structuring of the life course (Kertzer & Schaie, 1989) is a less applicable concept today, though a useful heuristic yardstick to measure such changes. Indeed, the idea of the “normal, expectable life cycle” (Neugarten, Moore & Lowe, 1965) is breaking down. The individualization of the life course (Settersen, 2003) is a result of the disintegration of these age-linked transitions, and is one of the focal points of my research. Third, the principle of linked lives emphasizes the idea that people’s lives are interdependent. Though the choices young adults make may stray, quite radically in some cases, from what their parents’ and other extended family members might have chosen, young adults’ values and behaviors are often not that different from their parents (Bengtson, Biblarz & Roberts, 2002). Although young adults are becoming more liberal concerning sexual attitudes, they are still strongly connected to ideals of marriage and family life (Thornton & Young-DeMarco, 2001; Treas, 2002). Much of the influence of parents on the attitudes and behaviors of their children arises from experiences in childhood. Although one of the principles of the life course perspective--human agency-- reminds us that decisions are up to the individual, attitudes and values are anchored within a family context. The effect of family background on young adults’ union transitions--measured by parents’ marital history, socioeconomic status and religiosity--is a cornerstone of my dissertation research. 13 Family Socialization The socialization hypothesis posits that children acquire attitudes and values and develop particular behaviors which are influenced by parents and other adults with whom they spend their formative years. There are several mechanisms through which parents and other family members influence their children’s attitudes, values and behaviors. One of the most important ways is through the attitudes transmitted between parents and children concerning marriage and cohabitation. Children who see their parents divorce are less likely to believe in the permanency of marriage, often because mothers’ attitudes towards marriage are less favorable after divorce (Axinn & Thornton, 1996; 1993). Attitudes toward cohabitation are also strongly transmitted between parents and children (Axinn & Thornton, 1996). Another important mechanism through which parents influence their children’s marital behaviors are through the modeling of parents’ own marital behaviors. A large body of research demonstrates the intergenerational transmission of divorce (Amato, 1996; Cherlin, Kiernan & Chase-Lansdale, 1995; Teachman, 2002). Likewise, parents who cohabit have children who are increasingly likely to cohabit with a partner (Axinn & Thornton, 1996; Cherlin, Kiernan & Chase-Lansdale, 1995). About 40 percent of children will, at some point during childhood, coreside with the unmarried partner of the custodial or biological parent (Bumpass & Lu, 2000). Thus parental cohabitation, particularly in the case of remarriage, has introduced a model of marriage behavior that children are increasingly likely to emulate. Parental religiosity is a third avenue of socialization that defines the timing and type of union formed among young adults. Religiously conservative families socialize 14 girls to value motherhood and childrearing over paid work (Glass & Jacobs, 2005). As adults, these women accrue less time in the labor force and have lower lifetime earnings due to earlier family formation (Glass & Jacobs, 2005). Similarly, many conservative Protestants view secular education as antagonistic to Christian values (Darnell & Sherkat, 1997). Given this view, many parents may be resistant, or unwilling, to support their children’s educational pursuits if it is not explicitly religious in nature (Sherkat & Darnell, 1999). In this case, religious socialization, where incongruous, has a strong, negative effect on the future life chances of children who stray from their religion of origin. Individuation Theory Most of the studies on young adults’ union transitions focus on the effects of family background, such as the occurrence of parental divorce and parents’ socioeconomic status, on the timing and type of union entered among young adults (Axinn & Thornton, 1992; Teachman, 2003; Wu, 1996). The influence of parents’ attitudes on their children’s union formation and childbearing decisions has also been studied (Barber, 2001; Axinn & Thornton, 1993). However, parents’ influence on their children’s union formation decisions has weakened over time. The reason for parents’ waning influence on the union formation decisions of their children stems from the rise in the ‘independent life stage’ among young adults (Rosenfeld & Kim, 2005). Young adulthood has taken on quite a different form than fifty years ago. Young adults today leave their parents’ home at earlier ages than in the past (Goldscheider, 1997). Marriage is a less frequent route through which young adults first leave home with many choosing to live with romantic partners or roommates (Goldscheider & Goldscheider, 1998). The 15 increased geographic mobility of young adults has lessened the control parents have over their children’s choice of partners (Rosenfeld, 2006). Indeed, the rise in independent living among young adults can explain, in part, the marked increase in interracial and same-sex partnerships over the last fifty years (Rosenfeld & Kim, 2005). The rise in non- traditional unions may provide evidence that union formation decisions are increasingly a personal decision rather than under parental control. The rise in independent living is one characteristic of the second demographic transition (SDT), a constellation of macro-level societal changes in living arrangements and fertility decisions (Lesthaeghe & Surkyn, 1988; Lesthaeghe, 1983). Other characteristics of the second demographic transition include the delay of marriage and childbearing and the rise in cohabitation. These trends, as linked to the SDT, have most often been studied in Western Europe, but trends in the United States closely resemble these countries (Lesthaeghe & Neidert, 2006). One of the mechanisms that drives the second demographic transition is a shift in values among young adults toward ‘higher- order’ needs (Maslow, 1954), or the push toward self-actualization and personal autonomy. Young adults today have many opportunities for personal growth and identity exploration (Arnett, 2000). The availability of non-family living arrangements encourages the view of the self as separate from one’s family of origin, enacting changes in family-related attitudes and values (Waite, Goldscheider & Witsberger, 1986). Young adults often shift their attitudes, preferences and norms according to their life experiences and the availability of alternatives (Barber, 2001). For example, young women who live apart from their parents are more likely to have positive attitudes about mothers in the labor force, have more egalitarian gender role attitudes and want fewer children than 16 those who live with their parents at the same ages (Waite, Goldscheider & Witsberger, 1986). In turn, residential independence exposes young adults to an array of new experiences and a variety of potential romantic partners. The individuation theory is a cultural theory of demographic change that attributes the declines in marriage and increase in non-marital childbearing, cohabitation and divorce to the emphasis placed on individual tastes and preferences (Lesthaeghe & Surkyn, 1988; Lesthaeghe, 1983). This theory provides an important framework for explaining why parents today have less influence over their children’s union formation decisions than in the past. As mentioned, residential independence is a major contributor to changes in attitudes and values among young adults. Changes in living arrangements are one example of a cultural shift, characterized by widespread attitudinal and value changes which are associated with changes in demographic behaviors. There are three types of attitudinal/values changes characterized by the individuation theory which are relevant to my dissertation: 1) religious secularization; 2) tolerance for non-conformist behaviors; and 3) increased emphasis on individualism, autonomy and egalitarianism (Lesthaeghe & Surkyn, 1988). These three attitudinal/values shifts have direct relevance for the theoretical framework of my dissertation and will be discussed in turn. Religious secularization refers to the decline in religious service attendance and the decreasing salience of religious beliefs that has characterized most of Western Europe and the United States. Whether Americans have become less religious over time has come under heated debate (Chaves, 1989; 1990; Hout & Greeley, 1990). Research on religiosity in the United States shows high levels of religious salience and religious 17 service attendance relative to Western European countries (Sherkat & Ellison, 1999). With increasing religious pluralism, however, religiosity is expressed and practiced much differently than at the turn of the last century (Wuthnow, 2005; 1998). Sociology of religion scholars have illustrated that many teenagers are inarticulate about their religious beliefs, opting for a declaration of faith that is much more humanistic (e.g. Do good unto others) than spiritual in origin (Smith, 2005). Similarly, most teenagers report that religion operates in the background of their lives, and is not a source of conflict with parents (Smith, 2005). These findings suggest that religion does not matter much to most American teens and may not have as large an impact on their life choices, such as premarital sexual activity and the timing of marriage and childbearing, as it would for religiously conservative teens. However, the decline in religiosity among American teenagers does not necessarily imply the moral influence of the family has receded completely. Rather, many teens cite the importance of following in the footsteps of their parents’ religious beliefs, and attending religious services, not because they are forced to, but because they want to (Smith, 2005). These responses indicate that parents and other relatives, to some extent, remain important in passing down religious beliefs, values and practices to their offspring. Whether the content of the message is distinctly religious in nature depends on how the family defines religion, as well how children use these messages to shape their identity as adults. Non-conformism to social rules is another ideational change that has led to variation in family formation behaviors. Widespread attitudinal changes are often due to cohort succession or the replacement of older generation members with a younger cohort 18 with different world views (Ryder, 1965). Variation in sexual practices, and increased tolerance for such deviation, is one example of social change that is due, in large part, to cohort succession. For example, in the late 1990’s, young adults were less likely to agree that homosexuality is wrong compared to older birth cohorts, suggesting tolerance for diverse sexual practices (Treas, 2002). Similar trends of liberalization are evident for other family-related behaviors as well, such as the rise in the ideal age at marriage and an increase in the acceptance of nonmarital heterosexual cohabitation among young adults (Thornton & Young DeMarco, 2001). In this case, young adults have more permissive attitudes toward sex and family-related behaviors because of increased education and a youthful receptivity to new ideas (Treas, 2002; Ryder, 1965). The third type of attitudinal change that characterizes the second demographic transition is the increase in attitudes and values which emphasize individualism, autonomy and egalitarianism. Several studies have charted the rise in individualistic and materialistic values among young adults in the United States from the 1970’s to the late 1980’s (Bellah, Madsen, Sullivan, Swidler & Tipton, 1985; Crimmins, Easterlin & Saito, 1991; Easterlin & Crimmins, 1991). However, the preoccupation with the self relative to family obligations and civic participation may have slowed in recent years. A more recent study has shown the trend toward increasing individualism has declined among young adults in the 1990’s (Bengtson, Biblarz & Roberts, 2002). In turn, young adults are placing more emphasis on collectivistic values than on individual goal attainment. Collectivistic values include those values which prize social networks, religious participation and family commitment. 19 What can account for the shift toward collectivistic values among young adults? First, many young adults report a ‘good’ marriage as important for a happy and fulfilling life (Thornton & Young DeMarco, 2001). Young adults in the 1990’s highly value the sanctity and permanency of marriage to a greater degree than in the mid-1970’s (Thornton & Young DeMarco, 2001). Young adults express a strong commitment and desire for marriage; thus, they are more likely to delay marriage until finding a suitable partner (Manning, Longmore & Giordano, 2007). Similarly, alternatives to marriage, such as cohabitation, allow individuals to test their relationship (Casper & Bianchi, 2002; Smock, 2000). Marriage is a significant life event for many young adults because it symbolizes maturity and economic stability (Cherlin, 2004). In turn, young adults are delaying marriage until reaching their personal and financial goals. Much like a rewarding career, young adults consider marriage as an important achievement which confers many benefits if it is a mutually satisfying relationship. Second, parenthood remains an important role for men and women. The majority of young men and women expect to be parents at some point in their lives (Chandra, Martinez, Mosher, Abma, & Jones, 2005). However, voluntary childlessness is viewed as an increasingly acceptable life choice, particularly among women (Thornton & Young DeMarco, 2001). The salience of the parenthood role, particularly in industrial societies where childrearing is expensive and children are not instrumentally valuable, is due in large part to the acquisition of social capital with the birth of a child (Schoen, Kim, Nathanson & Astone, 1997). Social capital is “a resource of individuals that emerges from their social ties” (Portes & Landolt, 1996). Individuals with children have increased access to instrumental and emotional assistance from extended family and neighbors. 20 Similarly, raising a child integrates parents into other social institutions, such as health care, schools, religious groups and other organizations. The presence of children widens adults’ social networks and bolsters existing family and community ties (Astone, Nathanson, Shoen & Kim, 1999). This is particularly true for men, who have fewer social and familial connections without children than women (Eggebeen & Knoester, 2001). In turn, the shift toward collectivism among young adults reflects, in part, the increase in the density of social networks as a result of marrying and/or having children. Theories of Attitudinal/Values Change Theories of attitudinal change help explain the rise in collectivistic values among young adults over time. Shifts in individual values can be attributed to developmental changes and role transitions. The ‘impressionable years’ hypothesis is often used to explain values changes as a result of developmental changes, such as aging (Alwin & Krosnick, 1991). This theory states individual’s attitudes are malleable in young adulthood but become increasingly stable as people age. Young adults are more amenable to attitudinal/value changes as they begin dating, have the ability to participate in political processes, attend college and enter the labor force (Mannheim, 1952; Ryder, 1965; Visser & Krosnick, 1998). Each of these life transitions leaves a “stamp” on young adults’ attitudes and world-views. Older adults, it is argued, are entrenched in social networks which reinforce their attitudes and values, and become more resistant to new ideas over time (Petersen, 1999). However, older adults have shown to change their attitudes over time, particularly on issues of gender and racial tolerance, civil liberties and privacy rights (Danegelis, Hardy & Cutler, 2007). Thus age is an important 21 contributor to attitudinal change, particularly concerning the timing of life events (e.g. age at time of marriage, parenthood, retirement). Individuals who marry and/or have children often change their attitudes concerning other domains of life, such as education and career development (Barber, 2001; Clarkberg, Stolzenberg & Waite, 1995). In other words, the adoption of multiple life roles engenders a reevaluation in one’s attitudes and values. Attitudes and values about family life vary by gender (Kapinus, 2004; Thornton & Young DeMarco, 2001), social class (Edin & Kefalas, 2005) and race/ethnicity (Crissey, 2005). The degree to which attitudinal/values change depends on the salience attributed to these social roles (Thoits, 1992). In particular, work demands compete with family responsibilities. The ways in which individual attitudes and values change with the adoption of multiple adult roles often depends on the balance between work and family (Bianchi, Casper & King, 2005). For example, men may scale back their work hours, or work longer, to provide for their family once having children (Kaufman & Uhlenberg, 2000). Likewise, higher- earning women may choose to delay marriage in order to pursue their career goals, choosing instead to remain single or cohabit (Clarkberg, 1999). Life course transitions during young adulthood are frequent and interrelated (Guzzo, 2006). Thus young adults’ career aspirations influence their attitudes toward marriage and childbearing and vice versa (Johnson, 2005). The timing and context of life transitions have changed rapidly over the past fifty years. These demographic changes have important consequences for attitudinal/value changes and are discussed in more detail in the next section. 22 D. Literature Review Changes in the Timing and Context of Union Formation in the United States The timing and context of marriage have undergone radical changes in the past fifty years. In particular, the postponement of marriage since 1970 has led to a substantial increase in never-married young adults (Fields, 2003). Cohabitation is often replacing marriage as a choice for first union (Bumpass, Sweet & Cherlin, 1991). Half of single young adults are cohabiting with a romantic partner (Casper & Bianchi, 2002). However, these living arrangements are often short-lived. The majority of cohabitations end within a year and there has been a decline in the proportion of cohabitating couples who marry (Bumpass & Lu, 2000). These trends point to the various contexts and functions of cohabitation as it becomes an increasingly normative living arrangement in the United States (Smock, 2000). Most adolescents expect to marry eventually; however, these expectations often include cohabitation as a testing ground for compatibility with a potential marriage partner (Casper & Sayer, 2000; Manning, Longmore & Giordano, 2007; Thornton & Young-DeMarco, 2001). In the United States, cohabitation is seen as a pathway to marriage, and also, less frequently, as a substitute for marriage (Smock & Gupta, 2002). The meaning attributed to a cohabiting relationship is closely tied to whether these couples marry or break up (Brown, 2000), as well as how union formation is viewed in relation to childbearing and other life goals (Lye & Waldron, 1997). Cultural explanations for the rise in cohabitation focus on the values and attitudes that give rise to shifts in the pattern of union formation and how these differ across social groups (Smock, Manning & Porter, 2005). Increasing emphasis on individual goal attainment relative to 23 family life has been widely used to explain the rise of cohabitation in Western Europe (Lesthaeghe, 1983, Lesthaeghe & Surkyn, 1988). The United States has seen a similar shift in attitudes toward self-fulfillment, termed the rise of individualism, which has fundamentally altered the way people view romantic relationships in relation to the self (Bellah, Madsen, Sullivan et al, 1985; Giddens, 1991). The Rise of Materialism, Individualism and the Union Formation Process There is mounting evidence young adults have become more materialistic over time (Easterlin & Crimmins, 1991). Many young adults desire jobs high in monetary rewards and prestige (Crimmins, Easterlin & Saito, 1991). In turn, many young adults with high materialistic aspirations may postpone or forgo marrying or starting a family to obtain a career that will allow them to have economic freedom in later life. Among many people, marriage may be seen as a transition made once achieving what one deems to be an adequate standard of living (Oppenheimer, 1988). Cohabitation is a living arrangement often chosen during a time of economic uncertainty, often when transitioning from school to the labor force, particularly among men (Oppenheimer, 2003). Young women may also delay marriage to gain a foothold in the labor force to bolster the competing demands of marriage and motherhood. In turn, cohabitation may be an attractive living option among high-earning women who do not want to trade their labor force participation for unpaid household labor (Clarkberg, 1999; Becker, 1981). The rise in individualistic values, or a set of beliefs focusing on the self relative to family obligations, has fostered a dynamic discussion of how attitudes toward the self contribute to family formation decisions among young adults (Cherlin, 2004). Since the 1960’s, participation in the labor force among women, particularly married women with 24 children, has risen (Sayer, Cohen & Casper, 2004). Although the increase in working married mothers has slowed over the last decade, the increased amount of time women allocate to paid work has called into question whether women’s economic independence destabilizes marriage (Sayer & Bianchi, 2000). However, this argument is mounted on an antiquated model of gender specialization within marriage (Becker, 1981; Oppenheimer, Kalmijin & Lim, 1997). Women do not only derive benefit from their husband’s labor force participation, but from his increasing assistance with household chores and childrearing (Sayer, 2005). Similarly, there have been widespread attitudinal changes toward egalitarian roles within marriage (Thornton & Young-DeMarco, 2001). Rising gender equality has allowed men and women the choice to delay marriage to fulfill their personal career goals, while providing them with the option to cohabit until economic solvency is reached. Variations in the Union Formation Process by Race/Ethnicity and Class The attitudes and expectations attributed to romantic relationships and marriage differ by race/ethnicity and social class. The dearth of ‘marriageable men’ in the African American community, a hypothesis first purported by Wilson (1987), proposes the chances of marriage among Black women in poor communities are lower because of neighborhood factors, such as unemployment and incarceration. The attitudes toward marriage among Black women are less favorable than whites (Crissey, 2005; Manning, Longmore & Giordano, 2007) and Black women report a mean age of marriage as a year and a half older than that of white women (East, 1998). Black women are less likely to cohabit than whites, but they are more likely to live with, rather than marry, their first romantic partner (Bramlett & Mosher, 2002; Raley, 1996). Among Black women, 25 cohabitation may be a long-term solution to not attaining the economic stability viewed as necessary for a successful marriage (Smock, Manning & Porter, 2005). Racial/ethnic explanations of marital timing would be remiss if they did not include how the African American family, both culturally and historically, has differed from that of whites (Sarkisian & Gerstel, 2004; Stack, 1974). Indeed, the definition of a family and its functions vary widely across racial/ethnic subgroups. However, economic explanations of marital timing work across racial/ethnic lines. Regardless of racial/ethnic background, many poor women place more importance on childbearing than marriage to maintain their independence, or to escape or avoid harmful romantic relationships (Edin & Kefalas, 2005). As a social institution, marriage implies financial security, as well as emotional readiness and maturity (Cherlin, 2004). For economically disadvantaged women, particularly those with children, marriage may be seen as an unattainable goal. Welfare policies, such as the Healthy Marriage Initiative, have targeted these ‘fragile families’, defined as an unmarried couple raising children, to participate in programs that encourage marriage (McLanahan, 2005). However, marriage is not necessarily an economic panacea for poor women (Lichter, Graefe & Brown, 2003). Further, studies have shown these programs do not significantly increase marriage or reduce non-marital childbearing (Blau, Kahn & Waldfogel, 2004; Moffitt, 1998). The relationship between educational attainment, labor force participation and marriage must be further examined to understand why social class matters in marital timing. School/Work Experiences and the Union Formation Process The transition into a cohabitation/marital union is often based on perceived economic stability, which is intimately tied to the amount of education one achieves, or 26 aspires to obtain. Young adults who are currently enrolled in college are often less likely to form unions because they are less financially stable than those who have completed a college degree (Thornton, Axinn & Teachman, 1995). Indeed, the increasing number of educationally homogamous marriages suggests educational attainment is a signal to others of one’s future economic success (Clarkberg, 1999; Schwartz & Mare, 2005). The relationship between education and likelihood of marrying has weakened over time, explained in part by the increase in cohabitation, which may serve as a substitute for marriage in times of economic uncertainty (Oppenheimer, 2003; Sassler & Goldscheider, 2004). Similarly, the timing and duration of schooling is increasingly varied, with many young married people still in school, or returning to obtain another degree (Jacobs & King, 2002). In particular, the rise in college completion among women has resulted in new strategies for selecting a suitable marriage partner (Blackwell & Lichter, 2004; Buchmann & DiPrete, 2006). Steady, well-paying employment is also an important predictor of marital timing, although it is increasingly less so among men (Sassler & Goldscheider, 2004). The job- search theory of marital timing helps explain partner choice, which is largely dependent on one’s labor market position (Oppenheimer, 1988). As stated, increasing opportunities for women in the labor market have initiated debate as to whether emphasis on women’s work roles result in fewer gains to marriage for women (Becker, 1981), which has since been refuted (Sayer & Bianchi, 2000). Indeed, women’s earnings have shown to increase the likelihood of marriage over time (Sweeney, 2002). The fluctuations of the labor market, as well as changing work-family roles, contribute to the pattern of delayed marriage among men and women. However, focusing on union transitions as the result 27 of economic status in young adulthood ignores the ways in which attitudes toward union formation are developed in childhood and influence subsequent union formation behaviors. The effects of social origins on union formation, namely parents’ marital history, parents’ income/education and childhood religiosity, are discussed next. Influence of Parents’ Marital Transitions on Children’s Union Formation Parents’ attitudes toward marriage, cohabitation and childbearing are important influences on children’s family formation behaviors (Barber, 2000; Axinn & Thornton, 1993). Parents’ attitudes have independent effects on the attitudes of their children via socialization techniques. Children may also adopt their parents’ attitudes toward marriage by modeling their parents’ marital behaviors. For example, parents’ marital quality is one route through which attitudes about marriage, cohabitation, divorce and premarital sex are shared between parents and their children (Cunningham & Thornton, 2005). In the case of marital dissolution, parents may develop less favorable attitudes toward marriage after divorce which their children may adopt (Axinn & Thornton, 1996). Similarly, parents’ cohabitation exposes young adults to alternative family living arrangements that may influence their own romantic relationships (Axinn & Thornton, 1993). In turn, parents’ marital behaviors influence the union formation behaviors of their children. In addition to their influence on children’s union formation, parents’ marital behaviors influence the timing and context of other life transitions experienced by children. Parents’ cohabitation and/or remarriage may result in adolescents leaving home at an earlier age, which may transform young adults’ attitudes toward marriage and affects the timing of union formation (Goldscheider & Goldscheider, 1998). For 28 example, children who leave home early as a result of marital disruption are more likely to engage in premarital sexual activity and cohabit with a partner (Cherlin, Kiernan & Chase-Lansdale, 1995). Young adults who experience marital disruption or parental cohabitation are also less likely to graduate high school (Biblarz & Gottainer, 2000; Raley, Frisco & Wildsmith, 2005) or attend college (Ploeg, 2002). Children’s educational aspirations stem from the availability of family resources and parental involvement (Teachman & Paasch, 1998). Educational attainment is one example of the long-standing effects of parental influence on the life course trajectories of their offspring. Influence of Parents’ Socioeconomic Status on Children’s Union Formation The resources afforded by one’s parents influence the transition into a cohabitation and/or marital union. Young adults whose parents have a college education and higher income are likely to delay marriage because of increased access to economic resources which increases the likelihood young adults remain at home longer (Axinn & Thornton, 1992). Highly educated parents are also more likely to instill similar educational aspirations in their children and encourage them to delay cohabitation, marriage and premarital childbearing until after finishing their education (Teachman & Paasch, 1998). The transmission of educational attainment from parents to children can be explained, in part, by parental investment in children’s educational success. However, parental investment in children is lower in disrupted families, particularly among blended families where investment in biological offspring may trump that of stepchildren (Astone & McLanahan, 1994). 29 Educational attainment is important in terms of young adults’ family formation because it is a clear marker of social mobility in adulthood (Biblarz & Raftery, 1999). Less educated parents who invest less time and money in their children’s education are setting them on a path that could include early marriage, increased likelihood of cohabitation and non-marital childbearing. For example, reduced parental investment in education among religiously conservative parents lowers the educational attainment of children (Sherkat & Darnell, 1999). Whether as a result of religion, marital disruption, lack of economic resources or uninvolved parenting, parental investment is a crucial indicator of children’s educational success. In sum, parents’ education and income have a demonstrable direct influence on the life course trajectories of their children, as well as an indirect influence through the value parents place on education and work. This socialization process begins in early childhood and has long-lasting effects into adulthood. Religious Socialization and the Union Formation Process Among Young Adults The family has been heralded as the premiere site for the socialization of beliefs, values and practices that accompany a religious tradition. The religious ideologies held and the religious rituals performed within the family have a defined impact on decisions made over the life course. Where agreed upon, religion serves to solidify emotional ties between family members (Pearce & Axinn, 1998). These tight social bonds provide a firm foundation for a family culture that leaves a lasting imprint on younger generations. Indeed, much research has indicated that the influence of religion pervades every aspect of social life, including demographic patterns of marriage and fertility (Bramlett & 30 Mosher, 2002; Mosher, Williams & Johnson, 1992), as well as education and occupational pursuits (Darnell & Sherkat, 1997; Glass & Jacobs, 2005; Lehrer, 1999). Given the meaning and practice of religion has changed in recent decades, parents’ ability to socialize children within a particular religious tradition has been called into question (see Edgell, 2006 for a review). The percent of married couples that both report being raised with no religion climbed from 16% in the early 1970’s to 32% in the late 1990’s (Hout & Fischer, 2002). Indeed, the likelihood of being raised with no religion, as well as marrying a spouse with no religious preference, seems to be gaining momentum. Today, people are more likely to choose a spouse of a similar educational background, rather than base their marriage decisions solely on cultural similarities, such as religion (Kalmijn, 1998; 1991). Boundaries between religious groups that were once quite distinct have blurred, while the economic opportunities afforded those with a college education, versus those without, have incurred more sharply divided social lines. The rise in interfaith marriages demonstrates that religion, as an ascribed characteristic, is taking a “backseat” to other status markers, such as education, in spouse selection. Parents who adhere to religious teachings often pass these beliefs on to their children through both their attitudes regarding non-marital sex, their beliefs regarding cohabitation, as well as how they behave in their own romantic relationships (Thornton, Axinn & Hill, 1992). Indeed, adolescents often ascribe to the religiosity of their parents (Smith, 2005). Thus the family has not lost its religious and moral influence and continues to pass on religious beliefs and traditions to children. Increasingly, rising longevity has afforded grandparents the opportunity to share their religious beliefs with grandchildren (King & Elder, 1999). The first chapter of my dissertation examines the 31 extent to which grandparents and parents influence the religiosity of children, and tests how these socialization processes are influenced by gender and parental divorce. 32 Chapter Two: Data Description A. The Longitudinal Study of Generations One data source for this dissertation is the University of Southern California’s Longitudinal Study of Generations (LSOG). Begun in 1971, grandparents, parents, and their adolescent or young adult children have been surveyed at regular intervals about their perceptions of family relationships as well as their values, religious beliefs and practices, political and social attitudes, physical and psychological health, work life, and personal goals. The data collection began just after the so-called Decade of Protest—the 1960s—when the generation gap between youth and elders became cause for concern around the world. An important goal of the study was to chart both behavioral and cognitive-affective dimensions of interactions between parents and children and grandparents and grandchildren across time and to examine the consequences of close or distant relations between generations. In 1985, the study was turned into a longitudinal panel study of aging families. Since 1985, respondents have been surveyed at three-year intervals, providing a total of eight waves of data (1991, 1994, 1997, 2000; and 2005). A fourth generation of great-grandchildren of the original grandparents was added in 1991 as they turned age 16. The LSOG began in 1971 as a cross-sectional study of three-generation families. The sampling frame was identified via a multi-stage stratified random sampling procedure from a population of 840,000 individuals enrolled in Southern California’s first Health Maintenance Organization (HMO) – Kaiser Permanente. The original sampling frame consisted of 2,044 respondents ages 16-91 from 328 three-generation families. The sampling strategy was to select grandparents (G1) who had grandchildren between the ages of 16 and 26. The HMO primarily served union members in the steel industry which resulted in a 33 sample of mainly White, middle and working class families. Minority families were underrepresented; however, educational levels were comparable with national norms (Bengtson, Biblarz & Roberts, 2002). The response rate between Wave 1 and Wave 2 was 65% and has averaged 74% between waves since then. The response rate is comparable to most longitudinal surveys. The sample is continuously replenished by the addition of spouses and great-grandchildren who reach 16 years of age. The LSOG contains many unique features which make it particularly useful for my dissertation research. First, the multigenerational design of the study allows me to explore the extent to which attitudes, values and behaviors are transmitted from parents to children and from grandparents to grandchildren and under what circumstances. These data show that parents have considerable influence on the educational aspirations of their offspring (Bengtson, Biblarz & Roberts, 2002). However, the study of social mobility using these data shows considerable weakening of status inheritance over time between parents and children (Biblarz, Bengtson & Bucur, 1996). Thus young adults today are more likely to have similar attitudes, aspirations and value orientations as their parents, but are not as likely to share their parent’s occupational status. In fact, each successive generation of offspring in the LSOG have higher occupational attainment than the previous generation. The LSOG is an invaluable resource for understanding intergenerational change and continuity over time. Second, the study’s generation-sequential design provides a method for examining different birth cohorts at approximately the same ages. By holding age constant, the historical milieu which gives rise to changes in attitudes and values can be explored. For example, young adults are often cited as having more individualistic and selfish motives for behavior. However, an age-matched comparison of Generation X youth in 1997 with their 34 parents twenty-six years earlier demonstrates young adults have higher aspirations, higher self-esteem and are more collectivistic than their parents (Bengtson, Biblarz & Roberts, 2002). As described in Chapter Four, collectivistic values were measured with four items: 1) importance of loyalty to family and community, 2) friendship, 3) patriotism and 4) religious participation. These findings underscore the importance of age-matched designs for understanding how attitudes and values change across generations and over time. Third, the LSOG is useful for understanding family relationships over time because of its longitudinal design. There are eight waves of data and most of the items are comparable across time. For example, the Rokeach Value Survey is asked at each time wave and provides important data on within-individual value changes over the life course. This last feature is particularly useful for Chapter Four which examines how values change in response to age and life course events in young adulthood. B. The National Survey of Families and Households The second data source used for my dissertation is the National Survey of Families and Households (NSFH). The first wave of the NSFH was conducted in 1987-88 and involved a multistage probability sample of 13,007 adult respondents (Sweet, Bumpass & Call, 1988). The NSFH followed up these respondents in 1992-1994 at which time their children were added to the sample. Focal children range in age from 10-23 at Wave 2. Recently, data collection was completed on a third wave of data (2001-2002) and extensive interviews were conducted of the focal children of the main respondent, ages 18-34. The NSFH offers several advantages to the study of union formation, which is the focus of Chapter Five. First, Wave 3 offers complete union histories of both parents and children, allowing me to measure the exact dates of entry and exit from romantic unions, 35 how they began (marriage or cohabitation) and how they ended (marriage, separation, divorced, widowhood). The timing and duration of unions is needed in order to conduct event history analyses which require time-dependent covariates. Second, the NSFH contains over-samples of Blacks, Mexicans, Puerto Ricans, one- parent families, families with stepchildren, cohabitors and recently married persons. The racial/ethnic over-samples are particularly important because the LSOG does not include an adequate number of minority families. The oversampling of cohabitors is also useful because these unions are usually of short duration (Bumpass & Lu, 2000). Similarly, the over- sampling of alternative family types, such as stepfamilies, is important so that union transitions, which are often frequent and complex, can be adequately assessed. Third, the NSFH offers a range of important family background variables, such as parents’ religiosity, parental supervision of child’s activities, the availability of family economic and social resources and parent-child relationship closeness. These variables are important for measuring childhood circumstances and the affective nature of parent-child relationships. 36 Chapter 3: The Transmission of Religious Beliefs Across Generations: Do Grandparents Matter? A. Introduction Although the transmission of religious values and beliefs across generations is one of the primary means by which continuity is maintained in the American family, cultural critics of contemporary family life maintain that the capacity of older generations to instill such values and beliefs in younger generations has attenuated as a result of marital instability and rampant individualism (Bellah, Madsen, Sullivan, Swidler & Tipton, 1985; Popenoe, 1993). From this perspective, the ability of parents to reproduce religious orientations in their children has diminished over the last several decades, presumably undermining the moral grounding of children. Changes in the religious landscape of the nation are also said to contribute to the devaluation of families as conduits of religious traditions, most notably the growth in secularism and the development of alternative spiritual movements that prize self-fulfillment and religious expression over religious conformity and the inerrancy of religious doctrine (Inglehart & Baker, 2000; Roof, 1993; Smith, 2005; Wuthnow, 1998). However, there have also been significant challenges to the assertion that the family has become too fragile an institution to pass down core values and beliefs to the next generation (Bengtson, Biblarz & Roberts, 2002; Bengtson, 2001; Cherlin, 2002). In this investigation I examine the intergenerational transmission of religious beliefs to adolescents and young-adults. Using a unique data set consisting of three linked generations, I test whether grandparents and parents, independently and jointly, influence the beliefs of grandchildren in the family’s youngest generation. Further I 37 examine whether family structure and the gender of sending and receiving generations condition the strength of this downward influence. Despite considerable social change in family and religious life over recent decades, I suggest that contemporary families possess a stock of religious capital sufficient to influence the religious orientations of successive generations. While religiosity is often considered (and measured) as a characteristic of an individual, it is also a social product forged by early experiences within nuclear and extended families. The literature suggests three mechanisms by which parents transmit their values and beliefs to their children: (1) Socialization or the direct training and instruction of children; (2) Social learning or role modeling that demonstrates desired outcomes to children; and (3) Status inheritance by which parents situate their children to particular social and economic contexts that predispose particular sets of values and world-views (Acock & Bengtson, 1980; Moen, Erikson & Dempster-McClain, 1997; Yi, Chang & Chang, 2004). Any or all of these mechanisms may be operating in the transmission process to maintain continuity across generations. Studies have shown that parents exert a lasting imprint on the religious ideologies and commitments of their children (Glass, Bengtson & Dunham, 1986; Myers, 2004; 1996; Sherkat, 1998) and that a majority of American teens prefer to adopt the religious traditions of their parents rather than seek out other religions (Smith, 2005). Less understood is the role that the wider family plays in the transmission of religious beliefs, values and practices. Indeed, children are embedded in a complex web of family relationships and receive messages about religious beliefs, values and expectations from parents, grandparents and other relatives (Mueller & Elder, 2003). 38 Until relatively recently, grandparents have been little considered in social science theory and research on intergenerational transmission and its consequences for youth. However, recent research indicates grandparents play an increasingly important role in the lives of their grandchildren (Bengtson, 2001; Casper & Bianchi, 2002). Owing to increases in life expectancy over the 20 th century, grandchildren spend more time with grandparents in their lives than ever before, increasing opportunities for shared activities and mutual benefit (Uhlenberg, 2005). Grandparents commonly share stories and experiences with their grandchildren, providing a “cultural window” into family history and traditions (King & Elder, 1999; Pratt & Fiese, 2004). Thus it is not surprising that grandparents, in maintaining a continuing presence in the lives of their grandchildren, also share similar views and values with them (Silverstein, Giarrusso, & Bengtson, 2003). Religion represents one of several intergenerational threads that may bring grandparents and grandchildren closer together. More religious grandparents report stronger emotional relationships and greater social interaction with their grandchildren than less religious grandparents (King & Elder, 1999). Yet, it is also likely that close relationships with grandchildren enable these more religious grandparents to pass down their religious beliefs, values, and practices to younger generations in the family. The role of parents as agents of religious socialization have also come under greater scrutiny as changes in family structure over the last half century--mostly due to increasing rates of divorce--have altered family functions as well (Casper & Bianchi, 2002). Parents may face greater challenges training their offspring to religious traditions in today’s more complex and fluid families. Indeed, scholars maintain that changing family forms have major implications for the religious training of children and their 39 consequent religious affiliations in adulthood (Edgell, 2006; Roof, 1999). Extrapolating from research findings that fathers tend to reduce their contact with children following a divorce (King, 1994), I might expect that marital disruption would principally interfere with the ability of fathers to influence the religious orientations of their children. In addition, parental divorce and remarriage may restrict access to grandchildren, particularly on the paternal side of the family, thereby inhibiting the religious influence of grandparents. The gender of “sending” and “receiving” generations may also pattern the intergenerational transmission process. Generally, religious continuity is stronger within same-sex relationships than it is within opposite-sex relationships, with the greatest religious congruence found between mothers and their daughters, (Aldous & Hill; 1965; Axinn & Thornton, 1993). Since mothers and daughters tend to be the intergenerational pair with the closest relationship, it follows that their communication around religious issues is stronger as well (Pearce & Axinn, 1998). However, intergenerational religious continuity is not restricted to female lineages, as mothers also influence their sons’ religiosity (Bao, Whitbeck, Hoyt & Conger, 1999). In addition, more religious fathers— in part because they are more involved with their children—influence the religiosity of their offspring to a greater degree than less religious fathers (Bartkowski & Xu, 2000; Clark, Worthington & Danser, 1988; King, 2003; Wilcox, 2002). Gender of grandparents and grandchildren may also play a role in religious transmission. As kin-keepers in the family, grandmothers are more likely than grandfathers to serve as confidants to their grandchildren and be involved with them in social, leisure, and religious activities (Silverstein & Marenco, 2001). Since 40 grandmothers are particularly close to their granddaughters (Hagestad, 1985) and women are generally more religious than men (Miller & Hoffman, 1995), I expect that grandmothers will have a stronger impact on the religious beliefs of their granddaughters than on the religious beliefs of their grandsons. In addition, the lineage of grandparents may be consequential to intergenerational religious transmission, as maternal grandparents—particularly maternal grandmothers— tend to see their grandchildren more often than do paternal grandparents, and report more satisfaction with the grandparenting role (Uhlenberg & Hammill, 1998; Somary & Stricker, 1998). Since the middle generation serve as gatekeepers to grandparents, it is not surprising that grandchildren have better access their mothers’ parents (particularly maternal grandmothers) than their fathers’ parents (Chan & Elder, 2000). Therefore, I expect maternal grandmothers will exert the greatest religious influence over their grandchildren. B. Research Questions Drawing on an expanded model of family socialization, I examine whether grandmothers influence the religiosity of their grandchildren independently of, and in conjunction with parents. I also examine whether the influence of grandmothers and parents is weaker in the context of parental divorce, and how grandchildren’s and parents’ gender, and grandmothers’ lineage condition the transmission process. I expect that the absorption of religiosity will be strongest in same-sex lineages, particularly those consisting of women. I focus my attention exclusively on grandmothers rather than grandfathers for theoretical and empirical reasons. Previous literature has emphasized and demonstrated 41 the important “kin-keeping” role of grandmothers and their crucial contributions to the well-being of their adult children’s families, especially when under stress (see Denham & Smith, 1989). Further, preliminary analyses of my data revealed substantially stronger correlations between grandmothers and grandchildren on almost all dimensions of religiosity compared to similar correlations between grandfathers and grandchildren, confirming my expectation of grandmothers’ exceptionalism. In summary, this investigation will address the following research questions: 1) Do grandmothers influence the religious beliefs of their grandchildren independently of their grandchildren’s parents? 2) Do grandmothers influence the religious beliefs of their grandchildren more strongly when parents are less religious (i.e., compensatory) or when parents are more religious (i.e., complimentary)? 3) Does parental divorce suppress the transmission of religious beliefs between parents and children, and between grandmothers and grandchildren? 4) Is the intergenerational transmission of religious beliefs stronger through women in the lineage, that is, from maternal grandmothers and mothers, and to granddaughters? C. Methods Sample Data for my analyses are from the Longitudinal Study of Generations (LSOG), a study of over 3,000 respondents ages 16-91, from 350 three- and four-generation families (for more details see Bengtson et al., 2002). Individuals eligible for sample inclusion 42 were generated from the families of grandparents randomly selected in 1970 from the membership of a large (840,000 member) prepaid health maintenance organization in the Los Angeles area. The sample pool was generally representative of white, economically stable and working class families. Self-administered questionnaires were mailed to the grandparents and their spouses (G1s), their adult children (G2s) and their grandchildren who were aged 16 or older (G3s). In 1985, 1,331 of the original sample were surveyed again, and since then data have been collected at three-year intervals through 2004. The response rate between 1971 and 1985 was 65%, and has averaged 74% between waves since then. Starting with the 1991 wave of data collection and continuing up to the present, great-grandchildren (G4s) Were accumulated into the study as they turned 16 years of age. In 2000, the response rate for G4s was 69%. At this wave, they averaged 23 years of age, ranging from 16 to 40. The analytic sample consists of participating G4 grandchildren in 2000 linked to their G3 mother and fathers in 1985, and their G2 grandmothers in 1971. The lineage of the grandmother is determined by the original sampling design and can be either on the maternal or paternal side of the family. The survey periods were intentionally chosen over a broad range of historical time to maximize the degree to which generations can be matched on age--a characteristic to which religiosity tends to be sensitive. This represents an advance over cross-sectional surveys because generations are better equated on the stage-of-life at which they are assessed. The eligible sample consisted of 565 grandchildren who matched to at least one participating parent and/or grandmother. Altogether, respondents derived from 208 common three-generation families. Grandchildren with participating siblings were paired 43 with the same parent(s) and/or grandmother. (I addressed non-independence across observations by statistically adjusting standard errors for clustering.) After excluding 34 grandchildren who had missing data on the dependent variable, the final sample was 529 grandchildren of whom 81%, had participating grandmothers, 72% had participating mothers, and 51% had participating fathers. Incomplete data due to non-response were imputed for all variables, except grandchildren’s religious beliefs, using the expectation maximization algorithm found in the statistical software SOLAS 3.0 (SOLAS, 2001). Under the assumption that incomplete data are conditionally missing at random based on observed variables, this approach produces unbiased estimates and is superior to relying on valid data only (Little & Rubin, 2002). Table 3.1 presents the number of valid observations for each independent variable. I note that means of variables based on imputed values are very close to those based solely on valid observations (not shown). Measures Dependent variable Religious beliefs were measured as the strength of agreement or disagreement with a series of statements conveying conservative religious attitudes and literal Biblical beliefs (Comrey & Newmeyer, 1965; see Glass, Bengtson & Dunham, 1986). Respondents evaluated the following statements: (1) “Every child should have religious instruction.” (2) “This country would be better off if religion had a greater influence in daily life.” (3) “God exists in the form as described in the Bible.” (4) “All people today are descendents of Adam and Eve.” Response categories were coded as follows: 1 = strongly disagree, 2 = disagree, 3 = agree, 4 = strongly agree. The dependent variable for 44 my analyses consisted of the summed responses to these four-items. The scale is highly reliable in the G4 generation, with Cronbach’s alpha = .92. Independent variables The same measures used to assess religiosity of grandchildren were also used to assess parents and grandmothers. Reliability coefficients for mothers, fathers, and grandmothers were .87, .86, and .82, respectively. To evaluate the strength of intergenerational transmission in religious beliefs and attitudes, I relied on the magnitude of correspondence in the summed scale between grandchildren and two generations of their elders. I controlled for several characteristics of grandchildren, including age in years, age in years squared (to capture non-linearity in life-stage effects), gender (0 = male; 1 = female), education (0 = no college experience; 1 = attended college); marital status (0= unmarried; 1=married), and parental status (0 = no children; 1 = has at least one child). In addition, I controlled for marital history of parents (0 = intact; 1= ever divorced), education of parents (0 = neither had college experience; 1 = at least one attended college), education of grandmothers (0 = no college experience; 1 = attended college), and the lineage through which grandchildren are related to their grandmothers (0 = paternal; 1 = maternal). D. Results Table 3.1 presents descriptive characteristics for each of the three generations represented in the sample. Agreement with conservative religious ideology clearly weakened across generations, particularly for grandchildren, demonstrating a sharp cohort or historical decline in religious conservatism from the 1970s to the end of the 45 1990s. Not coincidentally, this trend occurred alongside a rise in education across the generations, with 39 percent of grandmothers having completed at least some college compared to 58 percent of mothers and 66 percent of fathers. Although many grandchildren had not completed their education at the time of the survey, 60 percent had already completed some college by 2000. At this time, a sizable number of grandchildren had formed their own families, with 26 percent having married and 23 percent having had children of their own. Almost half (46%) of the grandchildren experienced the divorce of their parents. In the following tables I present multivariate models to test intergenerational sources of conservative religious beliefs in grandchildren. I present only standardized coefficients since, as partial correlations, they have desirable properties for gauging transmission intensity, ranging in magnitude from zero (no correspondence) to one (total correspondence). As there are no intercepts in the standardized model, cohort differences were not parameterized and transmission effects are interpreted as cross- generational mappings of religiosity relative to the central tendency of each generation. I built my equations hierarchically, first including only the characteristics of grandchildren. Second I added grandparent’s religious beliefs to determine the total effect of grandmothers on their grandchildren. Third, I added parent’s religious beliefs to identify the degree to which grandparent’s influence, if any, was mediated through the religious beliefs of parents. Finally, I added several interaction terms to test how parents and grandparents reinforced each other’s religious transmission, how parental divorce moderated the strength of transmission from parents and grandparents, how lineage moderated the strength with which grandmothers transmitted religious beliefs, and how 46 Table 3.1: Characteristics of Grandmothers, Parents, and Grandchildren, LSOG 1971-2000 Mean (SD) % Valid N Grandchildren (2000) Conservative religious beliefs 2.4 (.92) 519 Age 22.9 (4.98) 524 Female 51.5 291 Experienced parental divorce 45.7 503 Married 25.7 515 Has children 23.2 563 Has at least some college 59.6 558 Grandmothers (1971) Conservative religious beliefs 3.1 (.83) 472 Age 43 (4.53) 474 Has at least some college 39.4 471 Maternal lineage 60.0 474 Mothers (1985) Conservative religious beliefs 2.9 (.90) 401 Age 31.6 (3.94) 398 Has at least some college 58.2 366 Fathers (1985) Conservative religious beliefs 2.7 (.93) 288 Age 33.8 (2.46) 280 Has at least some college 65.8 266 gender of grandchild moderated the degree to which religious beliefs were absorbed from older generations. Table 3.2 shows the estimates regressing children’s religious beliefs on sequential blocks of grandchild, parent, and grandmother variables. Standard errors were adjusted for clustering of G4 siblings and their intergenerational triads (see Stata, 2002). Model 1 shows the effects of grandchildren’s characteristics. Results revealed that 47 granddaughters, married grandchildren, those with children of their own, and those whose parents did not divorce tended to have more conservative religious beliefs than their counterparts. Overall, grandchildren whose families of procreation and orientation were more traditional in structure tended to subscribe to more conservative religious beliefs. In Model 2, religious beliefs of grandmothers were introduced. I note that the coefficient is significant and, at .22, suggests a moderately strong resemblance between grandmothers and their grandchildren. When parents’ characteristics were added in Model 3, I found that both mothers and fathers exerted an influence on their children, with mothers the stronger of the two. However, while the coefficient for grandmothers was still significant, it was substantially reduced, suggesting that half the influence of grandmothers was mediated through the parent generation. Both linear and quadratic terms for age of grandchildren are significant in this equation. I found that conservative religious beliefs weakened from adolescence to early adulthood but began to strengthen with further maturation (for similar findings, see Argue, Johnson & White, 1999; Chaves, 1991; Stolzenberg, Blair-Loy &Waite, 1995). Post-hoc tests revealed that religious beliefs began strengthening at about the age of 30. While these age effects are not based on longitudinal data, they are nevertheless suggestive of a rebound in conservative religious beliefs over the adult life course. In Table 3.3 I tested for factors that may moderate the intergenerational transmission process. I introduced blocks of interaction variables separately in order to minimize multicollinearity associated with product terms. Model 1 included an interaction term between the religious beliefs of the two older generations. 48 Table 3.2: Standardized Regression Estimates Predicting Religiosity in Grandchildren (N = 529) Predictors Model 1 Model 2 Model 3 Grandchild Characteristics (2000) Granddaughter .10* .10* .09* Age -.46 -.74* -.58 + Age 2 .44 .71* .66 + Has at least some college -.02 .01 -.03 Married .20*** .19*** .13* Has children .15*** .15*** .14* Experienced parental divorce -.15*** -.10* -.10* Grandmother Characteristics (1971) Conservative religious beliefs .22*** .11*** Has at least some college -.10* -.08* Maternal lineage .00 .00 Parent Characteristics, 1985 Mother’s conservative religious beliefs .23*** Father’s conservative religious beliefs .17*** Mother and/or father has at least some college .06 Adjusted R 2 .10 .15 .27 +p<.10, *p<.05, **p<.01, ***p<.001 Note: Standard errors are adjusted for clustering. The positive interaction between the religious beliefs of mothers and grandmothers (p<. 10) suggests that each strengthened the influence of the other. In other words, grandparents transmitted their religiosity to grandchildren more effectively when mothers were more religious. In Model 2 I introduced variables that interacted parental divorce with religious beliefs of the two older generations. Results showed that divorce significantly interfered with the transmission of religious beliefs from mothers. When combined with the main 49 effect coefficient for mother’s religious beliefs (.34, not shown), the net influence of mothers under conditions of divorce was essentially nil. The same pattern with respect to divorce emerged for transmission from fathers and grandmothers, but the coefficients for these two interactions were not significant. Thus, the transmission of religiosity from mother to child appeared to be the intergenerational pathway most disrupted by parental divorce. In Model 3 I tested whether maternal grandmothers transmitted their religious beliefs more strongly than paternal grandmothers, but no significant interaction was found. This finding fails to confirm my expectation that maternal grandmothers would influence the religiosity of their grandchildren more strongly than paternal grandmothers. Finally, interactions entered in Model 4 tested whether the path of religious transmission varied by grandchild’s gender. The strength of religious transmission did not vary by parents’ or grandchild’s gender, although trends that were not significant suggested a pattern that favored transmission to daughters over sons. In contrast, the coefficient for grandmothers, at .34, in Model 3 indicates grandmothers transmitted their religious beliefs more strongly to granddaughters than to grandsons. Overall, the results in Table 3.3 provide support for the importance of gender, at least for grandparents and grandchildren, in religious transmission. However, maternal lineage did not show a similar pattern. 50 Table 3.3: Standardized Regression Estimates for Interactions Predicting Religiosity in Grandchildren (N = 529) Predictors Model 1 Model 2 Model 3 Model 4 Grandmother’s religious beliefs x mother’s religiosity .61 + Grandmother’s religious beliefs x father’s religiosity -.10 Mother’s religious beliefs x parental divorce -.44* Father’s religious beliefs x parental divorce -.19 Grandmother’s religious beliefs x parental divorce -.15 Grandmother’s religious beliefs x grandmother’s lineage .20 Mother’s religious beliefs x granddaughter .28 Father’s religious beliefs x granddaughter .18 Grandmother’s religious beliefs x granddaughter .34 + Adjusted R 2 .29 .27 .27 .29 +p<.10, *p<.05, **p<.01, ***p<.001 Note: All covariates found in Table 3.2 are controlled and standard errors are adjusted for clustering. E. Discussion The aim of this chapter was to learn more about how religious beliefs are passed down from older to younger generations in nuclear and extended families. Specifically I was concerned with the unique roles that grandparents and parents play in the transmission process, and how the intergenerational transmission of religiosity is conditioned by marital disruption and gender. General socialization theory has tended to focus almost exclusively on the influence of parents on children, and has paid little attention to the wider family context in which children are situated. In this chapter I expanded coverage of the theory to include grandparent-to-grandchild religious 51 transmission, as well as the family contexts that facilitate or inhibit intergenerational religious continuity. Using three decades of data from a sample of three-generation families, I examined four research questions, all of which received at least partial affirmation. With respect to my first question, my results confirm the importance of grandmothers as independent conveyers of religious beliefs to their grandchildren, albeit at a magnitude half of that of mothers. Notably, the detection of this influence is robust to the almost thirty year gap in measurement between generations and the fact many responding grandmothers represent only one of two possible grandmothers of the referent grandchild. The fact that grandparents were measured before some of their grandchildren were even born speaks to a deep, but perhaps unacknowledged, current of religious consistency that runs through generations in the family. This speaks to the legacy that grandmothers in particular leave for their grandchildren and counters the position of some social scientists who declare and decry the decline of the family’s influence over the moral development of children. In addressing my second research question, I detected an intergenerational synergy between grandmothers and parents. The influence of older generations on the religiosity of grandchildren is greatest when both grandmothers and mothers are consistently strong in their beliefs. This adds to my confidence that the family operates systemically in the socialization of children to religion, and as a social organization represents more than the sum of its parts. Indeed, it may be more accurate to characterize intergenerational resemblance as the manifestation of a family-wide religious orientation. 52 My third question explored whether parental divorce weakens the intergenerational transmission of religious beliefs. I found that divorce adversely affects religious transmission from mothers, but not from fathers and grandmothers. While at first glance this set of results may seem puzzling, it is possible that divorce has greater potential to disrupt the normally stronger ties that insure continuity between mothers and children; the influence of divorced fathers—because they tend to have lower contact with children—is paradoxically less affected by divorce. Further, research suggests that more religious fathers—both married and divorced—tend to have closer relationships with their children than less religious fathers (King, 2003). Thus, more religious fathers, because they have greater family commitment, may continue to be involved in the lives of their children following divorce and continue to religiously influence their children. The relevance of gender formed the basis for my fourth research question. Although I expected stronger religious transmission among same-sex parent-child pairs, I found no difference in the strength of religious transmission to sons and daughters. However, I found that religious transmission from grandmothers was stronger to granddaughters than it was to grandsons. Gender similarity may promote stronger bonds in these relationships, enhancing the influence of grandmothers. This leads us to speculate that it is primarily within more discretionary family affiliations—such as those between grandmothers and grandchildren—that gender similarity forms a basis for religious training and emulation. My fourth research question also addressed how lineage of grandmothers influenced religious transmission across three generations. Contrary to evidence that shows the importance of lineage in the quality of grandparent-grandchild relationships 53 (Chan & Elder, 1999), my findings reveal that lineage had no bearing on the capacity of grandmothers to transmit religious beliefs to their grandchildren. Perhaps lineage is more important in the context of complete gender congruity (i.e., grandmothers’ transmission to maternal granddaughters) or under conditions of parental divorce (i.e., grandmothers’ transmission to maternal grandchildren of divorce). However, sample restrictions prevent us from testing these hypotheses with the necessary three-way interactions. In addition, maternal-paternal effects in my analysis are necessarily based on comparisons across two sets of grandparents, each with only one type of grandchildren represented. The distinction in religious continuity within maternal and paternal relationships may best be seen in grandparents with both types of grandchildren. Although the availability of data on religiosity across three-linked generations is an important advantage of the chapter’s design, there are limitations to my investigation related to sample, measures, and modeling that deserve mention. The sample on which I performed my analyses is regional and caution is advised when generalizing my findings to the national population. As a relatively affluent sample that consists almost entirely of white non-Hispanic individuals, it provides little opportunity to inspect how transmission might vary by social class or ethnic group. In addition, attrition and non-response may have selected out of the sample those family members who had wide disagreements with their parents, causing us to overestimate the extent intergenerational continuity. A second limitation is that my measure of religious beliefs and values taps a fairly narrow dimension of religiosity, namely belief in the inerrancy and authority of the Bible and valuation of religion as a social good. Other dimensions of religiosity such as service attendance and identification with a denomination may show different patterns of 54 continuity and discontinuity across generations. For example, research shows that marital disruption may interrupt continuity in religious attendance and affiliation among those in religious sects that discourage divorce (Lawton & Bures, 2001; Sherkat, 1991; Stolzenberg, Blair-Loy & Waite, 1995;). My measure of religious beliefs and values— endorsement of a fundamental theological narrative and moral order—may be resistant to changes in religious affiliation and modes of expression within the Judeo-Christian tradition. A third limitation of note is that I do not directly observe the transmission process itself and rely on the degree of correspondence between generations as evidence of intergenerational transmission. There are likely to be direct (e.g. socialization) and indirect (e.g. shared environments) means of transmission, as well as potential genetic sources, all of which are not directly observed. In addition, I acknowledge that correspondence between generations may partially result from young adults influencing the beliefs of their parents and grandmothers. These mechanisms must remain objects of speculation and await testing in future studies. Beyond my more immediate focus on the intergenerational transmission of religion lay broader cultural debates over the potential demise or continued resilience and solidarity of the American family. The contemporary family is under constant moral surveillance by the public, leading to questions about whether changes in the form and functions of the family have seriously weakened its ability to pass down a cogent set of values and rules to the next generation. It is within this larger framework that the debate between idealists and pragmatists continues. While I cannot comment on whether there has been a historical decline in the influence of older generations, my results suggest that 55 their influence is still robust in contemporary families. The sharp decline in conservative religious ideology over three generations spanning the 20 th century may seem at odds with the intergenerational continuity I observed. Perhaps it is more accurate to say that young adults tend to retain the religious outlooks of their parents and grandparents relative to their shift as a cohort toward secularism and non-traditional forms of religious practice. Even as divorce looms as a threat to disrupt intergenerational continuity, I suggest that my findings can serve to revise current stereotypes about the demise of nuclear and extended families in socializing younger generations to religious beliefs and values. 56 Chapter 4: The Effects of Age and Life Course Transitions on Changes in Values Toward Individualism and Collectivism Among Young Adults A. Introduction The transition from adolescence to adulthood involves a variety of experiences in school and the labor force, as well as the assumption of multiple adult roles. Forming a romantic partnership and entrance into parenthood are two examples of role transitions that shape the life pathways of young adults, both in the context in which they are experienced and in their ability to shape strategies of action, or beliefs about how the world works (Swidler, 1986). Research on the effects of values concerning marriage, cohabitation and parenthood focuses on how these values shape family formation behaviors, such as the timing of marriage and parenthood, as well as the decision to cohabit with a partner (Axinn & Thornton, 1993; Barber, 2000; Clarkberg, Stolzenberg & Waite, 1995). These studies show values influence the timing and context of life course transitions. However, the reverse association is less frequently studied--how transitions in young adulthood influence values. The study of how life transitions influence values is an important research question, particularly because values are responsive to experiences and life course transitions over time. The current chapter has four specific aims. First, this chapter examines whether young adults are more or less individualistic than previous generations. Several studies have charted the rise in individualistic and materialistic values among young adults in the United States from the 1970’s to the late 1980’s (Bellah, Madsen, Sullivan, Swidler & Tipton, 1985; Crimmins, Easterlin & Saito, 1991; Easterlin & Crimmins, 1991). However, the preoccupation with individual growth and self-fulfillment may have slowed 57 in recent years. More recent studies have shown the trend toward increasing individualism has declined among young adults in the 1990’s (Bengtson, Biblarz & Roberts, 2002; Roberts & Bengtson, 1999). This chapter reevaluates this question with more recent data. Second, this chapter measures how age influences values. Measuring the independent effect of age on values changes is a useful way to determine whether values change as a result of developmental processes, such as aging, the experience of life course transitions, or both. Since life course transitions often occur at specific ages, the effects of age and life events are analyzed separately to better understand the process of values change over the adult life course. Third, this chapter measures the effects of parents’ divorce, parents’ education, parents’ own values and number of siblings on their children’s values over time. Last, my research examines five transitions in young adulthood—college completion, college enrollment, marriage, cohabitation and parenthood—and explores how each of these life transitions changes values toward individualism and collectivism over time. The fourth study aim is important for understanding how the occurrence of life events and the adoption of new social roles influence values over the adult life course. Values as a Driving Force of Demographic Change In this chapter, values are defined as “enduring beliefs that a specific mode of conduct is personally or socially preferable to an opposite or converse mode of conduct or end-state of existence” (Rokeach, 1973). Values are different from attitudes in that attitudes refer to a concrete social object, while values are more abstract and are often placed more centrally in concepts of the self (Hitlin & Pillavin, 2004). There is evidence 58 to suggest young adults’ demographic transitions today are less socially prescribed by the family and wider society, and are more a product of individual preferences (South, 2001). With these changes, there has been a shift in life goals among young adults, many whom desire jobs high in monetary rewards and prestige, rather than careers that further the public good (Easterlin & Crimmins, 1991; Crimmins, Easterlin & Saito, 1991). Decisions to cohabit and/or marry a partner are also informed by a broader set of competing values, such as the value placed on individual achievement compared to broader social concerns, such as family life (Lye & Waldron, 1997). Values reflect one’s life goals (Shanahan, 2000), and as such are important for understanding young adult transitions over the life course. In particular, values about marriage have a direct relationship to values about other domains of life. For example, young adults who value career advancement and leisure time are more likely to cohabit than to marry (Clarkberg, Stolzenberg & Waite, 1995). Thus cohabitation and singlehood are often chosen by people who wish to avoid the constraints and demands of marriage. The increase in the pursuit of personal goals separate from family commitments has been one of the arguments made by those touting the decline of the collective functions and roles provided by the family (Popenoe, 1988; 1993). However, most young adults expect to marry and highly value family life (Manning, Longmore & Giordano, 2007; Hill & Yeung, 1999). The current chapter explores if/how values change as a result of life course transitions, such as college completion, college enrollment, marriage, cohabitation and parenthood, keeping in mind that values are fluid and can be reconstituted as one ages, and as one experiences life events within a particular time period. 59 Although young adults’ values are influenced by their own life transitions, these values are first shaped by family experiences in childhood and adolescence. Studies have shown young adults whose parents divorce have less positive attitudes about marriage that children whose parents are continuously married (Axinn & Thornton, 1996). Parents’ marital transitions have a strong influence on their children’s attitudes and values toward marriage which influences children’s values and union formation behaviors as young adults (Amato & Booth, 1991; Cherlin, Kiernan & Chase-Lansdale, 1995; Cunningham & Thornton, 2006). Parental divorce is included in these analyses to assess the extent to which young adults’ values toward individualism/collectivism are shaped by parents’ own marital transitions. Theories Used to Explain Attitudinal/Value Changes Over Time Shifts in individual values over time can be attributed to two processes: developmental changes and role transitions. The ‘impressionable years’ hypothesis is often used to explain values changes as a result of developmental changes, such as aging (Alwin & Krosnick, 1991). This theory states individual’s attitudes are malleable in young adulthood but become increasingly stable as people age. The majority of studies in this area measure sociopolitical, gender and sexual attitudes to determine whether changes are due to cohort succession or intracohort changes (Danigelis, Hardy & Cutler, 2007; Brooks & Bozendahl, 2004; Treas, 2002). Although findings vary depending on the attitude considered, most studies find changes are due, in large part, to cohort succession: the movement of younger cohorts with attitudes rooted in distinct historical context replacing the attitudes of older cohorts (Ryder, 1965). The receptivity of young adults to their environment makes them particularly likely to change their values, 60 compared to older adults who are more resistant to change (Alwin, 2002; but see Danigelis et al, 2007). The role transitions experienced by young adults are another avenue through which young adults experience values change. A large number of sociological studies have focused on the strategies with which people cope with role overload or the conflict and strain that ensues when roles compete for one’s energy and time (Staff & Mortimer, 2007; Eggebeen & Knoester, 2001; Moen & Yu, 2000; Gerson, 1985). The adoption of adult roles, such as marriage and parenthood, results in a change in one’s identity, depending on the salience of the particular role (Thoits, 1992). Similarly, marriage and parenthood, both independently and jointly, involve a reassessment of one’s values. As my findings will show, young adults who marry and/or have children develop more collectivistic values as a result of these life transitions. In contrast, those who cohabit often have a different set of values compared to the married. On average, young adults who cohabit have less traditional attitudes toward family life. For example, cohabitors have more accepting attitudes toward divorce and have more egalitarian gender attitudes (Axinn & Thornton, 1996; Clarkberg, Stolzenberg & Waite, 1995; Cunningham, Beutel, Barber & Thornton, 2005; Thornton, Axinn & Hill, 1992). Young adults who cohabit are also less religious, often because religious institutions discourage cohabitation (Stolzenberg, Blair-Loy & Waite, 1995) and because young adulthood is a period of less religious conviction and participation (Uecker, Regnerus & Vaaler, 2007). Hence it is likely cohabitors have higher individualistic values because they have fewer family commitments and are less connected with 61 religious institutions. This chapter tests whether young adults who cohabit without children are more individualistic than those who marry and/or have children. B. Research Questions The Rokeach Value Survey (1973) is a useful tool to explore value hierarchies and their perceived importance over the life course. The applicability of this survey has been explored, both for its measurement qualities and for its larger contribution to the study of values (Hitlin & Pilavin, 2004; Spates, 1983). Although the survey has been cited widely in social psychology and in cross-cultural studies, no sociological study to date has used this survey to measure values changes in young adulthood in the United States. The Longitudinal Study of Generations (LSOG) is used to explore these questions, and its multigenerational and longitudinal design is particularly useful for the study of values changes over time. Specifically, this chapter addresses four research questions: 1) Do young adults rank individualistic values as more important than their parents at similar ages? 2) How do young adults’ values toward individualism/collectivism change with age? 3) How do parental divorce, parents’ education and parents’ own values influence young adults’ values toward individualism/collectivism? 4) How do college degree attainment, college enrollment, marriage, cohabitation and parenthood influence young adults’ values toward individualism/collectivism? 62 C. Methods Sample The Longitudinal Study of Generations (LSOG) is a study of over 3,000 respondents ages 16-91, from 350 three- and four-generation families (Bengtson et al., 2002). Individuals eligible for sample inclusion were generated from the families of grandparents randomly selected in 1970 from the membership of a large (840,000 member) prepaid health maintenance organization in the Los Angeles area. The sample pool was generally representative of white, economically stable and working class families. Self-administered questionnaires were mailed to the grandparents and their spouses (G1s), their adult children (G2s) and their grandchildren who were aged 16 or older (G3s). In 1985, 1,331 of the original sample were surveyed again. Since 1991, data have been collected at three-year intervals: 1991, 1994, 1997 and 2000. Starting with the 1991 wave of data collection and continuing up to the present, great- grandchildren (G4s) were accumulated into the study as they turned 16 years of age. The response rate between 1971 and 1985 was 65%, and has averaged 74% between waves since then. In 2000, the response rate for G4s was 69%. There are 839 young adults who were interviewed in at least one of the four waves since 1991. There are two sample criteria by which the subsample for the analyses was selected. My definition of young adulthood includes individuals between the ages of 16 and 35, referred to in the study as G4s. This age range follows closely other definitions of young adulthood which define this period of life between 18 to 30 years of age (Rindfuss, 1991). Nine individuals did not meet these criteria and were deleted from the sample. Second, only those young adults who participated in at least two waves of data 63 can be included in the sample. This is an important distinction because the main research aim of the study is to measure changes in values over time which requires at least two waves of data. Thirteen percent of the sample had participated in all four waves, 20 percent had participated in at least three waves, 24 percent had participated in at least two waves and 43 percent had participated in only one wave. Out of the 839 young adults who participated in at least one of these waves, 479 participated in at least two waves. The sample for analyses was further reduced to 451 once taking into account those young adults’ who did not fill out the Rokeach Value Survey correctly (N=28). Measures Dependent Variable Eight items from the Rokeach Value Survey were used in these analyses, a collection of 18 terminal values which respondents rank in order of importance or desirability in life (Rokeach, 1973). This chapter is concerned with values which reflect individualism and collectivism. These scales each include four items and have been tested in previous studies using these data (Bengtson, Biblarz & Roberts, 2002; Bengtson & Roberts, 1999). The individualism scale includes: 1) an exciting life (novelty, adventure), 2) a sense of accomplishment (achievement), 3) personal freedom (independence, free choice, autonomy) and skill (being good at something you enjoy doing). The collectivism scale includes: 1) religious participation (working with others in your own church or organization); 2) friendship (meaningful relations with others who really care); 3) loyalty to your own (family and loved ones, church or group); and 4) patriotism (working for my country). The first two items in each scale were ranked in accordance to importance in life; the last two were ranked according to desirability in 64 life. These eight values were asked at each time wave. Phrases in parentheses were included alongside each value item to aid in the respondent’s understanding of its meaning. The purpose of these analyses is to determine the effects of age and life events on individualism and collectivism. For these analyses, these eight items were coded where positive coefficients represent a higher ranking of individualistic values and negative coefficients indicate a higher ranking of collectivistic values. Two additional items were added to the Rokeach scale in 1997 and 2000: career advancement (achieving success in your job or profession) and family life (working for the well-being of family members). These items were not included in the analyses, but the range of the scales varies over time with the addition of these items. The range of these values is -8 to 8 in 1991 and 1994 and -9 to 9 in 1997 and 2000. To make these scales comparable over time, the scale values of “8” and “9” in 1997 and 2000 were collapsed to correspond with the values scale in earlier waves. The five Rokeach value items measured of young adults and their parents are time-varying. This means that respondents receive a value for each wave in which they participated. Thus the dependent variable is interpreted as the change in individualism/collectivism associated with the change in the independent variables over time. The reliability coefficients for the individualism scale in 1991, 1994, 1997 and 2000 are .96, .93, .86 and .87, respectively. The reliability coefficients for the collectivism scale are .89, .86, .77 and .78 during the same years. Table 4.1 shows a correlation matrix for individualism and collectivism items in 2000 to illustrate the association among these items. 65 Table 4.1: Correlation Matrix for Items Included in the Individualism and Collectivism Scales Among Young Adults Ages 16-35, LSOG 2000 (N=451)† 1 2 3 4 5 6 7 8 1) An exciting life 1 .56 .55 .56 -.02 .52 .43 .12 2) A sense of accomplishment .56 1 .67 .77 .13 .65 .62 .32 3) Personal freedom .55 .67 1 .66 .22 .74 .62 .34 4) Skill .56 .77 .66 1 .27 .73 .63 .31 5) Religious participation -.02 .13 .22 .27 1 .29 .47 .43 6) Friendship .52 .65 .74 .73 .29 1 .75 .75 7) Loyalty to your own .43 .62 .62 .63 .47 .75 1 .40 8) Patriotism .12 .32 .34 .31 .43 .40 .47 1 †Each of these items is significantly different from 0 except for the correlation between "an exciting life" and "religious participation" Independent Variables Family background characteristics are measured at the earliest wave in which these data are available. For G4s, the earliest wave is 1991 and for their parents, the G3s, it is 1985. Family background variables include whether at least one parent received a college degree (=1). To assess family structure in childhood, the question, “Did your parents ever divorce?” (1=Yes), asked of G4s, was also included. Another variable that reflects family composition is the number of siblings reported by the focal child. Parents’ values on individualism/collectivism scales are also measured from the earliest wave available. This variable is coded in the same way as young adults’ individualism and collectivism scores. Variables which reflect young adult transitions include: 1) received college degree (=1); and 2) enrolled in school (but no terminal degree). Last, three measures of family formation were included, whether the G4 married (=1), cohabited (=1) or had a child (=1). Young adult transitions variables--college degree attainment, college enrollment, marriage, cohabitation and parenthood--are time-varying. This 66 means the effects of the independent variable on the dependent variable depend on whether an event occurred during a particular time interval. For example, young adults who reported being single in 1991 and married in 1994 would be coded as “0” in 1991 and “1” in 1994. The Timing of Life Transitions Using the LSOG The timing and sequence of such events are important when assessing values change. Date of current union formation was available, as well as the birth year of first and subsequent children born to young adults in the study. However, restricting the sample to those whose unions and/or births occurred between 1991 and 2000 further reduced the sample by 44 cases (N=407). This sample restriction was deemed unnecessary because young adults are only asked about their current union status and not the dates of entry and exit from each union experienced. Further, the date of college graduation was not asked in any time wave. Thus the temporality of whether life course events results in values changes, or changes in values results in the occurrence life course transitions, cannot be fully assessed in the current study. Plan of Analysis Fixed and random effects pooled time-series models are used to test the relationship between age and values toward individualism/collectivism. In a pooled time- series, respondents receive four records, one for each time of measurement. Pooled time- series techniques are necessary for this particular model so the effects of time-varying variables can be estimated. Random effects models are preferable to regression models with lagged dependent variables because age and period effects can be estimated in the same model (Allison, 1994). The random effects models assume the unobserved 67 variation between individuals which is constant over time are random variables, as shown in the equation: Y it =u + b it + a i + e it where u is an overall constant, a i is the constant effect for individual i, b is the effect of change on individual i in time period t, and e it is a within-individual error term. Random effect models are often more efficient than fixed effects models because time-invariant variables (gender, birth cohort) can be included. Fixed effect models hold the change between individuals over time constant, such that a + (Y it –Y i )=b(X it – X i ) + e it , where Y i and X i are means for individuals i across t waves. Because both the independent and dependent variables are expressed as the deviation from the individual’s mean scores, they sum to zero for each individual and the constant term a is dropped from the equation. Although the fixed effects approach cannot include time-invariant variables, its ability to control for differences within individuals that often cannot be measured, such as personality characteristics or biological and genetic differences make this approach particularly useful (Johnson, 1995). The choice to use either random and/or fixed effect models depends on whether time-invariant variables need to be explicitly measured and whether the model is correctly specified. Both random and fixed effects models are shown here to compare differences between model estimates. D. Results Table 4.2 displays the descriptive characteristics for variables used in the analyses. With the exception of parents’ education, parental divorce, parents’ value on individualism/collectivism, number of siblings and young adults’ gender, these variables are time-varying. In other words, the mean represents the average value across all time waves from which the data were available. Table 4.2 also provides a comparison of 68 parents’ and children’s individualism/collectivism scores. Parents’ mean individualism/collectivism score is 1.37, which is lower than their children’s score of 2.84. A t-test on this item shows the mean difference to be significant, suggesting young adults have significantly higher individualistic values than their parents over time (t=-4.91). Both parents and their children are highly educated. Approximately 45 percent of parents are college graduates and almost one-third of young adults receive a college degree by the end of the study period. About 60 percent of young adults were ever enrolled in college. Over half of these young adults also had experienced parental divorce. On average, young adults grew up in small families with one or two siblings. Union and parenthood transitions are common among the sample. About 44 percent of these young adults had ever been married and 23 percent had ever cohabited. About 36 percent ever had children. Table 4.3 shows the breakdown of marriage, cohabiting, childbearing and singlehood among young adults. Given the three-year intervals between surveys, it is likely, in some cases, a relationship may begin and end during the interval that is not captured here. However, this table provides a useful description of the types of transitions formed among young adults, as well as the configuration of these unions for a small sample of young adults. Although most young adults report marriage and no cohabitation, about 20 percent of the sample reported cohabiting and marrying a romantic partner. The majority of married young adults report having children, but over one-third do not. The large number of married people without children is most likely a reflection of marriage in its early years before children are born. 69 Table 4.2: Means, Standard Deviations and Descriptions of Variables Used in Random and Fixed Effects Models Among Young Adults Ages 16-35, LSOG (N=479) Description Mean Valid N Parents' Characteristics Parent individualism/ Parents' individualism/collectivism score a 1.37 455 collectivism score (7.37) College degree or higher At least one parent received college .45 462 degree or higher (.50) Ever divorced Asked of G4s: "Did your parents ever .53 479 divorce?" (1=Yes) (.50) Young Adult Characteristics Individualism/collectivism Young adults' individualism/ 3.19 451 score collectivism score a (8.15) Gender 1=Female .58 479 (.49) Age Mean age 22.5 479 Range is 16-35 (4.54) Number of siblings Reported number of siblings 1.98 479 Range is 0-13 (1.53) College degree or higher Received a college degree or higher .28 470 by end of the study period (.45) Enrolled in college b Enrolled in college, no degree .59 478 by end of the study period (.49) Married b Married by end of study period .44 473 (.50) Cohabited b Cohabited by end of study period .23 473 (.42) Have children b Have children by end of study period .36 474 (.48) a The range for the Rokeach items is -32 to 32. b These variables are presented here as the proportion of young adults ever experiencing these events. Standard deviation in parentheses 70 Although many married couples do not yet have children, childbearing most often takes place within marital unions in this particular sample. Looking at young adults who cohabit with no reported marriages shows only a small number ever having children (N=18). Finally, this table shows there are a large number of single young adults. Over forty percent of the total sample does not report any union during the study period. However, 6 percent of these single young adults report having children. For these individuals, it may be the timing of the survey may have missed the short-lived romantic unions where children were present. Table 4.3: Descriptive Statistics of Married, Cohabitating and Young Adults with Children, Ages 16-35, LSOG (N=451) Mean N Marriage Reported .42 190 No cohabitation reported .81 154 Cohabited .19 36 No children reported .33 62 Has children .67 128 No Marriage Reported, Cohabited .14 65 Cohabited, no children .72 47 Cohabited, has children .28 18 No Union Reported .43 196 No children reported .94 184 Has children .06 12 The first research question asks whether young adults rank individualistic values higher than their parents at similar ages. Comparing parents in 1971 and young adults in 2000 at age 20 shows some interesting differences. By holding age constant, the differences in ranked values can be compared across generations and over time. Figure 71 4.1 shows how the four items which make up the individualism scale are ranked between parents and their offspring. Each item on the individualism scale is significantly different between parents and young adults, though to varying degrees. Possessing a skill for something you enjoy doing shows the largest difference, with young adults ranking this as more desirable than their parents (t=9.96; p<.01). Having an exciting life is also ranked as significantly more important among young adults compared to their parents (t=5.90; p<.01). Similarly, personal freedom is ranked as more desirable among young adults (t=2.32; p<.05), followed by accomplishment where there is a slight increase in importance among young adults (t=1.91; p<.10). Figure 4.2 displays the differences in ranking of the four items that comprise the collectivism scale. Interestingly, loyalty to your family and/or community is ranked as 1 2 3 4 5 6 7 8 Skill Exciting life Personal freedom Accomplishment Individualism Items Average Ranking G3 Parent (N=68) G4 Child (N=40) Figure 4.1: A Comparison of G3 Parents and G4 Children’s Individualism Values at Age 20 in 1971 and 2000, LSOG 72 significantly more desirable among young adults compared to their parents thirty years earlier (t=9.72; p<.01). This finding replicates an earlier study using these data comparing parents in 1971 with their children in 2000 at average age 19 (Bengtson, Biblarz & Roberts, 2002; p. 44). The value attributed to the importance of religious participation, on the other hand, declined significantly among young adults compared to their parents (t=2.27; p<.05). There are no significant differences on the ranking of friendship and patriotism between parents and their offspring. There are two main findings attributed to these two figures. First, young adults possess values that are distinctly more individualistic than their parents at similar ages. An exciting life and the development of a skill for something one enjoys are important to young adults, as is the autonomy to pursue one’s interests. Despite the rise in self- oriented values, however, young adults are ‘turning back’ to family commitment and Figure 4.2: A Comparison of G3 Parents and G4 Children's Collectivism Values at Age 20 in 1971 and 2000, LSOG 1 2 3 4 5 6 7 8 Loyalty to your own Religious participation Friendship Patriotism Collectivism Items Average Ranking G3 Parent (N=68) G4 Child (N=40) 73 loyalty. Hence the rise in individualist values among young adults does not reduce the extent to which young adults’ value family life. In line with the second research question, age variables were included in both random and fixed effects models to determine the extent to which values toward individualism/collectivism change with age and over time. The random effects model (Table 4.4) indicates young adults become less individualistic—and more collectivistic— over time. This trend is most significant for women whose values toward collectivism change at a faster rate than do men’s values. This finding corroborates other research which shows women possess values with a higher collectivistic sentiment compared to men (Beutel & Marini, 1995). Gender is not controlled in the fixed effects model because it is a constant (Table 4.5). However, the significance of age in the fixed effects model suggests significant within-person change in individualistic and collectivistic values over time. The third research question asks how parents’ own individualistic/collectivistic values, parental divorce, parents’ college education and number of siblings influences changes in young adults’ individualistic/collectivistic values over time. Table 4.6 includes these background characteristics in a random effects model. In the random effects model, it is important to remember these coefficients take into account the variation between- and within-individuals. Similarly, the differences across groups (e.g. young adults’ whose parents divorce vs. those whose parents do not divorce) are assumed to be uncorrelated with the other covariates in the model (Johnson, 1995). This table 74 Table 4.4: Random Effects Regression Model Relating Age and Gender to the Change in Individualistic/Collectivistic Values Among Young Adults, Ages 16-35, LSOG M1 Age -.16 *** (.05) Female -1.45 ** (.71) Constant 7.50 Person-years 1,804 N 451 Standard errors in parentheses ***p<.001, **p<.05, *p<.10 Table 4.5: Fixed Effects Regresion Model Relating Age to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG M1 Age -.14 ** (.06) Constant 6.06 Person-years 1,804 N 451 Standard errors in parentheses ***p<.001, **p<.05, *p<.10 presents four important findings. First, parents with a high average value of individualism are likely to instill the same values in their children. Thus parents who rank individualistic values highly also have children who rank individualistic values as more important over time. 75 Table 4.6: Random Effects Regression Model Relating Age, Gender and Parental Characteristics to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG Age -.19 *** (.05) Female -1.84 *** (.67) Parents' individualistic/ .30 *** collectivistic value (.05) Parents' divorce 1.19 * (.69) At least one parent has college degree .60 (.67) Number of siblings -.59 ** (.22) Constant 8.38 Person-years 1, 804 N 451 Standard errors in parentheses ***p<.001, **p<.05, *p<.10 The second important finding is young adults whose parents divorce and those whose parents are continuously married have significantly different patterns of value change. Young adults whose parents divorce develop more individualistic values over time. Third, young adults who grew up in larger families develop higher collectivistic values. Fourth, age and gender remain important predictors of values changes with the addition of these family background variables. This indicates young adults rank collectivistic values higher with age, after controlling for characteristics of their family of origin. This finding provides evidence that values change in young adulthood as a result 76 of developmental processes. The next step is to see whether age remains a significant predictor of value change once adding young adults’ life course transitions to the model. Table 4.7 is a random effects model which includes five life events--college degree attainment, college enrollment, marriage, cohabitation and parenthood--to determine the extent to which these life transitions change young adults’ values. This table reveals six important findings. As shown in Model 1, young adults who are enrolled in college develop higher individualism scores over time. These findings make sense in light of three of the four values included in the individualism scale: personal accomplishment, freedom and skill. It is likely college-enrolled young adults would place high value on their accomplishments and skills. Similarly, young adults enrolled in college would also highly value the autonomy provided by higher education. This finding suggests college education increases young adults’ individualistic values while holding age constant. The second finding of this table is displayed in Model 1. Married young adults develop lower scores toward individualism—and higher collectivism scores—than the non-married. A similar finding is evident for young adults who have children, as shown in Model 2. Young adults who have children develop more collectivistic values than young adults without children. This finding suggests marriage and parenthood induces a shift in values among young adults from individualism to collectivism. It is important to distinguish whether values changes are due to the life event itself, or the age at which it is experienced. Young adults who marry are on average five years older than the non-married. Similarly, young adults with children are six years older than those without children. In order to test whether values change was due to age 77 Table 4.7: Random Effects Regression Models Relating Age, Gender, Parents' Characteristics and Young Adults' Life Course Transitions to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG M1 M2 M3 M4 M5 Parents' individualistic/ .30 *** .30 *** .29 *** .30 *** .30 *** collectivistic value (.05) (.05) (.05) (.05) (.05) Parents ever divorce 1.26 * 1.33 * 1.27 * 1.23 * 1.16 * (.66) (.68) (.67) (.66) (.66) At least one parent has college degree .08 .17 .24 .14 .07 (.66) (.68) (.67) .(.68) (.66) Young adult characteristics Age -.09 -.19 *** -.20 *** -.11 -.08 (.06) (.06) (.06) (.06) (.07) Female -1.83 *** -1.80 *** -1.82 *** -1.85 *** -1.80 *** (.65) (.66) (.65) (.65) (.65) College degree or higher a 1.00 .99 .94 .96 .83 (.70) (.70) (.69) (.70) (.70) Enrolled in college a .82 * .94 * .89 * .80 * .76 * (.45) (.45) (.45) (.45) (.45) Number of siblings -.48 ** -.54 ** -.52 ** -.48 ** -.47 ** (.21) (.22) (.21) (.21) (.21) Married a -2.74 *** -2.46 *** -3.01 *** (.62) (.64) (.71) Have children a -.95 * -.37 -1.79 ** (.49) (.51) (.97) Cohabited a 2.07 *** 1.41 ** 1.51 ** (.70) (.72) (.72) Married * Have children a -1.87 * (1.07) Constant 6.31 7.97 8.06 6.55 6.13 Person-years 1,804 1,804 1,804 1,804 1,804 N 451 451 451 451 451 a These variables are time-varying. Standard errors in parentheses ***p<.001, **p<.05, *p<.10 78 or life events, interactions between marriage, parenthood and age were conducted for each model (not shown). The interaction between age and marriage is not significant. Similarly, the interaction between age and parenthood is not significant. This suggests the experience of marriage and parenthood on individualistic/collectivistic values is due to the experience of marriage and having a child, not the age at which the transition occurs. The third finding is presented in Model 3. Cohabitation increases young adults’ values toward individualism relative to collectivism. Although cohabitors may be more individualistic before the inception of such unions, it is also likely the experience of cohabitation engenders a change in young adults’ values. This finding falls in line with other research which shows cohabitors have higher individualistic values, particularly concerning the importance of career success and personal autonomy, than the married (Casper & Sayer, 2000; Clarkberg, Stolzenberg & Waite, 1995). The fourth important finding of this table is that marriage and parenthood have separate and joint effects on young adults’ values. The combination of marriage and parenthood often results in the reorganization of attitudes and values in order to deal with the stresses incurred by these multiple social roles (Thoits, 1992). Model 4 includes marriage, parenthood and cohabitation in the same model to test the independent effects of each transition while holding the other transitions constant. This model shows marriage increase collectivistic values relative to individualistic values. However, the effect of parenthood on young adults’ values does not reach statistical significance. This is due to the high correlation between marriage and parenthood in the sample (.52). Model 4 also shows the age coefficient is reduced to non-significance when including 79 both marriage and parenthood as main effects in the model. In this case, the transitions into marriage and parenthood increase collectivistic values to a greater degree than the independent effect of age. Model 5 includes an interaction term between marriage and parenthood to gauge how the combination of these social roles influences values. This interaction is significant, suggesting value toward collectivism increases when both of these roles are assumed. In sum, the independent effects of marriage and parenthood, as well as the combination of these roles, increases collectivistic values among young adults. Whether they occur separately or concurrently, marriage and parenthood are stronger predictors of values change among young adults than age. The fifth important finding of this table is parents’ own values toward individualism/collectivism influence young adults’ values after adding controls for young adults’ life transitions. These results suggest parents’ values are important in directing young adults’ values toward greater individualism. The significance of parents’ own values on that of their children after including life course transitions suggests the intergenerational transmission of values influences the change in young adults’ values over time. The sixth important finding of Table 4.7 is parental divorce increases individualistic values among young adults. In line with research indicating children of divorce have less positive attitudes about marriage and childbearing than children with continuously married parents (Amato & Booth, 1991; Cunningham & Thornton, 2006; Kapinus, 2004), these results suggest young adults who experience parental divorce 80 develop more individualistic—and less collectivistic—values than young adults whose parents have not divorced. Table 4.8 is a fixed effects model with the same variables included in Table 4.7, except for time-invariant variables. With the exception of cohabitation, these models confirm there is significant within-individual variation in values as young adults’ marry and/or have children. These results suggest the change in values toward collectivism over time is not due to differences between groups, such as the married and non-married, but rather due to within-individual changes in values over the life course. Random effects models are often preferred over fixed effects because they can include time- invariant predictors. However, there are no clearly defined criteria for choosing one method over the other in family research. The Hausman (1978) chi-square test was conducted to test differences between the coefficients in Tables 4.7 and 4.8. The chi- square value with 6 degrees of freedom (χ=-26.13) suggests there is a significant difference between the coefficients in these models, in which case the fixed effects estimates may be more appropriate (Johnson, 1995). In this case, there is sufficient evidence to suggest within-person variation on individualistic/collectivistic values over time as a result of marriage and parenthood among young adults. E. Discussion To address the first research question, parents’ values in 1971 and young adults’ values in 2000 were compared to assess whether young adults’ values were more/less individualistic over time by holding age constant. These findings revealed young adults’ ranked individualistic values higher than their parents at similar ages, but also reported “loyalty to one’s own” as more desirable in life. This finding suggests young adults’ 81 Table 4.8: Fixed Effects Regression Models Relating Age, Gender and Young Adult Life Course Transitions to the Change in Individualistic/Collectivistic Values Among Young Adults Ages 16-35, LSOG M1 M2 M3 M4 M5 Age -.07 -.13 ** -.15 ** -.06 -.03 (.08) (.08) (.07) (.08) (.08) College degree or higher .38 .37 .39 .37 .23 (.84) (.84) (.84) (.84) (.84) Enrolled in college .94 * 1.00 * 1.00 * .93 * .89 * (.51) (.51) (.51) (.51) (.51) Married -1.72 ** -1.66 ** -2.31 ** (.68) (.72) (.81) Have children -.29 .03 -1.87 (.56) (.57) (1.17) Cohabited .82 .38 .50 (.79) (.81) (.81) Married * Have children -2.26 * (1.25) Constant 4.46 5.56 5.80 4.55 3.88 Person-years 1,804 1,804 1,804 1,804 1,804 N 451 451 451 451 451 Standard errors in parentheses ***p<.001, **p<.05, *p<.10 have higher individualistic sentiments than their parents, but still highly value commitment to family and community. The second research question asked how values change in relation to age. Random and fixed effects models that included age and period effects were tested. These models confirm young adults become more collectivistic as they age. This finding suggests individualistic/collectivistic values are highly sensitive to developmental changes over the life course. The third research question focused on the effects of family background, namely parents’ divorce, education and parents’ own values, on young adults’ values. Parents’ 82 individualistic/collectivistic values had a significant influence on young adults’ values toward individualism, even after controlling for life course transitions. Further, random effects models (Tables 4.6 and 4.7) show young adults whose parents do and do not divorce have different patterns of values change. In this case, young adults whose parents ever divorce experience an increase in their individualistic values to a greater degree than young adults with continuously married parents. These findings suggest socialization within families, whether through parents’ own values or marital behaviors, influences young adults’ values over time. The fourth research question was to determine whether values change as a result of life course transitions such as college graduation, college enrollment, marriage, cohabitation and parenthood, independently of age. With the exception of college graduation, the effects of these transitions on the change in values over time were significant. These findings showed young adults enrolled in college develop stronger values toward individualism over time. Higher education increases individualistic values because college provides a venue for self-exploration and skill development. Also, college-educated individuals are often afforded more freedom in their personal lives and in the careers they choose. As shown in the random and fixed effects models, the effect of college enrollment on values change is a result of between-individual (difference between those with and without a college degree) and within-individual variation. Not surprisingly, marriage and parenthood increase values toward collectivism. The acquisition of these social roles, both independently and jointly, shifts young adults’ values from self-oriented values to values which emphasize involvement with one’s family and community. The effect of age in increasing collectivistic values was not 83 significant once adding young adults’ marriage and parenthood in the same model (Tables 4.7 and 4.8, Models 4 & 5). In this case, the significant life events that characterize adulthood, such as marriage and parenthood, had larger effects on value change relative to age. In contrast to the effects of marriage and parenthood, the experience of cohabitation increases individualistic values. This finding suggests cohabitors and married individuals, with or without children, have a different set of values, which has been shown in other research (Clarkberg, Stolzenberg & Waite, 1995). These differences are due, in part, to selection effects, as well as the social contexts in which these attitudes are formed (Cunningham, Beutel, Barber & Thornton, 2005). The non-significance of cohabitation on values change in the fixed effects model is a further indication the shift in values toward individualism is due to differences between those who have cohabited and not cohabited, not due to within-person change. Although the intergenerational linkages and longitudinal design of the LSOG makes it particularly useful for the study of values change, there are several limitations of the study’s sample and measures that deserve mention. First, the study of values change requires at least two waves of data, which restricted the sample to 479 young adults. The sample was further reduced to 451 cases because approximately 28 young adults did not fill out the Rokeach Value Survey correctly. The small sample requires a cautionary interpretation of the results. The small sample also did not allow for separate models to be run by gender. Moreover, the sample is comprised of white young adults from mainly working and middle-class families. Thus these results are not generalizable to a nationally representative population. It is likely the occurrence of young adults’ 84 transitions have different effects on values change by race/ethnicity that cannot be measured here. This study provides a first look at how values are shaped by the new experiences and social roles adopted in young adulthood. Collectivistic values among young adults increase in importance as they age, but are also dependent on whether they graduate college, as well as the types of relationships they enter. Young adults’ values are also embedded within a family context. Transmission of values between parents and their children remains strong, even after taking into account variation in values change between and within-individuals over time. The receptivity of young adults to new ideas about the self in relation to the social world is evident in this particular sample. The fluid and responsive nature of values to social contexts is important to remember when studying the life course transitions characterized by young adulthood. 85 Chapter 5: The Effects of Parents’ Education on the Union Formation Process of Young Adults: Are There Differences by Gender and Union Type? A. Introduction Fifty years ago, the decision to marry a romantic partner was often proscribed by the characteristics of one’s family of origin. Many young adults chose to marry a person who matched their parents’ preferences for a suitable life partner. Parents’ education and income were status markers by which these union formation decisions were made. Mate selection also followed religious and cultural norms as well (Kalmijn, 1994). A half- century later, the ascriptive processes that defined the search for a spouse have widely dissipated. Today, marriage is increasingly seen as a personal decision, rather than influenced by normative family pressures (Cherlin, 2004). In particular, the rise in cohabitation has altered the timing and sequence of young adult life transitions. Many young adults are choosing to cohabit with one or several romantic partners (Casper & Bianchi, 2002). Experiences in school and the labor force are tightly coupled with these union formation decisions (Guzzo, 2006). Young adults may delay marriage and childbearing until finishing school and obtaining a career. Others balance work and family responsibilities which requires juggling multiple life roles (Sayer, Cohen & Casper, 2004). Rather than follow an age-graded normative path, the life course today is increasingly guided by individual choices and preferences (Lesthaeghe, 1983). The current chapter examines the effects of ascribed family background--namely parents’ education--on their children’s union formation relative to their school experiences. 86 Past research has estimated the effects of young adults’ education on marriage and cohabitation without taking into account how the intergenerational transmission of educational attainment between parents and children influences union formation (Thornton, Axinn & Teachman, 1995; but see Axinn & Thornton, 1992). There are three central contributions of my analyses. First, I measure the effects of parents’ education on their offspring’s first union relative to young adults’ educational experiences. By including young adult’s family background and school experiences in the same model, this chapter examines how both ascribed and achieved characteristics of young adults influence the union formation process. Second, I examine whether the effects of parents’ education on their children’s union formation varies by first union type (marriage or cohabitation). Men and women are increasingly likely to cohabit, either as a step in the path to marriage or as an alternative to marriage (Heuveline & Timberlake, 2004; Smock & Gupta, 2002). Studies indicate young adults with highly educated parents often marry at later ages than those whose parents have lower education (Axinn & Thornton, 1992; South, 2001). However, no studies to date have measured how parents’ education influences the rate at which children cohabit. This question is important for understanding the relationship between education and cohabitation, as well as the process by which educational attainment is transmitted within families. Third, I test how the effects of parents’ education and young adults’ school experiences on union formation vary by gender. The effect of parents’ educational attainment on delaying marriage has shown to be stronger for men than women (Axinn & 87 Thornton, 1992). One reason for this is because women who marry at younger ages often curtail their educational attainment, thereby reducing intergenerational transmission of education from parents to their daughters compared to sons (Marini, 1978). However, a growing number of women are marrying at older ages and this effect is due, in part, to the high college graduation rates among women (DiPrete & Buchmann, 2006). This chapter builds on existing literature by exploring whether the effect of parents’ education on young adults’ decision to marry or cohabit varies by parent and child’s gender. My research draws on the National Survey of Families and Households (NFSH), which contains detailed union histories of parents and children. The chapter first outlines previous research on the relationship between parents’ education and young adults’ union formation, then posits several questions about intergenerational influences on young adults’ first union experiences and whether gender differences exist. After the data and measures are explained, hazard ratios from Cox regression models are presented. The Influence of Status Attainment on Union Formation Parents’ education is the strongest predictor of children’s educational attainment (Blau & Duncan, 1967). Parents’ invest in their children’s school success by helping with homework and providing financial resources for higher education (Astone & McLanahan, 1991; Sewell & Shah, 1968; Steelman & Powell, 1991). Highly educated parents are likely to have children who delay marriage for four reasons. First, highly educated parents are likely to encourage their children to leave home at later ages and to finish their education before marrying (Goldscheider & Goldscheider, 1998). Second, young adults with 88 highly educated parents often adopt their parents’ educational expectations (Axinn & Thornton, 1992). Due to high educational aspirations, young adults are more likely to attend college and less likely to marry while in school (Thornton, Axinn & Teachman, 1995). Third, parents’ education leads to greater economic resources, which plays a large role in the development of high aspirations for material goods among children (Axinn & Thornton, 1992). Adolescents with high consumption aspirations report the desire to marry and have children at older ages once an adequate standard of living is reached (Crimmins, Easterlin & Saito, 1991). Fourth, parents’ high socioeconomic status improves eventual educational outcomes among youth (Kao & Thompson, 2003). Parents’ education also influences the type of union selected by their offspring. Parents’ education influences their children’s type of union through the transmission of attitudes and values (Axinn & Thornton, 1993). In particular, attitudes toward union formation are shaped by attitudes toward other life activities and circumstances. For example, cohabitors have a distinctly different set of attitudes and beliefs about the division of household labor and the importance of career success than the married (Clarkberg, Stolzenberg & Waite, 1995). These attitudes are developed, in large part, from education, work and other life experiences. Highly educated parents who have strong, positive attitudes toward educational attainment and career success are likely to have children who share these attitudes. In turn, young adults who highly value education and work, particularly women, are likely to cohabit because these 89 attitudes often diverge from the gendered family roles associated with marriage (Cunningham, Beutel, Barber & Thornton, 2005; Johnson, 2002). Gender Differences Historically, the father’s occupational status was used as a proxy for social background and was considered the best indicator of social mobility (Blau & Duncan, 1967). However, mothers’ education also has a strong influence on the educational attainment of offspring through their educational expectations for children (Thornton, Axinn & Xie, 2007). Both men and women are likely to achieve higher occupational status if their parents also have high education (Featherman & Hauser, 1978). However, the transmission of educational attainment from parents to their offspring varies by young adults’ gender. Fathers’ education has the strongest effect on their sons’ occupational status, while mothers’ education has the strongest effect on the occupational status of daughters (Hauser, Warren, Huang & Carter, 2000). In turn, the transmission of social status within families is strongest in same-sex family relationships. Similar to status attainment, union formation behaviors are also influenced by gender. Attitudes toward family-related behaviors, such as cohabitation and divorce, are more likely to be shared between mothers and daughters (Axinn & Thornton, 1993; Kapinus, 2004). Mothers have different preferences for their daughters’ family formation behaviors, which then influence the rate at which daughters marry and become parents (Barber, 2000). The current chapter tests whether the effect of parents’ education on their offsprings’ union formation decisions varies by parent and child’s gender. These findings contribute to other studies which examine the intergenerational 90 transmission of attitudes and family formation behaviors by gender (Axinn & Thornton, 1992; Kapinus, 2004). Young Adults’ Educational Experiences and the Union Formation Process The transition into a cohabitation/marital union is often based on perceived economic stability, which is intimately tied to the amount of education one achieves, or aspires to obtain. Young adults who are currently enrolled in college are often less likely to form unions because they are less financially stable than those who have completed a college degree (Thornton, Axinn & Teachman, 1995). The relationship between education and likelihood of marrying has weakened over time, explained in part by the increase in cohabitation, which may serve as a substitute for marriage in times of economic uncertainty (Oppenheimer, 2003). Similarly, the timing and duration of schooling is increasingly varied, with many young married people still in school, or returning to obtain another degree (Jacobs & King, 2002). The high college completion rates of recently born cohorts, particularly women (DiPrete & Buchmann, 2006), presents a set of new questions concerning whether educated women will choose to remain single given greater economic independence (Becker, 1981). Contrary to this assertion, there has been an increase in marriage among women who are college graduates compared to less educated women (Goldstein & Kinney, 2001). This finding suggests the relationship between educational attainment and marriage will continue to shift with changes in the economy and the social value attributed to marriage, education and work. The current chapter examines the effects of college enrollment and college degree attainment on young adults’ union 91 formation to gauge how education influences the union formation process in light of these macro-level changes. B. Research Questions This investigation addresses the following research questions: 1) How do young adults’ college enrollment and college degree attainment influence their transition into marriage or cohabitation as a first union? 2) How does parents’ education influence young adults’ transition into a first union after controlling for young adults’ educational experiences? 3) To what extent does the effect of parents’ education on their children’s first union vary by union type (marriage versus cohabitation)? 4) To what extent does the effect of parents’ education on their children’s union formation vary by parent and child’s gender? C. Methods Sample This chapter measures the effect of family structure and parental marital transitions on young adults’ union formation using the detailed union histories of the National Survey of Families and Households (NSFH). Data for the first wave of NSFH were collected in 1987 and 1988 and involved interviews with over 13,000 respondents, including a main cross-section and an over-sample of minorities, single-parent families, families with stepchildren, cohabiting couples and recently married persons. In each household, an adult was randomly selected as the primary respondent and the spouse or cohabiting partner was asked to complete a shorter, self-administered questionnaire. The second wave of data 92 (NSFH2) was fielded between 1992 and 1994. Recently, data collection was completed on a third wave of data (2001-2002) and extensive interviews were conducted of the focal children of the main respondent, ages 18-34. Detailed union history data among Wave 3 respondents make the NSFH a rich source for analyzing union formation. The sample for this analysis consists of children who were interviewed at Wave 2 and/or Wave 3 (N=2,388). These children range in age from 18 to 26 at Wave 2 and 18 to 35 at Wave 3. Approximately 654 children participated at both time waves, 1,298 participated at Wave 3 but not Wave 2 and 436 participated at Wave 2 and not Wave 3. Ninety-three percent of these children are the biological child of the main respondent, 4 percent are stepchildren and 3 percent are adopted, foster children or the child of the respondent’s partner. Measures Dependent Variable At Wave 3, main respondents and their children were asked a series of questions about each union experience, including how the union began (cohabitation or marriage), how it ended (separated, divorced, widowed, intact) and the date of union entry and exit. At Wave 2, complete union histories were not obtained. In this case, two questions asking about marital and cohabitation status were used to construct relationship variables for respondents who participated at Wave 2 but not Wave 3. Among women, 56 percent had experienced either a cohabiting or marital union. Forty-four percent of men experienced either type of union. Approximately 39 percent of young adults 93 (N=934), remained single during the study period. For women who had experienced a union, 15 percent had married their partner without ever cohabiting and 41 percent exited single life through cohabitation. A smaller percentage of men married without first cohabiting (9 percent), while 35 percent cohabited with a romantic partner. About 40 percent of men and women whose first union was cohabitation went on to marry that partner. Since the focus of this chapter is on the transition into a first union, those who experienced a marital union or cohabitated during the study period were analyzed separately as 1) those that married as a first union (N=339); and 2) those that cohabitated as a first union (N=1,075). Table 5.1 provides summary statistics for the variables used in these analyses, comparing those who experienced a union with those who remained single during the study period. In this case, the study period is defined as the person-months of observation from when the child first entered a union, coded here as age 16. The study period begins in 1987 and ends when the respondent was last interviewed. For children whose last interview was Wave 2, this year was coded as either 1992 or 1993. For children whose last interview was Wave 3, the year was coded as 2001, 2002 or 2003. Nineteen respondents reported marrying or cohabiting before age 16; however, these cases are left-censored. Independent Variables Family transitions Parents’ union transitions are constructed from the main respondent’s detailed union history reports at Time 3. These transitions include the experience of 1) a parental 94 separation or divorce (N=1,168); 2) parental remarriage (N=936); 3) parental cohabitation (N=822) and 4) parental death (N=157). Twenty-seven percent of the sample (N=650) did not have union history reports at Wave 3. In this case, marital histories were constructed from Wave 1 variables asking the respondent how many times they had been married, whether the marriage had ended, and how this marriage ended (separation, divorce, widowed). If the respondent reported being currently married, cohabitation was ascertained by a variable asking whether the respondent lived with any previous spouse and/or the current spouse before marriage. If the respondent was not married at the time of interview, cohabitation was ascertained by a variable asking whether main respondent had ever lived with a romantic partner. Parent income and education Parents’ income is the log of total household income at Time 1. Mother and father’s education is a summary measured constructed at Time 1, which is an extraction of the primary respondents’ formal educational attainment from their detailed educational attendance and degree history. Education is measured in years of formal education completed from 0-20, where education of 16 years or more is considered “College degree or more”, which is coded into a dichotomous variable. Respondents who have “Some college, no degree” were coded as 13 years, but less than 16 years of schooling. This coding scheme was chosen to test whether the effects of parents’ education on their children’s union formation differed if parents completed college or not. The spouse’s educational attainment is also obtained from Time 1, where available. Thirty-nine percent of the spouses 95 in the study sample did not participate at Time 1 (N=926). In this case, a question which asks the main respondent about any previous spouse’s education is used. If these two variables are missing, as in 34 percent of cases (N=816), the constructed education variable for spouses at Time 2 is used. After this step, there are twenty cases missing for female spouses (.08%) and 126 male spouses missing (5.2%). In preliminary models, a variable was included to represent these missing values in the analyses to test if there were differences in young adults’ union formation for those with missing parental information. This variable was not significant in any of the models tested and was removed from the analyses for the sake of parsimony. For the cases where no spousal information was available, the main respondents’ education was used. Table 5.1 presents the breakdown of these missing values by young adults’ union type. Young adult demographic characteristics The analysis includes controls for gender (1=female) and two dummy variables for race (Black, Other; White is referent). Measured at Time 3, enrollment in school is measured by a series of questions which ask the young adult if they have ever been enrolled in a four-year college or university. School enrollment is added as a time- varying covariate to the model. Thus each person contributes one record for each month in the study until a union occurs, and receives a “0” or “1” according to whether they were enrolled in school either full or part time, in any month during the study period. The person is censored on school enrollment once a union occurs. The variable for enrollment was lagged by one year in order to assess the effects of school enrollment before a union occurs. 96 Young adults’ attainment of a college degree was measured at Time 3 using a series of questions which ask young adults if they have received a bachelor’s degree or higher and the date of degree attainment. College degree attainment is also a time-varying covariate. In this case, attainment of bachelor’s degree is coded “1” if young adults report attaining a bachelor’s degree in a given month and “0” otherwise. The person is censored on college degree attainment once a union occurs. Similar to school enrollment, the variable for college degree attainment was lagged by one year to assess the effect of obtaining a college degree before a union occurs. Plan of Analysis The analysis models the effects of parents’ education and other explanatory variables on the transition into first marriage or cohabitation using a Cox competing risks model (Cox, 1972). In the model, the hazard or rate of entering a union, h ij (t), is defined as the conditional probability that the event occurs between t and t+∆t and the event is of type j, given the person has not experienced the event by time t : h ij (t) = lim Pr{t < T j <t + Δt, J i =j|T i >t } , j=1…5. In the case of type-specific hazards, the occurrence of one type of event removes the individual from risk of other event types (Allison, 1995). In other words, men and women whose first co-residential union is a marriage are censored from the analysis at time of marriage. Likewise, those who cohabit as a first union are censored from the analysis at time of the cohabitation. Finally, children Δt 97 who remained single during the study period are right-censored because they did not experience either type of union during the study period. Due to the addition of time-varying covariates, each person receives a separate record for person-month of observation, which spans from 1987 to 2003. This creates 128,263 person-months of observation for men and 137,022 person-months for women from which the analysis is conducted. Most competing risks models are based on the assumption that alternative states, or events, are stochastically independent. This assumption for studying cohabitation and marriage is a tenuous one since individuals who are more likely to form a union may be more similar to one another than those who remain single. Methods to control for shared unobserved risk factors between the two groups have been conducted using multinomial logistic regression (Hill, Axinn & Thornton, 1993) and those who have employed these techniques have found that it does not significantly change their results (Barber & Axinn, 1998). In one study, however, accounting for unmeasured heterogeneity diminishes the effect of cohabitation on childbearing outside of marriage, suggesting non-marital fertility is markedly similar to childbearing within marriage (Musick, 2006). For these analyses, I will assume conditional independence between cohabitation and marriage, which falls in line with the view that cohabitation and marriage are family building behaviors, not necessarily independent events (Brein, Lillard & Waite, 1999; Hachen, 1988). 98 D. Results Table 5.1 displays the descriptive characteristics of men and women who remain single and those who enter a cohabitation or marital union. A lower proportion of young adults who cohabit as a first union have parents who are college graduates compared to young adults who remain single or marry. Parents of young adults who cohabit as a first union also earn considerably less than parents of young adults who marry as a first union. The differences in parents’ educational attainment and household income indicate young adults who cohabit as a first union come from lower socioeconomic origins than those who remain single or marry. Family transitions also show important differences between young adults who cohabit versus those who marry as a first union. About 60 percent of cohabitors experience parental marital dissolution, compared to 42 percent of married and single young adults. Similarly, about 50 percent of young adults who cohabit as a first union have parents who also cohabited compared to 21 percent among those who marry, indicating a pattern of socialization and/or modeling of parents’ union formation behaviors. A higher proportion of those who cohabit or marry as a first union also have parents who remarry than young adults who remain single. This finding corroborates another study which shows an increase in cohabitation among young adults who spend some portion of their lives living with stepparents (Teachman, 2003). Looking at young adults’ educational experiences shows a higher proportion of singles enrolled in college compared to those who marry or cohabit. 99 Table 5.1: Unweighted Descriptive Statistics for Variables Used in Analysis of the Rate of Union Formation: National Survey of Families and Households (N=2,388) No Union Marriage Cohab (N=934) (N=339) (N=1,075) Family background characteristics Mother has college education or more .21 .15 .12 (.41) (.36) (.35) Mother has some college, no degree .18 .19 .26 (.39) (.40) (.42) Mother has less than some college (referent) .60 .64 .62 (.49) (.48) (.49) Mother education missing .01 .02 .01 (.08) (.09) (.10) Father has college education or more .28 .26 .19 (.45) (.44) (.40) Father has some college, no degree .14 .14 .15 (.35) (.35) (.36) Father has less than some college (referent) .54 .56 .61 (.50) (.50) (.49) Father education missing .04 .04 .05 (.24) (.20) (.22) Total household income $52,994 $54,788 $49,519 ($52,520) ($51,751) ($52,078) No parental union transition (referent) .48 .50 .32 (.50) (.50) (.47) Parental separation/divorce .42 .42 .58 (.49) (.49) (.49) Parental remarriage .32 .37 .47 (.47) (.49) (.50) Parental cohabitation .30 .21 .47 (.46) (.41) (.49) Parental death .06 .08 .06 (.24) (.27) (.24) Young Adult Characteristics Female .48 .62 .54 (.50) (.49) (.50) Age at time of first union -- 22.08 21.54 (3.06) (.3.14) White (referent) .70 .87 .82 (.46) (.34) (.37) Black .21 .07 .12 (.41) (.26) (.33) Other race .09 .06 .06 (.29) (.24) (.24) Enrolled in college during study period a .45 .32 .41 (.50) (.46) (.49) Obtained bachelor's degree during study period a .15 .32 .20 (.36) (.47) (.39) a Presented here are the frequencies for young adults who ever enrolled in college or received a college degree. In the analyses, these variables are time-varying for each month between 1987-2003. Standard deviation in parentheses. 100 However, a higher proportion of cohabitors enroll in college compared to the married. This finding suggests young adults may not be finished with their education at the time of cohabitation. However, young adults who marry as a first union are more likely to have obtained a college degree compared to the cohabiting group. Odds ratios are presented in the following tables. An odds ratio of 1.0 indicates no relationship with the rate of union formation. An odds ratio of less than 1.0 indicates a slower rate of union formation, and an odds ratio greater than 1.0 indicates a faster rate of union formation. The percent change in the rate of union formation associated with a one-unit change in the variable under consideration is equal to (e B -1) X 100. Effect sizes of coefficients between event types and by gender are tested with the Wald chi-square statistic: (b 1 -b 2 ) 2 /[s.e.(b 1 )] 2 + [s.e(b 2 )] 2 . The exact partial method is used to control for the number of ties. Table 5.2 displays the effects of parents’ education, young adults’ educational experiences and other family background characteristics on the rate of marriage and cohabitation as a first union among men. My first research question asks how young adults’ college enrollment and college degree attainment influence the transition into marriage or cohabitation as a first union. Model 1 demonstrates college enrollment among sons reduces the monthly odds of marriage by 53 percent. In contrast, college degree attainment increases the monthly odds of marriage among sons by 74 percent. These findings suggest 101 Table 5.2: Odds Ratios of the Effect of Parents' Education and Men's Educational Experiences on First Union Formation, NSFH Marriage a Cohabitation b M1 M2 Men's Educational Experiences Enrolled in college in past month .47 ** 1.01 (.13) (.21) Bachelor's degree in past month 1.74 *** 1.08 (1.17) (.75) Family Background Characteristics c Mother has college degree or more .53 ** .73 ** (.14) (.10) Mother has some college, no degree .82 1.16 (.26) (.17) Father has college degree or more 1.04 .76 ** (.24) (.11) Father has some college, no degree .94 1.13 (.34) (.19) Total household income (Log) 1.17 1.01 (.15) (.06) Race d Black .42 ** .78 (.16) (.12) Other Race .12 ** .82 (.11) (.18) Family Transitions e Parental separation/divorce 1.04 1.67 *** (.32) (.25) Parental remarriage 1.90 ** 1.08 (.60) (.17) Parental cohabitation .32 ** .85 (.08) (.10) Parental death 1.46 1.54 ** (.51) (.29) Log Likelihood -1358 -4821 χ 2 /df 55/13 62/13 N 2,388 2,388 Person-months 128,263 128,263 *p<.10, *p<.05, **p<.01, ***p<.001 a For this analysis, people who marry without cohabiting are censored from the analysis at the time of the marriage. b For this analysis, people who cohabit without marrying are censored from the analysis at the time of the cohabitation. c The omitted category is mother/father with less than some college. e The omitted category is no family transition. 102 educational experiences have a significant influence on the timing of marriage among men. However, Model 2 shows no effect of college degree attainment and college enrollment on sons’ odds of cohabitation as a first union. Looking back at Model 1 shows the effects of parents’ education on their son’s rate of marriage. In answer to my second research question, these findings suggest mother’s college degree attainment significantly decreases the monthly odds of union formation among sons after controlling for sons’ educational experiences. Mothers’ college education reduces the likelihood of marriage as a first union among sons by 47 percent. Put another way, sons whose mothers have less than a college education are almost two times as likely to marry as a first union compared to sons whose mothers have a college degree or more. This finding suggests sons whose mothers are college graduates marry at older ages compared to sons whose mothers do not have a college degree. Further, this finding indicates the acquisition of a college degree among mothers—not only some college—reduce the odds of marriage as a first union among sons. Taken together, the negative effect of mothers’ college education and the positive effect of son’s college degree on young men’s odds of first marriage suggest sons whose mothers are college graduates are likely to delay marriage until graduating college themselves. Model 2 displays the effects of parents’ college education on their sons’ rate of cohabitation. Both mother’s and father’s college degree attainment reduces the monthly odds of cohabitation among sons by 27 and 24 percent, 103 respectively. Similar to the significant effect of parents’ college degree attainment on their children’s rate of first marriage, this finding suggests parents’ college degree attainment significantly decreases their son’s rate of cohabitation after including son’s own educational experiences in the model. My third research question asked the extent to which the effect of mother and father’s education on their children’s union formation differed by union type. Testing the coefficients across models for mothers who have a college degree or more demonstrates no significant difference on their sons’ monthly odds by union type. In contrast, fathers’ college degree attainment significantly reduces their sons’ monthly odds of cohabitation, but has no effect on marriage. When comparing the coefficients for fathers’ education between models, the Wald chi- square statistic is significantly higher than the chi-square critical value for 1 degree of freedom (17.75>3.84). This finding suggests father’s college degree attainment has a stronger effect on reducing their sons’ odds of cohabitation compared to marriage as a first union. Table 5.3 presents results concerning the effect of mother’s and father’s education on women’s rate of union formation. Looking first at young women’s educational experiences in Model 1 shows college enrollment reduces the monthly odds of marriage among daughters by 34 percent. However, daughters who obtain a college degree have 60 percent higher monthly odds of marriage compared to those without a college degree. This finding supports another study which shows women with a college degree are more likely to marry than are women with less than college (Goldstein & Kinney, 2001). As shown in Model 2, 104 Table 5.3: Odds Ratios of the Effect of Parents' Education and Women's Educational Experiences on First Union Formation, NSFH Marriage a Cohabitation b M1 M2 Women's Educational Experiences Enrolled in college in past month .66 ** 1.02 (.15) (.22) Bachelor's degree in past month 1.60 *** 1.21 (1.56) (.26) Family Background Characteristics c Mother has college degree or more .68 ** .71 ** (.10) (.11) Mother has some college, no degree .86 .92 (.17) (.15) Father has college degree or more .78 .72 ** (.16) (.09) Father has some college, no degree .97 .95 (.29) (.15) Total household income (Log) 1.01 .93 (.10) (.06) Race d Black .33 *** .43 (.10) (.07) Other Race 1.43 .72 (.36) (.15) Family Transitions e Parental separation/divorce .78 1.75 *** (.21) (.27) Parental remarriage 1.53 .79 (.45) (.12) Parental cohabitation .47 *** 1.34 (.09) (.16) Parental death 1.35 1.15 (.40) (.26) Log Likelihood -2159 -5371 χ 2 /df 49/13 107/13 N 2,388 2,388 Person-months 137,022 137,022 *p<.10, *p<.05, **p<.01, ***p<.001 a For this analysis, people who marry without cohabiting are censored from the analysis at the time of the marriage. b For this analysis, people who cohabit without marrying are censored from the analysis at the time of the cohabitation. c The omitted category is mother/father with less than some college. e The omitted category is no family transition. 105 however, college enrollment and college degree attainment do not have a significant effect on women’s odds of cohabitation as a first union. Looking again at Model 1 shows mothers’ college degree attainment reduces the monthly odds of their daughter’s marriage by 32 percent. That is, daughters who have mothers with less than a college education have 1.5 greater odds of marrying as a first union compared to daughters whose mothers have a college degree or more. In turn, mothers who are college graduates have a stronger influence on reducing their daughter’s odds of marriage as a first union compared to mothers who are not college graduates. This finding is similar to the effect of mothers’ college degree attainment on sons’ rate of marriage, as shown in Table 5.2 (Model 1). Comparing Tables 5.2 and 5.3 suggest mothers’ college degree delays marital formation among sons and daughters. However, mothers who attend college and do not graduate have no effect on their sons or daughters rate of first marriage. These findings emphasize the importance of mothers’ college degree attainment on reducing their sons’ and daughters’ rate of union formation. Model 2 presents the effect of parents’ education on daughters’ rate of cohabitation. Here, it is evident fathers’ college degree attainment is an important predictor of reducing their daughter’s odds of cohabitation. Mother and father’s college education reduces the odds of their daughters’ cohabitation by 29 and 28 percent, respectively. When testing the coefficients across event types, there are no significant differences of mothers’ or fathers’ college degree on their daughter’s odds of marriage or cohabitation as a first union. Thus mothers and 106 fathers have the same effect of reducing their daughters’ monthly odds of marriage and cohabitation. This finding stands in contrast to Table 5.2 which showed fathers’ college education has a stronger effect on delaying sons’ odds of cohabitation compared to marriage. Although not the immediate focus of this chapter, the effects of parents’ union transitions have interesting and significant effects on young adults’ union formation. Table 5.2 shows sons whose parents have ever separated or divorced have 67 percent higher monthly odds of cohabitating as a first union compared to sons with continuously married parents. Looking at Table 5.3 shows a similar effect: daughters whose parents ever separated or divorced have 75 percent greater odds of cohabitating compared to daughters whose parents were continuously married. The effect of parents’ separation and/or divorce are statistically different across event types for sons (36.77>3.84) and daughters (54.17>3.84). This finding suggests the occurrence of parental separation and/or divorce encourages the practice of cohabitation among children, a finding which has been supported in other research (Axinn & Thornton, 1996; Cherlin, Kiernan & Chase-Lansdale, 1995). Looking back at Table 5.2 shows parental cohabitation also reduces the monthly odds of marriage among sons by 68 percent. Similarly, parental cohabitation reduces the monthly odds of marriage among daughters by 53 percent, as shown in Table 5.3. This effect is significantly different across event types for men (12.49>3.84) and women (41.32>3.84). This finding resembles other research which shows parents’ cohabitation increases the odds children may 107 also cohabit, thereby delaying marriage to older ages (Teachman, 2003). My fourth research question asks whether the effects of parents’ education on their children’s rate of union formation differ by parent and child’s gender. Tables 5.4 and 5.5 display the same odds ratios presented in Tables 5.2 and 5.3, but compare the odds ratios by gender rather than union type. Table 5.4 compares the effects of parents’ education and other family background characteristics by gender on the odds of first marriage. Table 5.4: Odds Ratios of the Effect of Parents' Education and Men and Women's Educational Experiences on Marriage as a First Union, NSFH a Men Women M1 M2 Young Adults' Educational Experiences Enrolled in college in past month .47 ** .66 ** (.13) (.15) Bachelor's degree in past month 1.74 *** 1.60 *** (1.17) (1.56) Family Background Characteristics b Mother has college degree or more .53 ** .68 ** (.14) (.10) Mother has some college, no degree .82 .86 (.26) (.17) Father has college degree or more 1.04 .78 (.24) (.16) Father has some college, no degree .94 .97 (.34) (.29) Race d Black .42 ** .33 *** (.16) (.10) Other Race .12 ** 1.43 (.11) (.36) Family Transitions e Parental separation/divorce 1.04 .78 (.32) (.21) Parental remarriage 1.90 ** 1.53 (.60) (.45) 108 Table 5.4, Continued Men Women M1 M2 Parental cohabitation .32 ** .47 *** (.08) (.09) Parental death 1.46 1.35 (.51) (.40) Log Likelihood -1358 -2159 χ 2 /df 55/13 49/13 N 2,388 2,388 Person-months 128,263 137,022 a For this analysis, people who marry without cohabiting are censored from the analysis at the time of the marriage. b The omitted category is mother/father with less than some college. Table 5.5 compares these same effects by gender on the odds of cohabitation. Comparing Tables 5.4 and 5.5 shows parents’ education has the same effect on their sons and daughter’s rate of union formation. In sum, the effect of parents’ education on their children’s union formation does not vary by child’s gender. E. Discussion The current investigation builds on other studies which suggest status attainment within families influences the union formation process. These results corroborate other studies which indicate young adults’ educational experiences have a significant influence on the timing and type of union selected by young adults. In particular, young adults who are enrolled in college are less likely to marry as a first union. College students may delay marriage until finishing school 109 Table 5.5: Odds Ratios of the Effect of Parents' Education and Men and Women's Educational Experiences on Cohabitation as a First Union, NSFH a Men Women M1 M2 Young Adults' Educational Experiences Enrolled in college in past month 1.01 1.02 (.21) (.22) Bachelor's degree in past month 1.08 1.21 (.75) (.26) Family Background Characteristics b Mother has some college, no degree .73 ** .71 ** (.10) (.11) Mother has at least some college 1.16 .92 (.17) (.15) Father has some college, no degree .76 ** .72 ** (.11) (.09) Father has at least some college 1.13 .95 (.19) (.15) Race d Black .78 *** .43 (.12) (.07) Other Race .82 .72 (.18) (.15) Family Transitions e Parental separation/divorce 1.67 *** 1.75 *** (.25) (.27) Parental remarriage 1.08 .79 (.17) (.12) Parental cohabitation .85 1.34 (.10) (.16) Parental death 1.54 ** 1.15 (.29) (.26) Log Likelihood -4821 -5371 χ 2 /df 62/13 107/13 N 2,388 2,388 Person-months 128,263 137,022 a For this analysis, people who cohabit without marrying are censored from the analysis at the time of the marriage. b The omitted category is mother/father with less than some college. 110 because marriage requires the adoption of a social role which may be incongruous with the student role (Thornton, Axinn & Teachman, 1995). Compared to marriage, college enrollment does not significantly reduce the odds of cohabitation among men and women. This finding provides evidence young adults may be more likely to cohabit than marry while enrolled in college. In contrast to college enrollment, college degree attainment increases the odds of marriage as a first union. College degree attainment is a signal to others of one’s future economic success (Clarkberg, 1999; Schwartz & Mare, 2005). Men and women who are college graduates are more likely to marry than cohabit because educational attainment increases the likelihood of steady, well-paying employment. In turn, economic stability increases the odds of marriage (Sassler & Goldscheider, 2004; Sweeney, 2002). In line with research showing the increase in marriage among the highly educated (Goldstein & Kinney, 2001), this chapter adds to other studies which show college degree attainment is an important precursor to marriage among men and women. A contribution of this chapter to the existing literature on status attainment and union formation is the inclusion of parents’ and their children’s educational experiences in the same model to test the relative effects of educational attainment on union formation. My findings in Tables 5.2 and 5.3 show parents’ education has a strong effect on delaying union formation among young adults after controlling for young adults’ educational experiences. These results indicate parents’ education has a strong influence on the timing of young adults’ unions. Similarly, these results also suggest the effect of fathers’ education varies by 111 union type. In particular, fathers’ college education has a stronger influence on reducing their sons’ odds of cohabitation as a first union compared to marriage. Since college degree attainment increases the odds of marriage among men, it is likely fathers who are college graduates transmit the desire to graduate college among sons. In turn, sons who graduate college will be more likely to marry than cohabit as a first union. The fourth research question asked the extent to which the effect of parents’ education varied by parent and child’s gender. Tables 5.4 and 5.5 showed parents’ education has the same effect on the timing of their children’s unions. In other words, there were no gender differences of parents’ education on young men and women’s union formation decisions. This findings stands in contrast to other studies which demonstrate the transmission of attitudes and behaviors is strongest among same-sex family pairs (Axinn & Thornton, 1993; Barber, 2000; Hauser, Warren, Huang & Carter, 2000). This study looks at another area of interfamilial transmission—status attainment—and its effects by child’s gender. This chapter provides new evidence status attainment processes within the family may not be gender-specific. The chapter has several limitations concerning the items used to measure parents’ status attainment. There are several variables, many of which are available in the NSFH, which may have better captured parents’ investment in child’s educational success, such as items that measure supervision of child’s school activities and parent-child relationship quality. Other measures of parents’ economic status could also have been included, such as parents’ welfare receipt 112 and asset accumulation. Since my findings show parents’ education is a strong predictor of delaying young adults’ union formation, my future research will incorporate more detailed measures of the family’s economic status to further confirm this relationship. A second area for future inquiry is to explore why parents’ education delays union formation. High educational aspirations between parents and children are one explanation for the delay in young adults’ union formation. The NSFH contains measures of parents’ educational aspirations for their children and young adults’ own educational aspirations. However, this measure is only asked of parents with children ages 5 to 17 at Wave 1. Further, only children ages 10 to 17 at Wave 2 were asked their future educational aspirations. Thus including this measure in the study would have required a considerable truncation of the sample. In conclusion, my findings demonstrate the link between young adults’ social origins and their future life course decisions. Though young adults have more autonomy as to the timing and type of romantic unions they enter, their partnering decisions are strongly influence by their parents’ educational attainment. The exploration of status attainment and its effect on union formation is a relatively new area of research. The outcomes of these studies promise to provide a more complete explanation of the trend in delayed marriage among men and women. These studies will also clarify how educational attainment influences union formation. Further exploration of the relationship between educational attainment and union formation is important because the transitions in and out of educational institutions are increasingly varied and often stretch long into the life 113 course. This chapter is a first attempt to uncover how young adults fare in a world of increasingly varied childhood circumstances and complex life choices. 114 Chapter Six: Discussion and Conclusions A. Summary Each of the three chapters of my dissertation has made a meaningful contribution to the study of family socialization. The first chapter showed grandparents, particularly grandmothers, have a strong influence on the religiosity of their grandchildren. Congruence of religious beliefs and values across three generations counters the view the family is losing moral ground in socializing young people within a religious tradition. However, parental divorce weakens the intergenerational transmission of religious beliefs between mothers and their children. In turn, divorce may interrupt socialization processes within families and may undermine the continuity of religious beliefs across generations. My second chapter explored how values toward individualism and collectivism change over the young adult life course. Contrary to the widely held belief young adults are eschewing family life for their own personal goals, my second chapter found young adults rank loyalty to family and community as more desirable in life than their parents thirty years earlier. The significant, independent effects of college graduation, marriage, cohabitation and parenthood on young adults’ values demonstrate values shift in accordance with life events. Most importantly, my results indicate life course transitions influence the change in values among young adults. For example, young adults who marry and/or have children had higher collectivistic values than single and childless young adults. In turn, these life events engender a shift in one’s values toward family loyalty and community involvement. Another important finding was that young adults who have cohabited rank individualistic values higher than young adults who have not 115 cohabited. This finding indicates young adults who cohabit have a different set of values toward family life compared to married young adults. Family background, such as parents’ marital history and family size, was also shown to influence young adults’ values toward individualism and collectivism. The third chapter examined a different avenue of family socialization—parents’ educational attainment. This chapter addressed how status attainment processes within the family influence children’s union formation decisions. My findings demonstrate parents’ college degree attainment delays union formation among young adults. However, the extent to which parents’ education influences their children’s union formation is conditioned by the type of first union selected (marriage or cohabitation). In particular, the effect of father’s educational attainment reduced son’s monthly odds of cohabitation to a greater degree than marriage. Last, this chapter demonstrated there were no significant differences by young adults’ gender of parents’ education on their children’s union formation. In sum, this chapter demonstrated young adults’ decisions to cohabit and/or marry a partner are closely tied to their socioeconomic origins. The pattern of influence of parents on their children’s union formation is remarkably similar for men and women. The transitions associated with young adulthood are occurring at a different pace than fifty years ago. Likewise, the increasing diversity of romantic partnerships has permanently altered the form and function of romantic relationships. Despite these macro-level social changes, the decision to cohabit and/or marry a partner is anchored 116 within a family context. As a whole, my dissertation provides evidence that parents— and grandparents—influence young adults’ beliefs, values and behaviors over the life course. B. Limitations As described in the second chapter of this dissertation, one of the datasets used for the analyses—the LSOG—presents several weaknesses that are important to note. The most relevant limitation for these analyses was the lack of information on the timing of life course transitions, such as college graduation, marriage and cohabitation. For example, respondents were asked the year of their current union status, rather than the date of their first union. Moreover, the three-year intervals between study waves after 1991 leaves some guesswork as to when and if romantic unions occurred during non- study periods. However, these temporality issues do not affect the finding in the third chapter that the occurrence of life events, at any particular time, changes young adults’ values. A second limitation concerns the selectivity of the sample. The LSOG sampling frame is based on a “middle America” sample, the majority of which are White, economically stable middle- and working-class families (Bengtson, 1996). Careful analyses of attrition have shown older adults, males, ethnic minorities and those with lower education are more likely to exit the study. However, the sample is replenished when the youngest family members turn age 16. For example, the number of G4 male respondents rose from 206 in 1997 to 324 in 2000. Likewise, the number if G4 female respondents went up from 259 in 1997 to 356 in 2000. The availability of a web survey in 2005 is also expected to increases response rates over time. 117 The weaknesses of the LSOG must be viewed in light of its strengths. The availability of data from four generations of family members makes the LSOG a unique resource for understanding how attitudes, values and behaviors are transmitted from one generation to another. Similarly, the generation-sequential design allows for the comparison of younger generations to their parents at similar ages. This feature was useful in my third chapter which compared religious beliefs and values between grandparents in 1971 with their parents in 1985 and children in 2000. Matching on age was particularly important in this chapter because religious beliefs and values are age- specific (Argue, Johnson & White, 1999). This design was also important in my fourth chapter which compared young adults’ values toward individualism and collectivism in 2000 with their parents at similar ages in 1971. Another advantage of the study is its longitudinal design. With the addition of another survey round in 2005, the LSOG consists of eight waves of data. The replication of items at every time wave, such as the Rokeach Value Survey, allows for the study of how values have changed over time. Similarly, the addition of a fifth generation of young adults in future waves will make the LSOG an even more important data source for the study of family socialization processes and its effects on life course transitions in adulthood. C. Implications for Current Knowledge and Future Research This dissertation adds to the cumulative body of knowledge and bears implications for future research by demonstrating how social origins influence life choices. Using a life course framework, my research shows family socialization has important implications for young adults’ beliefs, values and behaviors. As a whole, this 118 dissertation makes the following contributions: First, my dissertation provides evidence to suggest the family has an enduring influence on the religious beliefs and values of young adults. Young adults are situated in a complex web of family relationships and receive messages about religious beliefs, values, and expectations from parents, grandparents, and other relatives. Until recently, grandparents have not been considered in socialization theory and research. Grandparents were relegated to the periphery of the nuclear family model, and their influence on youths’ outcomes was not explored. However, the third chapter of my dissertation demonstrates grandmothers have a significant influence on the religious beliefs and values of their grandchildren. This finding suggests the continued resilience and relevancy of families for the passing on of religious traditions and beliefs to younger generations. Contrary to the predictions of the secularization thesis, the influence of religion appears to be on the rise in American civic life today (Roof, 1999; Smith, 2005; Wuthnow, 1998). Thus it is important that we know more about how religious values and practices are socialized in children and how, or whether, family influences have changed over time. We need to understand patterns of religious influence in families across generations and how religion helps or hinders family members to cope with contemporary crises and burdens such as poor health or caregiving for elderly parents or other dependent family members. The only study in this area showed religious congruence increased the level of intergenerational assistance--such as help around the house, financial assistance and child care--between adult children and their parents (Myers, 2004). Similarly, this study found the link between religion and 119 intergenerational assistance was negative when exchanges involved a divorced parent but positive when continuously married parents were involved. This finding demonstrates parental divorce reduces the flow of intergenerational assistance and weakens the close emotional ties between parents and children (Amato & Booth, 1997; Pearce & Axinn, 1998; Sherkat & Ellison, 1999; Silverstein & Bengtson, 1997). Further studies should explore how religious beliefs and values influence the provision of intergenerational assistance in later life. Second, my dissertation findings indicate young adults’ values toward individualism/collectivism are responsive to age and life course transitions. The fourth chapter of my dissertation showed young adults’ values change with age, as well as in response to life course transitions, such as college completion, marriage, cohabitation and parenthood. This finding corroborates other studies which show young adults’ values are malleable over the life course (Alwin & Krosnick, 1991). Young adults reorganize their values in accordance with the adoption of new roles and responsibilities. Future research on the timing of union formation should attempt to measure the extent to which role transitions engender value changes. In particular, the attitudes and values associated with young adult life transitions are likely to vary widely across racial/ethnic groups. My findings in Chapter 4 should be replicated in nationally representative samples with an adequate over-sample of race and ethnic minorities. A large-scale study on the change in values over the young adult life course by race/ethnicity and socioeconomic status would uncover the different meanings associated with young adult life transitions, and provide clues as to why the rates of marriage, cohabitation and non-marital childbearing vary across social groups. 120 Third, my dissertation demonstrates status attainment processes in childhood influence union formation decisions in adulthood. Proponents of the individuation theory assert the ascribed family characteristics, such as religious affiliation, cultural background and social class, have less influence on the mate selection process today compared to fifty years ago (Lesthaeghe, 1983; South, 2001). However, my findings indicate social status, as measured by parents’ education, have a strong influence on when a union occurs and what form it takes. Social mobility is an increasingly important marker of when (and if) individuals choose to marry and have children. The strong relationship between educational attainment and union formation is particularly evident in a recent study which found college-educated women today are more likely to marry than women with less than a college education (Goldstein & Kinney, 2001). Future research should examine the relationship between status attainment and union formation to determine the extent to which parents’ social location influences their children’s life course transitions in young adulthood. Future studies should also explore other measures of parents’ human capital, such as the availability of resources and parental investment in schooling, to further explore the relationship between status attainment and union formation. 121 Bibliography Acock, A., & Bengtson, V. L. (1980). Socialization and attribution processes: Actual vs. perceived similarity among parents and youth. Journal of Social Research. 68: 151- 171. Aldous, J. & Hill, R. (1965). Social cohesion, lineage type and intergenerational transmission. Social Forces. 43(4): 471-82. Allison, P.D. (1995). Survival Analysis Using SAS: A Practical Guide. Cary, N.C.: SAS Institute, Inc. Allison, P. (1994). Using panel data to estimate the effects of events. Sociological Methods and Research. 23: 174-99. Alwin, D.F. & Krosnick, J.A. (1991). Aging, cohorts and the stability of sociopolitical orientations over the life span. American Journal of Sociology. 97(1): 169-95. Alwin, D.F. (2002). Generations x, y and z: Are they changing America? Contexts. 1(4): 42-51. Amato, P.R. & Cheadle, J. (2005). The long reach of divorce: Divorce and child well- being across three generations. Journal of Marriage and the Family. 67: 191-206. Amato, P.R. & DeBoer, D.D. (2001). The transmission of marital instability across generations: Relationship skills or commitment to marriage? Journal of Marriage and Family. 63(4): 1038-51. Amato, P.R. & Booth, A. (1997). A Generation at Risk: Growing Up in an Era of Family Upheaval. Cambridge: Harvard University Press. Amato, P.R. (1996). Explaining the intergenerational transmission of divorce. Journal of Marriage and Family. 58(3): 628-40. Amato, P.R. & Keith, B. (1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin. 110: 26-46. Argue, A., Johnson, D.R. & White, L.K. (1999). Age and religiosity: Evidence from a three-wave panel analysis. Journal for the Scientific Study of Religion. 38(3), 423- 35. Arnett, J.J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist. 469-80. 122 Astone, N.M, Nathanson, C.A., Schoen, R. & Kim, Y.J. (1999). Family demography, social theory and investment in social capital. Population and Development Review. 25(1): 1-31. Astone, N.M. & McLanahan, S.S. (1994). Family structure, residential mobility and school drop out: A research note. Demography. 31(4): 575-84. Axinn, W.G. & Barber, J.S. (1997). Non-family living and family formation in early adulthood. Journal of Marriage and Family. 59(3): 595-611. Axinn, W.G. & Thornton, A. (1996). The influence of parents’ marital dissolutions on children’s attitudes towards family formation. Demography. 33: 66-81. Axinn, W.G. & Thornton, A.(1993). Mothers, children and cohabitation: The intergenerational effects of attitudes and behavior. American Sociological Review. 58(2): 233-46. Axinn, W.G. & Thornton, A. (1992). The influence of parental resources on the timing of the transition to marriage. Social Science Research. 21: 261-85. Bao, W.N., Whitbeck, L.B., Hoyt, D.R. & Conger, R.D. (1999). Perceived parental acceptance as a moderator of religious transmission among adolescent boys and girls. Journal of Marriage and the Family. 61(2): 362-74. Barber, J.S. (2001). Ideational influences on the transition to parenthood: Attitudes towards childbearing and competing alternatives. Social Psychology Quarterly. 64(2): 101-27. Barber, J.S. (2000). Intergenerational influences on the transition into parenthood: Mothers’ preferences for family and non-family behavior. Social Forces. 79(1): 319-48. Barber, J.S. & Axinn, W.G. (1998). The impact of parental pressure for grandchildren on young people’s entry into cohabitation and marriage. Population Studies. 52(2): 129-44. Bartkowski, J.P. & Xu, X. (2000). Distant patriarchs or expressive dads? The discourse and practice of fathering in conservative Protestant families. The Sociological Quarterly. 41: 465-85. Becker, G.S. (1981). A Treatise on the Family. Cambridge: Harvard University Press. Bellah, R. N., Madsen, R., Sullivan, W. M., Swidler, A., & Tipton, S. M. (1985). Habits of the Heart: Individualism and Commitment in American Life. New York: Harper & Row. 123 Bengtson, V. L., Biblarz, T. J., & Roberts, R. E. L. (2002). How Families Still Matter: A Longitudinal Study of Youth in Two Generations. New York: Cambridge University Press. Bengtson, V. L. (2001). Beyond the nuclear family: the increasing importance of multigenerational bonds. Journal of Marriage and the Family. 63: 1-16. Bengtson, V.L. (1996).Continuities and discontinuities in intergenerational relationships over time. In V.L. Bengtson (Ed.) Adulthood and Aging: Research on Continuities and Discontinuities. New York: Springer. Beutel, A.M. & Marini, M.M. (1995). Gender and values. American Sociological Review. 60(3): 436-48. Bianchi, S.M., Casper, L.M. & King, R.B. (Eds.) (2005). Work, Family, Health and Well-Being. Mahwah, N.J: Lawrence Erlbaum Associates. Biblarz, T.J. & Gottainer, G. (2000). Family structure and children’s success: A comparison of widowed and divorced single-mother families. Journal of Marriage and Family. 62: 533-38. Biblarz, T.J. & Raftery, A.E. (1999). Family structure, educational attainment and socioeconomic success: Rethinking the “pathology of matriarchy”. American Journal of Sociology. 105(2): 321-65. Biblarz, T.J., Bengtson, V.L. & Bucur, A. (1996). Social mobility across three generations. Journal of Marriage and the Family. 58(1): 188-200. Blackwell, D.L. & Lichter, D.T. (2004). Homogamy among dating, cohabiting and married couples. The Sociological Quarterly. 45(4): 719-37. Blau, F.D., Kahn, Lawrence M. & Waldfogel, J. (2004). The impact of welfare benefits on single motherhood and headship of young women: Evidence from the Census. Journal of Human Resources. 1: 382-404. Blau, P.M. & Duncan, O.D. (1967). The American Occupational Structure. New York: John Wiley & Sons, Inc. Bramlett, M.D. & Mosher, W.D. (2002). Marriage, Divorce and Remarriage in the United States. National Center for Health Statistics. Vital and Health Statistics, 23(22): 1-93. Brien, M.J., Lillard, L.A. & Waite, L.J. (1999). Interrelated family-building behaviors: Cohabitation, marriage and nonmarital conception. Demography. 36: 535-551. 124 Brooks, C. & Bozendahl, C. (2004). The transformation of U.S gender role attitudes: Cohort replacement, social-structural change and ideological learning. Social Science Research. 33: 106-33. Brown, S.L. (2000). Union transitions among cohabitors: The significance of relationship assessement and expectations. Journal of Marriage and Family. 62: 833-46. Buchmann, C. & DiPrete, T.A. (2006). The growing female advantage in college completion: The role of parental resources and academic achievement. American Sociological Review. 71: 515-41. Bumpass, L.L. & Lu, H.H. (2000). Trends in cohabitation and implications for children’s family contexts in the United States. Population Studies. 54: 29-41. Bumpass, L.L., Sweet, J.A. & Cherlin, A. (1991). The role of cohabitation in the declining rates of marriage. Journal of Marriage and Family. 53(4): 913-27. Casper, L.M. & Bianchi, S.M. (2002). Continuity and Change in the American Family. Thousand Oaks, CA: Sage. Casper, L.M. & Cohen, P.N. (2000). How does POSSLQ measure up? Historical estimates of cohabitation. Demography. 37(2): 237-45. Casper, L.M. & Sayer, L.C. (2000). Cohabitation Transitions: Different Attitudes and Purposes, Different Paths. Working paper. Chan, C. G. and Elder, Elder, G. H. (2000). Matrilineal advantage in grandchild– grandparent relations. The Gerontologist. 40: 179-190. Clarkberg, M. (1999). The price of partnering: The role of economic well-being in young adults’ first union experiences. Social Forces. 77: 945-68. Clarkberg, M. Stolzenberg, R.M. & Waite, L. (1995). Attitudes, values and entrance in cohabitational unions. Social Forces. 74: 609-32. Chandra A., Martinez G.M., Mosher W.D., Abma J.C. & Jones J. 2005. Fertility, Family Planning, and Reproductive Health of US Women: Data from the 2002 National Survey of Family Growth. National Center for Health Statistics. Vital and Health Statistics, 23(25): 1-174. Chaves, M. (1991). Family structure and Protestant church attendance: The sociological basis of cohort and age effects. Journal for the Scientific Study of Religion. 39: 329- 40. 125 Chaves, M. (1990). Holding the cohort: Response to Hout and Greeley. Journal for the Scientific Study of Religion. 29(4): 525-30. Chaves, M. (1989). Secularization and religious revival: Evidence from U.S. church attendance rates: 1972-1986. Journal for the Scientific Study of Religion. 28(4): 464-77. Cherlin, A. (2004). The deinstitutionalization of American marriage. Journal of Marriage and Family. 66: 848-61. Cherlin, A. (2002). Public and Private Families: An Introduction. 3 rd edition. Boston: McGraw-Hill College. Cherlin, A., Kiernan, K. & Chase-Lansdale, L. (1995). Parental divorce in childhood and demographic outcomes in young adulthood. Demography. 32: 299-318. Clark, C.A., Worthington, E.L. & Danser, D.B. (1988). The transmission of religious beliefs and practices from parents to firstborn early adolescent sons. Journal of Marriage and the Family. 50(2): 463-72. Clarkberg, M. (1999). The price of partnering: The role of economic well-being in young adults’ first union experiences. Social Forces. 77: 945-68. Clarkberg, M., Stolzenberg, R.M. & Waite, L. (1995). Attitudes, values and entrance in cohabitational unions. Social Forces. 74: 609-32. Comrey, A.L. & Newmeyer, J.A. (1965). Measurement of radicalism-conservatism. Journal of Social Psychology. 67: 357-69. Cox, D.R. (1972). Regression models and life tables. (with discussion). Journal of the Royal Statistical Society. B34: 187-220. Crimmins, E.M, Easterlin, R.A. & Saito, Y. (1991). Preference changes among American youth: Family, work and goods aspirations: 1976-86. Population and Development Review. 17: 115-33.. Crissey, S.R. (2005). Race/ethnic differences in the marital expectations of adolescents: The role of romantic relationships. Journal of Marriage and Family. 67: 697-709. Cunningham, M. & Thornton, A. (2006). The influence of parents’ marital quality on adult children’s attitudes toward marriage and its alternatives: Main and moderating effects. Demography. 43(4): 659-72. 126 Cunningham, M., Beutel, A., Barber, J. and Thornton, A. (2005). Reciprocal relationships between attitudes about gender and social contexts during young adulthood. Social Science Research. 34:862-892. Cunningham, M. & Thornton, A. (2005). The influence of parents’ marital quality on adult children’s attitudes toward marriage: Main and moderating effects. Demography. 43(4): 659-72. Danigelis, N.L., Hardy, M. & Cutler, S.J. (2007). Intracohort aging and sociopolitical attitudes. American Sociological Review. 72(5): 812-30. Darnell, A. & Sherkat, D.E. (1997). The impact of Protestant fundamentalism on educational attainment. American Sociological Review. 62: 306-15. Denham, T. E., & Smith, C. W. (1989). The influence of grandparents on grandchildren: A review of the literature and resources. Family Relations. 38: 345-350. DiPrete, T.A. & Buchmann, C. (2006). Gender-specific trends in the value of education and the emerging gender gap of college completion. Demography. 43(1): 1-24. East, P.L. (1998). Racial and ethnic differences in girls’ sexual, marital and birth expectations. Journal of Marriage and Family. 60: 150-62. Easterlin, R.A & Crimmins, E.M. (1991). Private materialism, personal self-fulfillment, family life and public interest: The nature, effects and causes of recent changes in the values of American youth. Public Opinion Quarterly. 55(4): 499-533. Edin, K. & Kefalas, M. (2005). Promises I Can Keep: Why Poor Women Put Motherhood Before Marriage. Berkeley: University of California Press. Edgell, P. (2006). Religion and Family in a Changing Society. Princeton, NJ: Princeton University Press. Eggebeen, D.J. & Knoester, C. (2001). Does fatherhood matter for men? Journal of Marriage and Family. 63: 381-93. Elder. G.H., Johnson, M.K. & Crosnoe, R. (2003). The emergence and development of life course theory. In Mortimer, J.T. & Shanahan, M.J. (eds.). Handbook of the life course. (pp.3-19). New York: Kluwer Academic Press. Featherman, D.L. & Hauser, R.M. (1978). Opportunity and Change. New York: Academic Press. 127 Fields, J. (2003). America’s Families and Living Arrangements: 2003. Current Population Reports, P20-553. U.S Census Bureau, Washington D.C. George, L.K. (1993). Sociological perspectives on life transitions. Annual Review of Sociology. 19: 353-73. Gerson, K. (1985). Hard Choices: How Women Decide About Work, Career and Motherhood. Berkeley and Los Angeles: University of California Press. Giddens, A. (1991). Modernity and Self-Identity: Self and Society in the Late Modern Age. Stanford University: Stanford University Press. Glass, J. & Jacobs, J. (2005). Childhood religious conservatism and adult attainment among black and white women. Social Forces. 84(1): 555-79. Glass, J., Bengtson, V. L. & Dunham, C. (1986). Attitude similarity in three-generation families, Socialization, status inheritance or reciprocal influence? American Sociological Review, 51, 685-698. Glick, P.C. (1947). The family life cycle. American Sociological Review. 12(2): 164- 74. Glick, P.C. (1977). Updating the life cycle of the family. Journal of Marriage and Family. 39: 5-14. Glick, P.C. & Norton, A.J. (1977). Marrying, divorcing and living together in the United States. Journal of Marriage and Family. 42: 19-30. Goldscheider, F. & Goldscheider, C. (1998). The effects of childhood family structure on leaving and returning home. Journal of Marriage and Family. 60:745-56. Goldscheider, F. (1997). Recent changes in U.S young adult living arrangements in comparative perspective. Journal of Family Issues. 18: 708-24. Goldstein, J.R. & Kinney, C.T. (2001). Marriage delayed or marriage forgone? New cohort forecasts of first marriage for U.S. women. American Sociological Review. 66(4): 506-19. Goodman, C. & Silverstein, M. (2001). Grandmothers who parents their grandchildren: Affective relations across three generations and implications for psychological well-being. Journal of Family Issues. 22: 557-78. Guzzo, K.B. (2006). The relationship between life course events and union formation. Social Science Research. 35: 384-408. 128 Hachen, D.S. (1988). The competing risks model: A method for analyzing processes with multiple types of events. Sociological Methods and Research. 17(1): 21-54. Hagestad, G. (1985). Continuity and connectedness. In V.L. Bengtson & J.F. Robertson (Eds.) Grandparenthood. (pp.31-48). Beverly Hills, CA: Sage. Hausman, J. (1978). Specification tests in econometrics. Econometrica. 46: 1251-71. Hauser, R.M., Warren, J.R., Huang, M.H. & Carter, W.Y. (2000). Occupational status, education and social mobility in the meritocracy. (pp. 179-229). In K.Arrow, S. Bowles, & S. Durlauf (Eds). Meritocracy and Inequality. Princeton: Princeton University Press. Haveman, R. & Wolfe, B. (1995). The determinants of children’s attainments: A review of methods and findings. Journal of Economic Literature. 33(4): 1829-1878. Hecter, M. & Kanazawa, S. (1997). Sociological rational choice theory. Annual Review of Sociology. 23: 191-214. Heuveline, P. & Timberlake, J.M. (2004). The role of cohabitation in family formation: The United States in comparative perspective. Journal of Marriage and Family. 66(6): 1214-30. Hill, D.H., Axinn, W.G. & Thornton, A. (1993). Competing hazards with shared unmeasured risk factors. Sociological Methodology. 23: 245-77. Hill, M.S. & Yeung, W.J. (1999). How has the changing structure of opportunities affected transitions to adulthood? In A. Booth, A.C Crouter and M.J. Shanahan (Eds). Transitions to Adulthood in a Changing Economy: No Work, No Family, No Future? (pp. 3-39). Westport, CT: Praeger. Hitlin, S. & Pilavin, J.A. (2004). Values: Reviving a dormant concept. Annual Review of Sociology. 30: 359-93. Hout, M. & Fischer, C.S. (2002). Why more Americans have no religious preference: Politics and generation. American Sociological Review. 67(2): 165-90. Hout, M. & Greeley, A. (1990). The cohort doesn’t hold: Comment on Chaves (1989). Journal for the Scientific Study of Religion. 29(4): 519-24. Jacobs, J.A. & King, R.B. (2002). Age and college completion: A life history analysis of women aged 15-44. Sociology of Education. 75(3): 211-30. Johnson, D.R. (1995). Alternative methods for the quantitative analysis of panel data in family research: Pooled times series models. Journal of Marriage and the Family. 57(4): 1065-77. 129 Johnson, M.K. (2005). Family roles and work values: Processes of selection and change. Journal of Marriage and Family. 67:352-69. Johnson, M.K. (2002). Social origins, adolescent experiences and work value trajectories during the transition to adulthood. Social Forces. 80: 1307-41. Kalmijn, M. (1998). Intermarriage and homogamy: Causes, patterns, trends. Annual Review of Sociology. 24: 395-421. Kalmijn, M. (1994). Assortative mating by cultural and economic occupational status. American Journal of Sociology. 100(2): 422-52. Kalmijn, M. (1991). Shifting boundaries: Trends in religious and educational homogamy. American Sociological Review. 56(6): 786-800. Kao, G. & Thompson, J.S. (2003). Racial and ethnic stratification in educational achievement and attainment. Annual Review of Sociology. 29: 417-42. Kapinus, C.A. (2004). The effects of parents’ attitudes toward divorce on offspring’s attitudes: Gender and parental divorce as mediating factors. Journal of Family Issues. 25: 112-35. Kaufman, G. & Uhlenberg, P. (2000). The influence of parenthood on work effort of married men and women. Social Forces. 78(3): 931-47. Kertzer, D.I. & Schaie, K.W. (1989). Age structuring in comparative perspective. Hillsdale, N.J: Erlbaum. King, V. (1994). Variation in the consequences of nonresident father involvement for children’s wellbeing. Journal of Marriage and the Family. 56: 963-72. King, V., & Elder, G. H., Jr. (1999). Are religious grandparents more involved grandparents? Journal of Gerontology: Social Sciences. 54: S317-S328. King, V. (2003). The influence of religion on fathers’ relationships with their children. Journal of Marriage and the Family. 65(2): 382-95. Kao, G. & Thompson, J.S. (2003). Racial and ethnic stratification in educational achievement and attainment. Annual Review of Sociology. 29: 417-42. Lawton, L.E. & Bures, R. (2001). Parental divorce and the “switching” of religious identity. Journal for the Scientific Study of Religion. 40: 99-111. Lehrer, E.L. (1999). Religion as a determinant of educational attainment: An economic perspective. Social Science Research. 28: 358-79. 130 Lesthaeghe, R. & Niedert, L. (2006). The second demographic transition in the United States: Exception or textbook example? Population and Development Review. 32(4): 669-698. Lesthaeghe, R. & Surkyn, J. (1988). Cultural dynamics and economic theories of fertility change. Population and Development Review. 24(1): 1-15. Lesthaeghe, R. (1983). A century of demographic and cultural change in Western Europe: An exploration of underlying dimensions. Population and Development Review. 9(3): 411-35. Lichter, D., Graefe, D.R. & Brown, J.B. (2003). Is marriage a panacea? Union formation among economically disadvantaged unwed mothers. Social Problems. 50: 60-86. Little, R. J. A., & Rubin, D. B. (2002). Statistical Analysis with Missing Data. Hoboken, NJ: Wiley. Lye, D.N. & Waldron, I. (1997). Attitudes toward cohabitation, family and gender roles: Relationships to values and political ideology. Sociological Perspectives. 40(2): 199-225. Mannheim, K. (1952/1922). The problem of generations. In K. Mannheim (Ed.) Essays in the Sociology of Knowledge. (pp.276-322). London: Routledge and Kegan Paul. Manning, W.D., Longmore, M.A. & Giordano, P.C. (2007). The changing institution of marriage: Adolescent’s expectations to cohabit and to marry. Journal of Marriage and Family. 69(3): 559-575. Marini, M.M. (1978). The transition to adulthood: Sex differences in educational attainment and age at marriage. American Sociological Review. 43(4): 483-507. Martin, S.P. & Parashar, S. (2006). Women’s changing attitudes toward divorce, 1974- 2002: Evidence for an educational crossover. Journal of Marriage and Family. 68(1): 29-40. Maslow, A. (1954). Motivation and Personality. New York: Harper. McLanahan, S. (2005). Fragile families and the marriage agenda. In Fragile Families and the Marriage Agenda. L. Kowaleski-Jones & N. Wolfinger (Eds.) New York: Kluwer Academic Press. Miller, A., & Hoffman, J. (1995). Risk and religion: An explanation of gender differences in religiosity. Journal for the Scientific Study of Religion. 34(1): 63-75. 131 Moen, P. & Yu, Y. (2001). Effective work/life strategies: Working couples, working conditions and life quality. Social Problems. 47(3): 291-326. Moen, P., Erickson, M. A., & Dempster-McClain, D. (1997). Their mother’s daughters? The intergenerational transmission of gender attitudes in a world of changing roles. Journal of Marriage and the Family, 59, 281-293. Moffitt, R.A. (1998). Introduction: The effect of welfare on marriage and fertility. In Robert A. Moffitt (Ed). Welfare, the Family and Reproductive Behavior. Washington, D.C: National Academy Press. Mosher, D. W., Williams, L.B. & Johnson, D.P. (1992). Religion and fertility in the United States: New patterns. Demography. 29(2): 199-214. Mueller, M. M. & Elder, G. (2003). Family contingencies across the generations: Grandparent-grandchild relationships in holistic perspective. Journal of Marriage and Family, 65(2), 404-17. Muller, C. (1998). Gender differences in parental involvement and adolescents’ mathematics achievement. Sociology of Education. 71(4): 336-56. Musick, K. (2006). Cohabitation, Nonmarital Childbearing and the Marriage Process. Working Study. California Center for Population Research. Myers, S. M. (2004). Religion and intergenerational assistance: Distinct differences by adult children’s gender and parent’s marital status. Sociological Quarterly, 45, 67- 89. Myers, S. M. (1996). An interactive model of religiosity inheritance: The importance of family context. American Sociological Review, 61, 858-866. Neugarten, B.L., Moore, J.W. & Lowe, J.C. (1965). Age norms, age constraints and adult socialization. American Journal of Sociology. 70: 229-36. Nock, S.L. (1995). A comparison of marriage and cohabiting relationships. Journal of Family Issues. 15: 53-76. Oppenheimer, V.K. (2003). Cohabiting and marriage during young men’s career development process. Demography. 40(2): 127-49. Oppenheimer, V.K., Kalmijin, M. & Lim, N. (1997). Men’s career development and marriage timing during a time of rising inequality. Demography. 34: 311-30. Oppenheimer, V.K. (1988). A theory of marriage timing. American Journal of Sociology. 94:563-91. 132 O’Rand, A.M. (2006). Stratification of the life course: Social origins, life course capital and cohort inequality. In R.H. Binstock & L.K. George (Eds.) Handbook of Aging and the Social Sciences. 6 th ed. (pp. 145-62). New York: Academic Press. Pearce, L. D. & Axinn, W. G. (1998). The impact of family religious life on the quality of mother-child relations. American Sociological Review, 63, 810-828. Petersen, P.G. (1999). Gray Dawn: How the Coming Age Wave will Change America—And the World. New York: Random House. Ploeg, M.V. (2002). Children from disrupted families as adults: Family structure, college attendance and college completion. Economics of Education Review. 21: 171-84. Popenoe, D. (1993). American family decline: 1960-1990: A review and appraisal. Journal of Marriage and the Family. 55(3): 527-42. Popenoe, D. (1988). Disturbing the Nest: Family Change and Decline in Modern Societies. New York: Aldine de Gruyer. Portes, A. & Landolt, P. (1996). The downside of social capital. The American Prospect. (May-June). 18-26, 94. Pratt, M.W. & Fiese, B.H. (Eds.) (2004). Family Stories and the Life Course: Across Time and Generations. Malwah, NJ: Lawrence Erlbaum Associates. Raley, R.K., Frisco, M.L. & Wildsmith. (2005). Maternal cohabitation and educational success. Sociology of Education. 78: 144-64. Raley, R.K. (1996). A shortage of marriageable men? A note on the role of cohabitation in Black-White differences in marriage rates. American Sociological Review. 61: 973-83. Rindfuss, R.R. (1991). The young adult years: Diversity, structural change, and fertility. Demography. 28(4): 493-512. Roberts, R.E.L. & Bengtson, V.L. (1999). The social psychology of values: Effects of individual development, social change and family transmission of the life span. In C. Ryff & V. Marshall (Eds.) The Self and Social Processes. (pp. 453-82) New York: Springer. Rokeach, M. (1973). The Nature of Human Values. New York: Free Press. Roof, W. C. (1999). Spiritual Marketplace: Baby Boomers and the Remaking of American Religion. Princeton, NJ: Princeton University Press. 133 Roof, W. C., (1993). Generation of Seekers: The Spiritual Journeys of the Baby Boom Generation. San Francisco: Harper San Francisco. Rosenfeld, M. (2006). Young adulthood as a factor in social change in the United States. Population and Development Review. 32(1): 27-51. . Rosenfeld, M. & Kim, B.S. (2005). The independence of young adults and the rise of interracial and same-sex unions. American Sociological Review. 70(4): 541-62. Ryder, N.B. (1965). The cohort as a concept in the study of social change. American Sociological Review. 30: 843-61. Sarkisian, N. & Gerstel, N. (2004). Kin support among blacks and whites: Race and family organization. American Sociological Review. 69(4): 812-37. Sassler, S. & Goldscheider, F. (2004). Revisiting Jane Austen’s theory of marriage timing: Changes in union formation among young men in the 20 th century. Journal of Family Issues. 25(2): 139-66. Sayer, L.C. (2005). Gender, time and inequality: Trends in women’s and men’s paid work, unpaid work and free time. Social Forces. 84(1): 285-303. Sayer, L. C, Cohen, P.N. & Casper, L.M. (2004). Women, men and work. In The American People: Census 2000. Published by the Russell Sage Foundation and the Population Reference Bureau. Sayer, L.C. & Bianchi, M.S. (2000). Women’s economic independence and the probability of divorce. Journal of Family Issues. 21: 906-23. Schoen, R. Landale, N.S. & Daniels, K. (2007). Family transitions in young adulthood. Demography. 44(4): 807-20. Schoen, R., Kim, Y.J., Nathanson, C.A., Fields, J. & Astone, N.M. (1997). Why do Americans want children? Population and Development Review. 23(2): 333-58. Schwartz, C.R. & Mare, R.D. (2005). Trends in educational assortative mating from 1940 to 2003. Demography. 42: 621-46. Settersen, R.A. (2003). Age structuring and the rhythm of the life course. (pp. 81-98). In Mortimer, J.T. & Shanahan, M.J. (eds.). Handbook of the Life Course. (pp.3-19). New York: Kluwer Academic Press. 134 Sewell, W.H., Jr. (1992). A theory of structure: Duality, agency and transformation. American Journal of Sociology. 98: 1-20. Sewell, W.H. & Hauser, R.M. (1980). The Wisconsin longitudinal study of social and psychological factors in aspirations and achievements. Research in Sociology of Education and Socialization. 1: 59-99. Sewell, W.H. & Shah, V.P. (1968). Parents’ education and children’s educational aspirations and achievements. American Sociological Review. 33(2): 191-209. Shanahan, M.J. (2000). Pathways to adulthood: Variability and mechanisms in life course perspective. Annual Review of Sociology. 26: 667-692. Sherkat, D.E. & Darnell, A. (1999). The effect of parents’ fundamentalism on children’s educational attainment: Examining differences by gender and children’s fundamentalism. Journal for the Scientific Study of Religion. 38(1): 23-35. Sherkat, D., & Ellison, C. (1999). Recent developments and current controversies in the sociology of religion. Annual Review of Sociology, 25, 363–394. Sherkat, D. E. (1998). Counterculture or continuity? Competing influences on baby boomers’ religious orientations and participation. Social Forces. 76, 1087-1114. Sherkat, D.E. (1991). Leaving the faith: Testing theories of religious switching using survival models. Social Science Research. 20: 171-87. Silverstein, M., Giarrusso, R., & Bengtson, V. L. (2003). Grandparents and grandchildren in family systems. A socio-developmental perspective. In V. L. Bengtson & A. Lowenstein (Eds.). Global Aging and its Challenges to Families (pp. 75-102). New York: Aldine de Gruyter. Silverstein, M. & Marenco, A. (2001). How Americans enact the grandparent role. Journal of Family Issues 22, 493-522. Silverstein, M. & Bengtson, V.L. (1997). Intergenerational solidarity and the structure of adult-child parent relationships in American families. American Journal of Sociology. 103(2): 429-60. Smith, C. (2005). Soul Searching: The Religious and Spiritual Lives of American Teenagers. New York: Oxford University Press. Smock, P.J., Manning, W.D. & Porter, M. (2005). “Everything’s there except money”: How money shapes decisions to marry among cohabitors. Journal of Marriage and Family. 67: 680-96. 135 Smock, P.J. & Gupta, S. (2002). Cohabitation in contemporary North America. In A. Booth & A. Crouter (Eds.) Just Living Together: Implications for Children, Families and Public Policy (pp. 53-84). Lawrence-Erlbaum. Smock, P.J. (2000). Cohabitation in the United States: An appraisal of research themes, findings and implications. Annual Review of Sociology. 26: 1-20. SOLAS (2001). Users Manual 3.0. Statistical Solutions. Somary, K. & Stricker, G. (1998). Becoming a grandparent: A longitudinal study of expectations and early experiences as a function of sex and lineage. The Gerontologist, 38, 53-61. South, S.J. (2001). The variables effects of family background on the timing of first marriage: United States, 1969-1993. Social Science Research 30, 606-626. Spates, J. (1983). The sociology of values. Annual Review of Sociology. 9: 27-49. Stack, C. (1974). All My Kin: Strategies for Survival in a Black Community. New York: Harper & Row Staff, J. & Mortimer, J.T. (2007). Educational and work strategies from adolescence to early adulthood: Consequences for educational attainment. Social Forces. 87(3): 1169-94. Stata (2002). Users Manual 8.0. Stata Press. Steelman, L.C. & Powell, B. (1991). Sponsoring the next generation: Parental willingness to pay for higher education. The American Journal of Sociology. 96(6): 1505-29. Stolzenberg, R.M., Blair-Loy and Waite, L.J. (1995). Religious participation in early adulthood: Age and family life cycle effects on church membership. American Sociological Review. 60: 84-103. Sweeney, M.M. (2002). Two decades of family change: The shifting economic foundations of marriage. American Sociological Review. 67(1): 132-47. Sweet, J., Bumpass, L.L. & Call, V. (1988). The design and content of the National Survey of Families and Households. Madison: University of Wisconsin, Center for Demography and Ecology. Swidler, A. (1986). Culture in action: Symbols and strategies. American Sociological Review. 51: 273-86. 136 Teachman, J.D. (2003). Childhood living arrangements and the formation of coresidential unions. Journal of Marriage and Family. 65: 507-24. Teachman, J.D. (2002). Stability across cohorts in divorce risk factors. Demography. 39(2): 331-51. Teachman, J.D. & Paasch, K. (1998). The family and educational aspirations. Journal of Marriage and the Family. 60(3): 704-14. Thoits, P. (1992). Identity structures and psychological well-being: Gender and marital status comparisons. Social Psychological Quarterly. 55(3): 236-56. Thornton, A., Axinn, W.G. & Xie, Y. (2007). Marriage and Cohabitation. Chicago: University of Chicago Press. Thornton, A. & Young-DeMarco, L. (2001). Four decades of trends in attitudes towards family issues in the United States: The 1960s through the 1990s. Journal of Marriage and Family. 63, 1009-1037. Thornton, A., Axinn, W.G & Teachman, J.D. (1995). The influence of school enrollment and accumulation on cohabitation and marriage in early adulthood. American Sociological Review. 60(5): 762-74. Thornton, A., Axinn, W.G. & Hill, D.H. (1992). Reciprocal effects of religiosity, cohabitation and marriage. American Journal of Sociology. 98(3): 628-51. Thornton, A. (1991). Influence of the marital history of parents on the marital and cohabitational experiences of children. American Journal of Sociology. 96: 868-94. Treas, J. (2002). How cohorts, education and ideology shaped a new sexual revolution on American attitudes towards nonmarital sex, 1972-1998. Sociological Perspectives. 45(3): 267-83. Uecker, J.E., Regnerus, M.D. & Vaaler, M.L. (2007). Losing my religion: The social sources of religious decline in early adulthood. Social Forces. 85(4): 1-26. Uhlenberg, P. R. (2005). Historical forces shaping grandparent-grandchild relationships: Demography and beyond. In M. Silverstein (Ed.), Intergeneration Relations across Time and Place (pp. 77-97). New York: Springer. Uhlenberg, P., & Hammill, B. G. (1998). Frequency of grandparent contact with grandchild sets: Six factors that make a different. The Gerontologist, 38, 276-285. Visser, P.S. & Krosnick, J.A. (1998). The development of attitude strength over the life cycle: Surge and decline. Journal of Personality and Social Psychology. 75: 1389- 1410. 137 Waite, L., Goldscheider, F. & Witsberger, C. (1986). Nonfamily living and the erosion of traditional family orientations among young adults. American Sociological Review. 51:541-554. Warren, J.R., Hauser, R.M. & Sheridan, J.T. (2002). Occupational stratification across the life course: Evidence from the Wisconsin Longitudinal Study. American Sociological Review. 67: 432-65. Wilson, W.J. (1987). The Truly Disadvantaged: The Inner City, the Underclass and Public Policy. Chicago: Chicago University Press. Wu, L.L. (1996). Effects of family instability, income and income instability on the risk of a premarital birth. American Sociological Review. 61: 386-406. Wuthnow, R. (2005). America and the Challenges of Religious Diversity. Princeton, N.J: Princeton University Press. Wuthnow, R. (1998). After Heaven: Spirituality in America Since the 1950’s. Berkeley and Los Angeles: University of California Press. Yi. C., Chang, C. & Chang, Y. (2004). The intergenerational transmission of family values: A comparison between teenagers and parents in Taiwan. Journal of Comparative Family Studies, 35(4), 523-545.
Abstract (if available)
Abstract
This dissertation aimed to explain how young adults social origins influence their beliefs, values and behaviors in adulthood. More specifically, the objectives were: 1) to measure the extent to which there is intergenerational religious continuity in the family
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Intergenerational transmission of values and behaviors over the family life course
PDF
The measurement, life course patterns, and outcomes of intergenerational ambivalence among parent-adult child dyads
PDF
Normative obligations and parental care in social context
PDF
Intergenerational coresidence of older adults in contemporary Japan: traditional cultural norms in divergent styles
PDF
The immediate and long term legacy of relationships with grandparents for the well-being of grandchildren
PDF
Children 's migration and the financial, social, and psychological well-being of older adults in rural China
PDF
Assessing the impact of diversity courses on students’ values, attitudes and beliefs
PDF
Childlessness and psychological well-being across life course as manifested in significant life events
PDF
Internet communication use, psychological functioning and social connectedness at older ages
PDF
Life transitions, leisure activity engagement, and cognition among older adults
PDF
A neuropsychological exploration of low-SES adolescents’ life goals and their motives
PDF
Problematic alcohol use in Hispanic emerging adults: the role of perceived discrimination, cultural identity, and salient cultural values
PDF
Multilevel influences of care engagement and long-term survival among childhood, adolescent, and young adult cancer survivors
PDF
Anxiety symptoms and nicotine use among adolescents and young adults
PDF
Understanding the role of population and place in the dynamics of seasonal influenza outbreaks in the United States
Asset Metadata
Creator
Copen, Casey E.
(author)
Core Title
Transmission and transitions: young adults' beliefs, values and life course transitions in familial context
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Sociology
Publication Date
07/18/2008
Defense Date
05/29/2008
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cohabitation,intergenerational transmission,Marriage,OAI-PMH Harvest,Religion
Language
English
Advisor
Casper, Lynne (
committee chair
), Biblarz, Timothy J. (
committee member
), Crimmins, Eileen M. (
committee member
), Silverstein, Merril (
committee member
)
Creator Email
ccopen@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m1358
Unique identifier
UC178075
Identifier
etd-Copen-20080718 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-88453 (legacy record id),usctheses-m1358 (legacy record id)
Legacy Identifier
etd-Copen-20080718.pdf
Dmrecord
88453
Document Type
Dissertation
Rights
Copen, Casey E.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
cohabitation
intergenerational transmission