ASRXXX10.1177/0003122415606101American Sociological ReviewOjeda and Hatemi 6061012015

American Sociological Review 2015, Vol. 80(6) 1150–1174­ Accounting for the Child in © American Sociological Association 2015 DOI: 10.1177/0003122415606101 the Transmission of Party http://asr.sagepub.com Identification

Christopher Ojedaa and Peter K. Hatemia

Abstract The transmission of party identification from parent to child is one of the most important components of political socialization in the . Research shows that children learn their party identification from their parents, and parents drive the learning process. The vast majority of studies thus treats children as passive recipients of information and assumes that parent-child concordance equals transmission. Rather than relying on a single pathway by which parents teach children, we propose an alternative view by focusing on children as active agents in their socialization. In so doing, we introduce a two-step model of transmission: perception then adoption. Utilizing two unique family-based studies that contain self-reported measures of party identification for both parents and children, children’s perceptions of their parents’ party affiliations, and measures of the parent-child relationship, we find children differentially learn and then choose to affiliate, or not, with their parents. These findings challenge several core assumptions upon which the extant literature is built, namely that the majority of children both know and adopt their parents’ party identification. We conclude that there is much to be learned by focusing on children as active agents in their political socialization.

Keywords political affiliation, party identification, transmission, values, perception, child agency

The transmission of political party identifica- and partisan identities have increased in tion from parents to children remains one of importance (Baldassarri and Gelman 2008; the most studied concepts in political sociol- Brooks and Manza 1997; Inglehart 1990; ogy (Alwin and Krosnick 1991; Bengtson Iyengar and Westwood 2015). Laws and poli- 1975; Campbell et al. 1960; Glass, Bengtson, cies on marriage, how we can have sex, how and Dunham 1986; Jennings and Markus income is distributed, human rights, free 1984). Such attention is not surprising given speech, healthcare, freedoms and tolerance, the profound influence political parties and competing visions of equality of condition identities have on the individual and society. The classics of postwar political sociology initially focused on membership in social aThe Pennsylvania State University groups—such as ethnicity, class, and religion— as the source of political identities and con- Corresponding Author: Peter K. Hatemi, The Pennsylvania State flicts in democratic societies (Franklin, Mackie, University, Department of Political Science, 307 and Valen 1992). This view has given way to Pond Lab, University Park, PA 16802 more recent scholarship that finds ideological E-mail: [email protected] Ojeda and Hatemi 1151 versus equality of process, who is deserving interaction is needed to lead “us away from of representation and benefits, and every viewing parental behavior as something that other aspect of group life are influenced by is done to children or for children towards the electoral outcomes in western democracies. view that it is done with children.” This per- In the United States, political parties both spective, in contrast to historical work that encapsulate and set the policy stances for treats children as a tabula rasa onto which these issues. parents and other socializing agents inscribe Identification with political parties is political values, is increasingly reflected in equally pervasive in our personal lives. modern sociological explorations of values Whom we choose to affiliate with and marry, writ large (e.g., Calarco 2014; Corsaro and where we choose to live and go to school, and Fingerson 2006). Nevertheless, the majority so many other social experiences are guided of research on political value transmission in part by our political identities (Alford et al. continues to rely almost exclusively on mod- 2011; Cavior, Miller, and Cohen 1975; Kalm- els that present an asymmetric relationship ijn 1998; McAuley and Nutty 1982; Rubinson between parents and children. These models 1986; Stoker and Jennings 2006; Watson propose that children observe and imitate the et al. 2004; Zuckerman 2005). Understanding behaviors of authority figures, most com- the origins of political identity, specifically monly parents, with little control over what is the transmission of party identification, con- learned. tributes to a larger understanding of how Psychosocial development theories, in party politics define and influence social life. which children seek to establish an identity The majority of the political socialization lit- both separate and apart from their family erature proposes that through either active or (Erikson 1968), and self-determination theo- passive teaching, children “learn” their politi- ries, in which individuals internalize social cal party identification from their parents, and values by integrating them into their sense of this identification, in turn, becomes part of self (Grolnick, Ryan, and Deci 1991), extend their identity (Bengtson 1975; Jennings and the social learning approaches. Sociological Niemi 1968; Miller and Glass 1989). This and psychological studies that focus on these view is supported by the purportedly high mechanisms of child learning reveal that indi- concordance between parents’ and their adult vidual differences in child learning are as children’s political party identification, and important as parental teaching (Middleton the belief that positive transmission between and Putney 1963; Patterson, DeBaryshe, and parents and children occurs the majority of Ramsey 1989). Calarco (2014), for example, time. Indirect modes of transmission, such as uses longitudinal ethnography to show that the social milieu created by parents, have also the passage of class-based cultures from par- been proposed as operating on the socializa- ents to children is neither automatic nor tion of party identification, although to a immediate. She found that children were lesser degree (Knoke 1972; McAllister and often reluctant to adopt the problem-solving Kelley 1985). strategies modeled by parents. Children play These foundations have yet to be revisited a critical role in their learning; they create using more recent sociological theory. The their own realities, have their own cognitive sociology of childhood has continued to and emotive biases, selectively attend to develop and given birth to the social child, experiences, and differentially process infor- one who has more agency1 and competence mation, including that from parents. This is than is implied in the political socialization true for social values and norms, as well as literature (Haug 2013; Morrow 2002). In this religious and group affiliations (Bao et al. view, models of top-down parental socializa- 1999; Block and Block 2006; Fraley et al. tion do not adequately reflect the socializa- 2012; Knafo and Schwartz 2003; Koleva and tion process. Rather, as Maccoby and Martin Rip 2009; Miller and Glass 1989; Nelson and (1983:78) argue, an emphasis on parent-child Tallman 1969). If party identification adheres 1152 American Sociological Review 80(6) to the same mechanisms of learning common political values and then choose to adopt to other social traits, then a child’s perception them. Introducing a model that explicitly tests of parental party identification should be par- this assumption is important, because chil- amount to the transmission process. dren might discount, ignore, or choose to These countervailing assumptions about reject their parents’ values. This possibility is children, one in which children have no agency significant because the direct transmission and another in which children have agency, model treats children who accurately perceive raise issues about how children are ontologi- and adopt their parents’ party identification cally treated and studied in political socializa- the same as children who misperceive and tion research. Kohn’s (1983:1) critique of reject their parents’ party identification—thus parent-driven models of political socialization returning a false positive. Imagine, for exam- suggests there is a lack of research focusing on ple, Republican parents with two children. the child because “[t]he object of research has The first child chooses to identify as a Repub- generally been, not to demonstrate a similarity lican because she perceives her parents as in parents’ and children’ values, but to explain being Republican and wants to adopt this a similarity that was assumed to exist.” We affiliation; the second child, however, chooses attempt to resolve these inconsistencies and to identify as a Republican because he incor- continue to build on recent work in the sociol- rectly perceives his parents to be Democrats ogy of childhood by presenting a two-step and want to reject this affiliation. These cases model of political identity transmission that share empirically equivalent outcomes but are provides agency to children within the context qualitatively different from one another. of social learning. We argue that children Transmission occurs in the former but not the observe and often imitate parents’ attitudes and latter. Thus, the mechanics of parent-child behaviors, but they do so critically. Children transmission remain obfuscated by ignoring must perceive and evaluate parents’ attitudes children’s perceptions in the political sociali- and behaviors and then determine whether to zation process. adopt or reject those attitudes and behaviors as This study makes two important contribu- their own. While our focus is on the transmis- tions to understanding the intergenerational sion of party identification, we propose that transmission of party identification. First, and our two-step model is applicable to a wide most important, we show that children play a range of social identities, values, and attitudes. critical role in the transmission of party iden- Party identification is a good initial test of the tification; how children perceive and respond model because of the breadth of research on to parental values affects their own values. political socialization that uses the direct trans- Second, we show that one of the core assump- mission model, the enduring quality of party tions in political sociology, that children typi- identification over the life course, and the cally adopt their parents’ political values, is importance of party identification in shaping not valid. Rather, once measures of child other social attitudes and behaviors (Alwin and perceptions are included, we find parent-child Krosnick 1991). concordance of political identification is far Current models of partisan identification, lower than previously believed. what are labeled direct transmission models (see Jennings, Stoker, and Bowers 2009), point to trait-to-trait correlations as evidence Socialization and the that the parent, or other relevant socializing Transmission of Party agent, has successfully passed on values to Identification the child. Trait-to-trait evidence assumes transmission occurs if the correlation is posi- Political socialization, the process through tive. Yet in order for real transmission to which adolescents and young adults acquire occur, children must actually know parents’ political beliefs, attitudes, and identities and Ojeda and Hatemi 1153 begin to understand and engage the political socialization is less pertinent for political world around them, is fostered by, and medi- identity and more often focused on civic duty ated through, interactions with agents of social- and institutional understanding. For example, ization (Greenstein 1965). These agents, primary and high school environments often through exposure to information and social stimulate political learning through civics reinforcement, create environments and convey education, reciting the pledge of allegiance, perceptions and orientations to individuals, all or extracurricular activities such as student of which are mediated by societal context government (Hess and Torney-Purta 1967; (Boehnke 2001; Huckfeldt 1979; Weiner and Nie, Junn, and Stehlik-Barry 1996). College, Eckland 1979). The past century of research on the other hand, invites a more complex has identified the most important agents of process of political socialization through political socialization as the family, peers, edu- interaction with teachers and students, self- cation, schools, neighborhoods, religion, men- selective exposure to new ideas, and the iden- tors, the mass media, social movements, issue tity gained by voluntarily joining groups and and partisan salience, economic conditions and engaging in student activism (Alwin, Cohen, status, global events, opinion leaders, gender, and Newcomb 1991; Dey 1997; Nasatir 1968; and age (Gauchat 2012; Haynes and Jacobs Weiner and Eckland 1979). 1994; Janowitz and Segal 1967; Lopreato 1967; Peer influences, unlike familial and insti- Manza and Brooks 1997; Thompson 1971). tutional forces such as religion and primary The opportunities under which these agents can school that have a hierarchical and authorita- exert an influence are unlimited and mostly tive foundation, offer a form of group-identity, informal, especially outside the structure of the where fitting in has a more substantial role in election cycle. Politics is simply infused in the acquisition of beliefs (Walker, Hennig, almost every aspect of modern group life and Krettenauer 2000). Individuals, however, (Langton 1969). Young people are socialized most often gravitate toward groups that hold into politics through conversing with family beliefs similar to their own to minimize con- members at dinner, watching television, partici- flict and reinforce their own values. Never- pating in Facebook, or texting with friends, for theless, there is pressure to conform and example. young people desire acceptance. Thus, they Given its prominence in U.S. culture as the tend to preferentially select into like-minded most central component of political identity, groups and then continue to adopt the specific the acquisition of political party identification attitudes, viewpoints, and behaviors of those has been the subject of intense study. Extant groups (Harris 1995). research overwhelmingly shows that the fam- The mass media is also a powerful sociali- ily, parental influence specifically, is the sin- zation force (Krueger 2006; McLeod 2000; gle most important political socialization Merriam 1931). With few exceptions, expo- agent. There are many other social contribu- sure to political persons and events is experi- tors to party identification that have a lesser enced through the media, whether it is news role in the development and transmission of on the Internet or television shows depicting political identification. Upon closer review, political happenings. Starting in the 1980s, however, the literature provides evidence that mass-media experiences have become gradu- for many non-familial socializing agents, ally more individual experiences, especially children’s disposition at least partially guides for young adults, who spend an increasing their perception of and selection into these amount of time on smart phones and comput- experiences. Thus, before we turn to the fam- ers. The Internet and its unlimited forms of ily, we briefly review this literature. social media allow for personalized exposure Much research argues that education plays to information and permit people to self- a prominent role in political socialization select into their preferred forum for every (Haste and Torney-Purta 1992). Initially, this type and view of any political or social topic. 1154 American Sociological Review 80(6)

Social identities, such as class, ethnicity, The empirical foundation for this view, cultural groups, gender, sex, and religion, all that parents pass on their party identification affect political preferences and how individu- to their children, relies almost entirely on the als believe those preferences might be realized uniformly positive trait-to-trait correlations vis-à-vis political parties (Beutel and Marini found between self-reported parent and child 1995; Knoke and Felson 1974; Knoke and political party identification (Bengtson 1975; Hout 1974; Orum and Cohen 1973; Rubinson Campbell et al. 1960; Dalton 1980; Glass 1986; Sears and Valentino 1997). Exploring in et al. 1986; Jennings and Niemi 1968, 1981; detail the influences of every type of social Langton and Jennings 1968; Niemi and Jen- identity on party affiliation is beyond the nings 1991; Tedin 1974). Based on these cor- scope of this article; suffice to say that their relations, three important assumptions have influences during childhood and adolescence been made in this literature: (1) children are operate largely through parental transmission passive learners; (2) a majority of children and the environments parents directly or indi- know their parents’ party identifications; and rectly provide (Glass et al. 1986; Kohn, Slom- (3) a majority of children adopt their parents’ czynski, and Schoenbach 1986). party identifications. This view remains the Of the array of socialization processes, modal and near constant approach for study- Hyman’s (1959:51) proclamation over 50 ing the transmission of party identification. years ago that “foremost among agencies of Perhaps the past 50 years of research in this socialization into politics is the family” area can be summarized by Jennings and col- remains the majority view today. The family leagues (2009), who recently affirmed the is the primary supplier of the home environ- importance of the direct transmission model: ment, culture, and nurturance, and it provides “As expected on the basis of social learning the neighborhood, religion, schools, and theory . . . [t]he direct transmission model is resources that children grow up in during robust, as it withstands an extensive set of their formative years. A hierarchical structure controls. Early acquisition of parental charac- exists within all families, even in the most teristics influences the subsequent nature of non-authoritative families, in which parents adult political development.” Parents are set the rules and establish familial values. The believed to be the primary force in the forma- strong emotional relationships that exist tion of children’s political values because between parents and children compel many they control the flow of political information children to adopt behaviors and attitudes that that children receive (Easton, Dennis, and will please their parents, but at times children Easton 1969; Greenstein 1965) and provide also challenge or refuse parent-approved the environment in which children develop. behaviors and attitudes. Figure 1 shows this viewpoint as a simple In some cases, parents directly teach their model. children about political values, institutions, The parent-child correlations, ranging in and processes of government, but this is not value from .3 to .6, are high enough to sup- common. Rather, parents’ greatest influence port an important role for parents but low in the development of political values is enough to suggest that party identification is through political party identification (Kohn not perfectly transmitted like information 1983); as agents of political socialization, through a fax machine. Scholars have offered parents are most successful in passing on several explanations to explain the discrep- their political identity (Alwin and Krosnick ancy between the claims of the social learning 1991; Campbell et al. 1960; Hyman 1959; hypothesis and the often modest size of the Knoke 1972; Merelman 1986). In the United parent-child correlational evidence. The pri- States, this identity is most often gained mary explanation for the correlation not being through affiliation with either the Republican higher is that parents are not always able to or Democratic Party. send clear and consistent cues about their Ojeda and Hatemi 1155

Figure 1. The Direct Trait-to-Trait Transmission Model Note: PID = political party identification. beliefs (Beck and Jennings 1991; Knoke either case, yet parents and children still may 1972; Tedin 1980). That is, more successful hold the same party identification, and trait- transmission will supposedly occur with bet- to-trait correlation would erroneously count ter communication and higher quality parent- this as transmission. child interactions (e.g., Calarco 2014). Earlier sociological and psychological Additional research shows that the political studies that focus on child development find climate of the home, the salience of political the quality of value transmission depends on issues, and the nature of the times also have a the disposition of the child, with considera- role in the concordance between parents and tions based on developmental stage and age children deviating from unity (Boehnke, Hadjar, (Abendschön 2013; Piaget 1965; Whitbeck and Baier 2007; Sears and Valentino 1997). and Gecas 1988). More specifically, as chil- This work has yet to include modern dren grow up and develop social sophistica- research that embraces child agency and the tion, they gain understanding of the issues, mechanics of child learning (see Mayall and become increasingly aware of the import of Zeiher 2003; Middleton and Putney 1963; political identity, and can exert a greater influ- Morrow 2002; Nelson and Tallman 1969). ence on the transmission process. Political Including child agency has significant impli- identification crystalizes in adolescence and cations for understanding why the concord- young adulthood (Sears and Valentino 1997), ance between parents and children deviates a time at which individuals have developed a from unity, even when communication is high great deal of agency but are often still influ- and issues are salient. Unlike the direct trans- enced by parents. This balance between the mission model, we consider mismatches of child’s agency and the parent’s influence has political identification between parents and been all but overlooked in interpretations of children to be an inherent part of, rather than the greater political socialization research, an error in, the transmission process. We pro- although it is hinted at in numerous studies pose the transmission of political party iden- (e.g., Hyman 1959; Jennings and Niemi 1968). tification between parent and child is a social In addition, several studies, when combined, and psychological process that entails at least provide reason to specifically address this two steps: (1) one individual (i.e., the parent) issue. For example, children can direct the impresses information upon another individ- flow of information they receive about politics ual (i.e., the child), and (2) the latter individ- by initiating conversations with parents ual (i.e., the child) receives and evaluates that (McDevitt 2006; Saphir and Chaffee 2002) information. Both the parent and the child and often seek alternative information that have active roles in this process and might reflects their own values by talking to peers deviate, either intentionally or unintention- (Atkin and Gantz 1978). Similar variability in ally, from perfect transmission. Parents might information-seeking behaviors, which in part not disclose their partisanship, for example, are dispositional and reflect one’s personality to let their children choose a party identifica- (Belsky 1984; Pirolli and Card 1999), account tion for themselves. On the other hand, chil- for differences in how children perceive and dren may either misperceive or reject their evaluate parental information and values. parents’ values. Transmission fails to occur in Indeed, studies of households with more than 1156 American Sociological Review 80(6) one child, including children of the same age, Thus, while sociologists have long voiced attest to profound differences in interpretation concerns over the inadequacies of the unidi- of the same parental cues and messages (Heth- rectional conception of socialization (see erington, Reiss, and Plomin 2013). Kohn 1983), only a handful of studies provide The fact that party identification does not preliminary evidence that child agency is crystallize until late adolescence has two important for political party identification, important implications for this study. First, it one of the most important indicators of politi- is possible to study this two-step process of cal and social identity in the United States. transmission using a sample of adolescents and young adults. One problem that child socialization research encounters is a dearth A Perception-Adoption of data, often due to the difficulties of captur- Model ing socialization near or at the time it occurs. We operationalize and extend previous social Party identification is less susceptible to this learning and perceptual models (Bao et al. problem because of the developmental stage 1999; Westholm 1999) by further integrating during which it emerges. Second, adoles- the nature of child learning into the study of cence and young adulthood is a time when political affiliation. Figure 2 presents our individuals have developed enough agency to baseline perception-adoption model. The determine, at least in part, their own social model also depicts other factors that influence identity. This latter point is important because transmission, such as the social milieu of the it lends plausibility to the argument that mod- family and any potential feedback from the els of transmission should account for the role child to the parent. This milieu can be viewed of the child. as a larger latent construct that encompasses Critical to our argument, research shows the greater societal context (e.g., events, that children’s perceptions of their parents’ social movements, and economic conditions), values and attitudes are more important in as we do in this study, or it can be operation- value transmission and child behavior than alized to include those forces individually. are parents’ actual values and behaviors The model does not assume trait-to-trait parent- (Cashmore and Goodnow 1985; Knafo and child congruence is evidence of positive Schwartz 2003). Yet little empirical work transmission. Rather, it assumes parents and explores this area regarding political orienta- children both affect the transmission process. tions, with two exceptions. With respect to Children must first perceive their parents’ political attitudes, Acock and Bengtson party identification and then choose to adopt (1980) assessed 446 parent-child triads drawn it for transmission to occur. This extension from a population of Los Angeles residents allows us to answer several important ques- belonging to a shared health plan. They found tions: To what degree does the transmission that “within the family, the actual opinions of of party identification in the U.S. electorate parents appear to have little direct bearing on depend on the child’s perception and adoption children’s orientations, except as the actual of parents’ values? What enhances or orientations are perceived and reinterpreted depresses the ability of children to perceive by the children. It is not what parents think, their parents’ party identifications? Why do but what their children think they think, that some children choose to adopt their parents’ predicts their offsprings’ attitudes” (p. 513). political identification while others reject it? Tedin (1980), relying on a very small sample The first step in the transmission process is of parents and children ( n = 155), found that perception. Accurate perception depends on value transmission was greater in children the availability of relevant information and who knew their parents’ political preferences. the ability of the perceiver to detect and uti- Critically different from our view, he consid- lize that information (Funder 1995). Research ered child perception a measure of parental on party identification tends to focus on the communication and not a factor of the child. availability and relevance of information by Ojeda and Hatemi 1157

Figure 2. The Perception-Adoption Model Note: PID = political party identification. emphasizing top-down processes in which (Schönpflug 2001). Research on bottom-up parents exert control over the information processes focuses on children’s motivation to children receive (Easton et al. 1969; Green- adopt or reject parents’ values (Knafo and stein 1965). For example, research consist- Schwartz 2004). Motivation for adopting par- ently finds that the probability of transmitting ents’ values may be based on the perceived party identification modestly increases as par- moral desirability of those values, the desire ents’ cues about political partisanship become for a constrained self-identity, the anticipation more clear and consistent (Knoke and Hout of reward or the avoidance of punishment, the 1974; Lewis-Beck et al. 2008). However, desire to rebel against parents, or the desire to even the most communicative of homes have avoid feelings of shame or guilt that may a substantial gap in parent-child values, result from parent-child value incongruence because the role of the child in detecting and (Achen 2002; Knafo and Assor 2007; Middle- utilizing that information is not accounted for. ton and Putney 1963). The political socializa- The second step in the transmission pro- tion literature has given less attention to why cess is adoption. In this step, children decide children choose to adopt what they believe to whether to adopt their parents’ perceived be their parents’ party identification as their party identification or to reject it. The adop- own. The overwhelming assumption in the tion step has not been examined for political trait-to-trait model is that adoption generally affiliation, but it has been explored for other occurs when party identification is known. social values through both top-down and bottom- The final component of our perception- up perspectives. In the former, parents are adoption model is the social milieu. Social motivated to transmit values to their children milieu—sometimes called third-party, paral- (Whitbeck and Gecas 1988), but parenting lel experiences or zeitgeist—refers to factors styles influence the adoption of values common to both parents and children that 1158 American Sociological Review 80(6) affect the party identification of each ability to both correctly perceive parental (Boehnke et al. 2007; Nelson and Tallman values and choose to adopt or reject those 1969). The influence of the social milieu has values. Two established aspects of the quality warranted less attention in the transmission of of parent-child relationships—discussion and party identification. McAllister and Kelley social support—are important mechanisms (1985), for example, argue that “the social facilitating value transmission through a bet- milieu of the family has little or no effect on ter understanding of the parent (i.e., percep- the preferences of the children and it is the tion) and a desire to model parents’ behaviors partisanship of the parent that is the dominant (i.e., adoption). We elucidate their role in factor.” Nevertheless, recent years have seen more detail below. renewed attention to the social milieu, par- One consistent finding in the values social- ticularly regarding one’s perceptions of one’s ization literature is that communication within own values relative to that of society’s. a household helps facilitate the transmission Boehnke and colleagues (2007), for example, of values from parents to child (McPherson, show that the transmission of values from Smith-Lovin, and Brashears 2006; Myers parent to child is stronger when families have 1996). As Hooghe and Boonen (2013:143) values that appear further from the mean of note, “talking about politics within the family preferences within a given society, but trans- context can indeed significantly increase the mission is weak-to-insignificant when family likelihood of adolescents sharing the same values appear similar to the zeitgeist. Given voting intention as (one of ) their parents.” that we are providing the baseline model and Within the context of the perception-adoption focusing on the most proximal component of model, however, these findings raise addi- value transmission (parent to child) we do not tional questions. Which step of the transmis- explicitly include zeitgeist factors in these sion process is affected by discussion? Are analyses. We do, however, offer several spe- children better at perceiving parents’ party cific and implementable suggestions in the identification because of discussion, or are conclusion for incorporating the social milieu they more likely to adopt parents’ party iden- into the baseline model articulated here. tification due to affective or other reasons? Overall, the perception-adoption model Discussion is one way individuals reveal reveals the complexity of transmission by preferences to others and thereby set the stage including perception, adoption, and the social for change (Huckfeldt and Sprague 1991). milieu common to both parents and children. Nevertheless, discussion in and of itself does It does not propose the abandonment of past not immediately suggest change in one direc- research that relied on the direct transmission tion or another; discussion could as easily model. Rather, we consider the direct trans- lead to agreement as to argument. Thus, we mission model to represent a simpler, macro further refine the importance of discussion by model of transmission, because it reduces arguing that it allows children to better know perception and adoption into a single unidirec- their parents’ political party identifications, tional trait-to-trait relationship. The perception- but it does not necessarily lead to higher rates adoption model unpacks these steps, and in so of adoption. doing, gives agency to the child in the trans- A second component of the quality of the mission process. parent-child relationship found to have an important role in child learning is parent- child affect (Nelson and Tallman 1969). One What Informs Perception measure of affect—social support—has been and Adoption? widely used in the psychological literatures to We operationalize our perception-adoption measure the quality of parent-child interac- model by focusing on how the quality of the tion. Social support includes the size and parent-child relationship affects the child’s interconnectedness of social networks, the Ojeda and Hatemi 1159 perception of being reliably connected to oth- Methods ers, and the assistance provided by others Data and Measures (Barrera 1986). We propose social support is important to the transmission process because Data come from two sources. The first data- it brings about feelings of belonging and set, the Health and Lifestyles Study (HLS), is affiliation resulting from inclusion in family a 1988 survey of 8,636 families in the United and in-group activities (Kaplan, Cassel, and States, which includes parents, adult children, Gore 1977). Positive affect between parents spouses, and other relatives. Families in the and children enhances value transmission HLS were identified through state birth (Murray and Mulvaney 2012). For example, records and national advertisements (for a full Tedin (1980) classified children into parent- description of the data, collection techniques, oriented and peer-oriented groups depending and method see Eaves, Martin, and Heath on how much affect they expressed toward 1990). Figure 3 presents a diagram of the parents and peers; parent-oriented children family members represented in this study and had stronger attitude concordance with par- their respective sample size. What makes ents than did their peer-oriented counterparts. these data unique, and particularly suited for Research on similar affiliatory behaviors pro- this study, is that respondents answered ques- vides evidence that good relationships tions about other family members, including between parents and children, especially their political identification, and the dataset those characterized by openness and warmth, includes measures of the quality of parent- lead to a higher quality of parental transmis- child interactions. Arrows in Figure 3 indicate sion of religious commitment and affiliation the direction of cross-reporting. Cross-reports (Taris and Semin 1997). These findings dem- are crucial because they reveal the percep- onstrate that affect is an important facilitator tions of each family member, which allows of value transmission. one to compare a respondent’s self-reported In contrast to parent-child communication, attributes with others’ perceptions of those we propose that positive social support facili- attributes. tates the adoption of parents’ political identi- The second dataset consists of the 2006 and fication but does not necessarily enhance 2008 waves of the National Longitudinal children’s perceptions of it. Because social Study of Youth (NLSY, Center for Human support cultivates feelings of belonging and Resource Research 2004). The NLSY data affiliation, it is likely to improve children’s initially contained 12,686 respondents; addi- identification with parents. In contrast, chil- tional interviews were conducted as the women dren who perceive little support from parents in this dataset had children, thereby allowing will have a lesser sense of belonging and for a mother-child dyad dataset to be generated cohesion with their parents. This may have (3,356 families). Several questions about poli- one of two contrasting effects. First, children tics were asked in 2006 and 2008 of both who perceive lower social support will be less mothers and children, including children’s likely to identify with parents; they may look cross-reports on parents’ party identification for social support elsewhere, which opens up and measures of parent-child communication. the possibility that their party identification The combination of datasets presents sev- will be grounded in extra-familial sources eral strengths. First, it allows us to replicate (e.g., Huckfeldt and Sprague 1991; Tedin findings. Replicating a study in all its impor- 1980). Second, children who perceive lower tant details establishes the reliability of social support may be motivated to please results. Second, both datasets survey children their parents in the hopes of cultivating feel- in adolescence, young adulthood, and adult- ings of social support; in this case, children hood, thereby capturing the full range of the may be more likely to adopt their parents’ life course. Finally, the two datasets can be party identification. used to confirm the robustness of the findings 1160 American Sociological Review 80(6)

Figure 3. Sample Sizes of Kinship Linkages across a three-decade time span—HLS data self-reported PID in 2008. Self-reported PID in are from the late 1980s and NLSY data are the NLSY follows the American National from the late 2000s.2 Election Studies seven-point scale of “strongly To assess political party identification (PID) Republican,” “Republican,” “leaning Republi- in the HLS, respondents were asked to indicate can,” “neither Republican nor Democrat,” a political affiliation for themselves, their sib- “leaning Democrat,” “Democrat,” and “strongly ling, mother, father, and spouse, where appli- Democrat,” in which higher values indicate a cable. Self-reports and cross-reports rely on strong Democratic affiliation. The cross- the same measure. Responses include five reported PID, in which children report on their options: “always support Republicans,” “usu- mother’s PID, is limited to a three-point scale ally support Republicans,” “varies,” “usually of “Republican,” “Independent,” and “Demo- supports Democrats,” and “always support crat.” To make self-reports and cross-reports Democrats.” Higher values indicate a stronger commensurable, we collapse the self-reported Democratic affiliation. Values of children’s PID into the three-point scale; all subsequent self-reported PID are distributed as follows: 9 analyses of the NLSY data use the three-point percent always support Republicans, 30 per- measure. About 31 percent of children report a cent usually support Republicans, 36 percent Republican affiliation, 27 percent report no vary, 20 percent usually support Democrats, affiliation, and 42 percent report a Democratic and 5 percent always support Democrats. Sub- affiliation. sequent analyses of the HLS data use the five- point measure of party identification unless otherwise indicated. In the NLSY, child Assessing the Perception- respondents were asked about their self- Adoption Model reported and cross-reported PID in either 2006 As an initial assessment of our hypothesis or 2008; mothers were asked about their that parent-child concordance does not Ojeda and Hatemi 1161

Table 1. Transmission Does Not Occur in a Majority of Child-Parent Relationships

HLS Sample (Child-Mother Dyads) Correctly Perceive Incorrectly Perceive Adopt 46.5% 18.5% Reject 23.3% 11.7%

HLS Sample (Child-Father Dyads) Correctly Perceive Incorrectly Perceive Adopt 45.9% 19.0% Reject 22.9% 12.3%

NLSY Sample (Child-Mother Dyads) Correctly Perceive Incorrectly Perceive Adopt 48.8% 20.2% Reject 18.0% 13.0%

Note: Values are cell percentages of children who report on the perception of their parents’ political party identification (correctly perceive versus incorrectly perceive) and the evaluation of their parents’ party identification (adopt versus reject). True transmission occurs when a child correctly perceives and adopts the parent’s party identification. Sample size for HLS is 4,963 mother-child dyads and 3,466 father-child dyads. Sample size for the NLSY is 2,048 mother-child dyads.

necessarily reflect transmission, we examine transmission occurs the majority of time—is the frequency with which children correctly not valid. With the HLS mothers and fathers, perceive and adopt parents’ party identifica- true transmission occurs about 46 percent of tion. All significance tests are two-tailed the time. The remaining 54 percent of chil- unless otherwise noted. Table 1 is a 2x2 dis- dren either misperceive or reject parents’ play of the perception and adoption steps. In party identification and transmission does not the perception step, children either correctly occur.3 The results are similar in the NLSY, or incorrectly perceive their parents’ PID. In which is composed of younger adult children. the adoption step, children either adopt or Less than half of these children correctly per- reject parents’ PID. This leads to four potential ceive and adopt the mother’s party identifica- outcomes: correctly perceive and adopt (true tion. In contrast, about 20 percent of children transmission), correctly perceive and reject adopt a misperception of their mother’s PID, (true rejection), incorrectly perceive and adopt while another 18 percent of children reject the (false transmission), and incorrectly perceive correct perception of their mother’s PID. The and reject (false rejection). True transmission first results from our two-step approach pro- occurs only when a child correctly perceives vide evidence that the transmission process is and chooses to adopt the parent’s PID. more complicated than what is captured using The results show that true transmission only trait-to-trait evidence. (i.e., correctly perceiving and adopting) fails Table 2 reports correlations between chil- to occur in a majority of parent-child relation- dren’s self-reported PID, children’s percep- ships. This finding turns the extant literature tion of parents’ PID, and parents’ self-reported on its head and requires reconsideration of PID. These correlations are labeled trait-to- how to interpret the parent-child correlations trait (child self-report and parent self-report), from previous research on political value perception (parent self-report and child-report transmission, if not the main thesis behind the on parent), and adoption (child-report on par- similarity in parent-child party affiliation. ent and child self-report). Trait-to-trait corre- That is, we find that one of the core assump- lations are .39 and .37 for mothers and fathers, tions of the trait-to-trait model—that value respectively, in the HLS, and .46 for mothers 1162 American Sociological Review 80(6)

Table 2. Correlations Indicate an Important Role for Perception in the Transmission Process

Concordant HLS Discordant HLS HLS Sample Parents Parents NLSY Sample

Mothers Fathers Mothers Fathers Mothers Fathers Mothers

Trait-to-Trait .39 .37 .48 .48 .21 .09 .46 Perception Step .70 .68 .76 .77 .47 .38 .61 Adoption Step .51 .48 .58 .55 .41 .36 .63 Observations 4,963 3,466 2,228 2,228 942 942 1,810

Note: Direct or trait-to-trait transmission is the correlation between parent’s political party identification (PID) and child’s PID, perception is the correlation between parent’s PID and child’s perception of parent’s PID, and adoption is the correlation between child’s perception of parent’s PID and child’s PID. in the NLSY. These values reflect the tradi- political affiliation more closely reflects a tional approach to studying the transmission direct transmission model, then the parent’s of PID and are consistent with previous find- self-reported PID should be the strongest pre- ings (Acock and Bengtson 1980; Jennings dictor of the child’s political affiliation. If, on and Niemi 1968).4 The perception and adop- the other hand, the child plays the more impor- tion correlations reflect the two-step learning tant role in the transmission of political affili- process, in which a child first perceives the ation, then children’s perception of their parent’s PID and then chooses to adopt or parents’ PID should be the strongest reject it. The correlations for perception and predictor. adoption (.36 to .77) are significantly larger Each model has standard errors clustered than the trait-to-trait correlations (.09 to .48) at the family level to account for the structure in both samples, indicating that the trait-to- of the data. Analyses using HLS data control trait correlations mask important learning for sex, age, educational attainment, income, processes, again providing additional evi- political attitudes, and education of parents. dence for our hypothesis. The correlations Both respondent and parental education is when parents are concordant for party identi- measured with dummy variables, one indicat- fication, defined as jointly Republican, Dem- ing a high school degree and one indicating ocratic, or Independent, are uniformly higher some college education, with no high school than when parents are discordant, akin to degree serving as the reference category. extant findings (Acock and Bengtson 1980; Income is measured with income brackets but Beck and Jennings 1991). This indicates that operationalized for the analysis as the mini- children better perceive parents’ PID and are mum value of the income bracket in incre- more likely to adopt it when parents have the ments of $10,000. Political attitudes are same self-reported party identification. measured using a Wilson and Patterson (1968) To more adequately identify the role of conservatism index, with higher values cor- perception, we use a series of logistic regres- responding to more conservative attitudes. sions that predict children’s party identifica- Race and ethnicity are not included in the tion. These models compare the influence of HLS analyses because the sample is 98 per- parent’s self-reported PID to the child’s per- cent Caucasian. Analyses using NLSY data ception of parent’s PID. The first model is the include the same covariates, but with three direct transmission model and uses parent’s exceptions. First, the model does not include self-reported PID as the main predictor. The a measure of political attitudes, as there are second model is the perception-adoption not suitable ones in the NLSY. Second, we model; it builds on the direct transmission use parental income (in increments of model by including children’s perception of $10,000) rather than respondent income, their parents’ PID. If the transmission of because the sample consists primarily of Ojeda and Hatemi 1163

Table 3. Characteristics of the HLS and NLSY Samples

HLS Mothers HLS Fathers NLSY Mothers

Mean SD Mean SD Mean SD

Child Characteristics Black .21 Hispanic .08 Age 34.71 9.46 33.31 7.93 24.81 3.42 Male .43 .44 .52 High school degree .17 .14 .72 Some college + .80 .84 .19 Income 17,662 14,024 17,776 13,848 Attitudes 20.62 7.78

Parent Characteristics High school degree .35 .23 .77 Some college + .51 .61 .16 Income ($) 75,209 72,430

Observations 4,963 3,466 1,810

young adults, many living at their parents’ fathers represent the change in the probability home, for whom parental income is a better of claiming to “usually support Democrats” indicator of economic status. Third, we con- given a one standard deviation change in the trol for race by including dummy variables independent variable, when all other varia- for Black and Hispanic, with Caucasian serv- bles are held at their median value. A one ing as the reference category. The models use standard deviation change in the child’s per- sample weights to correct for the oversam- ception is associated with a 15 percent pling of Blacks, Hispanics, and the economi- increase (HLS mothers model), a 14 percent cally disadvantaged. Table 3 lists descriptive increase (HLS fathers model), and a 23 per- statistics for these variables. cent increase (NLSY mothers model) in the Table 4 reports results of the analyses for child reporting a Democratic affiliation.5 HLS mothers and fathers as well as NLSY These findings strongly suggest a mediating mothers. The direct transmission models con- effect of perception between parents’ self- firm the notion that parents’ PID is a strong reported PID and child’s self-reported PID. predictor of children’s political affiliation. We conduct mediation models and sensi- Children are more likely to report a strong tivity analyses as alternative modeling strate- Democratic affiliation when parents report a gies and report results in the online supplement strong Democratic affiliation, and likewise (http://asr.sagepub.com/supplemental). These with Republican affiliations. However, our results mirror findings from the main analy- perception-adoption models show that chil- ses by showing that the influence of parent dren’s perception of their parents’ PID is the PID occurs almost exclusively through the most important predictor of transmission. child’s perception. Because transmission of These models reveal that the actual PIDs of party identification from parent to child mothers and fathers offer very little additional occurs primarily during adolescence and benefit in explaining transmission when com- young adulthood, when individuals first pared to the influence of the child’s percep- encounter the political world and their place tion. This is seen most explicitly by examining in it, we conduct additional analyses by the first differences reported in Table 4. The restricting the sample to adolescents and first differences for the HLS mothers and young adults (i.e., respondents age 16 to 24).

0% 3% 4% 0% 7% 4% –3% –2% –9% 10% 23% First Differences * * * * * .997 .024

.857 .888 .996 .641

(.01) (.33) (.22) (.16) (.02) (.08) (.23) (.28) (.46) (.13) 1,810 1,216 1.161 1.213 1.484 1.860 4.141 1.298 Pathway Perceptual NLSY Mothers * * * * .999 .18 .986

.984 .641

Trait 1,810 1,216 (.01) (.28) (.92) (.88) (.02) (.08) (.24) (.38) (.23) 1.012 1.026 1.027 1.638 2.512 2.527 Trait-to- 3% 2% First –6% –1% –3% –6% –2% 14% –15% –11% Differences * * * * * * .941 .929 .873 .868 .512 .640 .891 3,466 1,824 (.10) (.11) (.00) (.03) (.12) (.15) (.00) (.06) (.05) 1.024 2.337 1.168 Pathway Perceptual HLS Fathers * * * * * * .863 .886 .874 .868 .430 .555 .908 Trait 3,466 1,824 (.09) (.10) (.00) (.02) (.11) (.14) (.00) (.07) (.07) 1.025 1.989 Trait-to- 0% 2% 1% First –2% –1% –3% –4% –1% 15% –13% Differences * * * * * .892 .895 .887 .841 .785 .974 .942 (.09) (.09) (.00) (.02) (.15) (.19) (.00) (.06) (.13) (.04) 4,963 2,592 1.019 2.580 1.094 Pathway Perceptual HLS Mothers * * * * .867 .863 .885 .850 .668 .873 .957 Trait 4,963 2,592 (.08) (.08) (.00) (.02) (.14) (.19) (.00) (.06) (.06) 1.021 1.949 Trait-to- PID Some college + High school degree Political attitudes Some college + High school degree p < .05 (two-tailed). Observations Clusters

Income

Parent Characteristics

Income

Age Male

Hispanic

Black Child Characteristics Child Perception of Parent

The dependent variable is child’s self-reported party identification. Models are ordered logistic regressions, cell entries odds ratios, and standard errors Note : The dependent variable is child’s are reported in parentheses. First differences percentage point changes the probability of reporting “usually supports Democrats” (HLS) or a Democratic affiliation (NLSY) given a one standard deviation change in the continuous variables or 0 to 1 dichotomous when all other are held at their median. * . The Perceptual Pathway Model Better Explains the Transmission of Party Identification 4 . The Perceptual Pathway Model Better Explains the Transmission Table Parent Self-Reported PID

1164 Ojeda and Hatemi 1165

Results using this restricted sample are con- party identification to improve: political dis- sistent (i.e., same direction and significance) cussion and social support. Here, we use with results of the full sample (see the online regression analysis to determine which step supplement). of the transmission process is affected by Until this point, our models treat parental these aspects of parent-child relationships. influences as independent of one another. Do children better perceive parent PID when Additional insight into the transmission pro- exposed to discussion or social support? Or cess can be gained by simultaneously mode- are children more likely to adopt perceived ling the effect of both parents. Modeling parent PID when exposed to discussion or parents together allows us to parse out the social support? Or do these factors operate on relative influence of mothers and fathers and both steps of the transmission process? To the effect they have on each other’s influence, answer these questions, we run three regres- an issue that has long been a concern in the sion models: a trait-to-trait model, a percep- study of political socialization. Figure 4 pre- tion model, and an adoption model. sents a structural model based on the assumed causal pathways. In this model, the PID of Trait-to-trait model: The dependent variable both parents may directly inform the child’s is the child’s self-reported party identifica- PID. This path is indicated by arrows from tion; the main independent variables are par- the mother’s and father’s PIDs to the child’s ents’ self-reported PID, discussion (or social PID. The child’s perception of parents’ PIDs support), and an interaction between parent is also informed by the parents’ PIDs. These PID and discussion (or social support). paths are indicated by arrows from mother’s and father’s PIDs to the child’s image of Perception model: The dependent variable mother and father. The child’s adoption of is the child’s perception of parent PID; the parental PID is indicated by arrows from the main independent variables are child’s per- child’s image of mother and father to the ception, discussion (or social support), and child’s PID. Child attributes, including sex, an interaction of the two. age, education, income, and political atti- tudes, are included in the model, indicated by Adoption model: The dependent variable is arrows from child attributes to child’s percep- the child’s self-reported PID; the main inde- tions and self-reported PID. For this analysis, pendent variables are parents’ PID, child’s we use only HLS data, which have measures perception, discussion (or social support), for both mothers and fathers. We conducted and an interaction between child’s percep- analyses using the SEM procedure in Stata tion and discussion (or social support). 12.0 and standard errors are clustered by fam- ily. Figure 4 shows that pathways from the Because each dataset contains only one of mother’s and father’s PID to the child’s PID these measures, we use NLSY data for the are not statistically significant after both par- analysis of political discussion and HLS data ents are included in the model. These results for the analysis of social support. All the con- illustrate the importance of the child’s percep- trol variables from the models in Table 4 are tion in the transmission process. That is, the included in these analyses. most important causal pathways are parents’ The political discussion hypothesis articu- PIDs on the child’s perception and the child’s lated earlier holds that discussion enhances perceptions on the child’s adoption. children’s perception of parents’ PID but will not necessarily lead to a higher rate of adop- tion. Political discussion is measured by ask- Informing Perception and Adoption ing respondents, “When you were growing Earlier we highlighted two mechanisms up, how often did you hear the adults in your through which we expect transmission of household talking about politics?” Response 1166 American Sociological Review 80(6)

Figure 4. The SEM Models Both Parents at the Same Time and Shows the Importance of the Child’s Perception in the Transmission Process Note: PID = political party identification. *p < .05 (two-tailed). options include “never” (about 22 percent of Perceived Social Support scale (Kessler et al. respondents), “once in a while” (48 percent), 1992): (1) How much do your parents listen to “moderately often” (18 percent), “very often” you if you need to talk about your worries or (8 percent), and “extremely often” (4 per- problems? (2) How much do your parents cent). Table 5 reports these results. The sig- understand the way you feel and think about nificant interaction term in the trait-to-trait things? (3) How much could you count on model suggests that discussion enhances the your parents to lend you a few hundred dollars overall transmission of party identification. if you really needed it? Response options Nevertheless, a trait-to-trait approach con- range from 0 to 4 and are “not at all,” “a lit- ceals the two-step process that underlies the tle,” “some,” “quite a bit,” and a “great deal.” transmission of PID and the possibility that The summed scale ranges from 0 to 12, where discussion has heterogeneous effects on per- 0 indicates no support and 12 indicates a ception and adoption. Indeed, the perception great deal of support (mean = 8.59, SD = 2.52, model shows that political discussion signifi- α = .78). cantly increases children’s ability to perceive Table 6 reports results of this analysis. The parent PID—the interaction term indicates significant interaction term in the trait-to-trait that mother’s self-reported PID becomes a model indicates that parent-child concordance stronger predictor of children’s perception as grows stronger as social support increases. political discussion increases. On the other The question is whether we are observing this hand, the adoption model shows that political effect because social support improves per- discussion has no effect on whether children ception, adoption, or both. The non-significant adopt their perception of mother’s PID. interaction terms in the perception models The social support hypothesis suggests that show that social support does not enhance the social support leads to higher rates of adoption degree to which parent PID predicts chil- but will not necessarily increase children’s dren’s perceptions. In other words, children ability to correctly perceive parents’ PID. who report high social support are no more Social support from parents was measured by likely to know their parents’ PIDs than are the sum of three questions from the Kessler children who report low social support. The Ojeda and Hatemi 1167

Table 5. Discussion Affects Children’s Ability to Perceive Parent PID but Not the Choice to Adopt or Reject Parent PID

Model: Trait-to-Trait Perception Adoption

Dependent Variable: Child PID Child Perception Child PID

Parent Self-reported PID 1.764* 2.746* 1.277* (.24) (.46) (.13) Child Perception of Parent PID 3.693* (.56) Discussion about Politics .517* .464* .771 (.10) (.12) (.15) Discussion x Parent PID 1.323* 1.442* (.11) (.16) Discussion x Child Perception 1.097 (.09) Child Characteristics Black 2.563* 2.379* 1.875* (.39) (.40) (.28) Hispanic 1.668* 1.409* 1.477* (.25) (.23) (.23) Male .632* .889 .637* (.08) (.12) (.08) Age .981 .973 .994 (.19) (.02) (.02) High school degree .999 1.296 .878 (.19) (.30) (.16) Some college + 1.025 1.462 .854 (.25) (.43) (.22) Parent Characteristics High school degree .968 .790 1.222 (.18) (.18) (.24) Some college + 1.076 1.061 1.217 (.30) (.33) (.35) Income .986 .966* .997 (.01) (.01) (.01)

Note: Data come from the NLSY. Sample size is 1,810. Models are ordered logistic regressions, cell entries are odds ratios, and standard errors are reported in parentheses. *p < .05 (two-tailed). significant interaction term in the adoption of their parents’ political party identification model, however, provides evidence in favor drives adoption, and this perception improves of our hypothesis that children are more likely when parents discuss politics with their chil- to adopt their perception of parents’ PIDs as dren, and parents’ PID is likelier to be adopted social support increases. when children receive greater social support from their parents. That is not to say that parents’ actual parti- Discussion san values do not play a critical role; they do The findings reveal that children’s perception and they are the main source informing what of parental values is a critically important children perceive. Parental communication determinant of political identification. The use significantly influences children’s perceptions, of two unique datasets with cross-reporting but perception is still influenced by parents’ provides evidence that children’s perceptions actual partisan values, even in households * * * * * * * * .939 .924

.874 .871 .564 .682 .873

.825 (.10) (.11) (.00) (.03) (.13) (.17) (.01) (.07) (.02) (.03) (.19) (.05) 1.021 1.060 1.408 1.174 Adoption Child PID * * * * .808 .900 .984 .987 .577 .610 .994 (.09) (.11) (.00) (.03) (.16) (.18) (.01) (.08) (.01) (.04) (.64) 1.008 1.014 1.019 4.944 Perception HLS Fathers Child Perception * * * * * * * * .864 .882 .874 .869 .478 .614 .897 .885 (.09) (.11) (.00) (.03) (.12) (.16) (.01) (.07) (.01) (.03) (.18) 1.022 1.033 1.497 Child PID Trait-to-Trait * * * * * * .905 .899 .887 .849 .820 .993 .949 .823 (.09) (.09) (.00) (.02) (.16) (.20) (.00) (.06) (.01) (.03) (.18) (.04) 1.015 1.063 1.545 1.089 Adoption Child PID * * * .925 .909 .978 .992 .732 .816 .999 (.10) (.10) (.00) (.02) (.16) (.18) (.00) (.07) (.01) (.03) (.48) 1.008 1.032 1.003 4.832 Perception HLS Mothers Child Perception * * * * * * * .885 .877 .885 .854 .684 .877 .971 .903 (.09) (.09) (.00) (.02) (.15) (.19) (.23) (.06) (.01) (.03) (.14) 1.018 1.029 1.518 Child PID Trait-to-Trait Some college + High school degree Political attitudes Some college + High school degree p < .05 (two-tailed). Data come from the HLS; sample size is 4,796 mother-child dyads and 3,369 father-child dyads. Models are ordered logistic regressions, cell entries odds dyads and 3,369 father-child Note : Data come from the HLS; sample size is 4,796 mother-child ratios, and standard errors are reported in parentheses. *

Parent Characteristics

Income

Age

Child Characteristics Male

Social Support x Child Perception

Social Support x Parent PID

Social Support

Child Perception of Parent PID

Table 6 . Social Support Increases the Probability Children Adopt Parents’ PID, but Not Their Ability to Correctly Perceive It Table Parent Self-reported PID Dependent Variable: Model:

1168 Ojeda and Hatemi 1169 where little to no political discussion occurs. In of the role of the child. By empirically testing this way, our approach and findings provide a extant assumptions, we have provided evi- more comprehensive understanding of how dence that they are not valid, opening the parents matter than is offered by considering door to explore any number of topics. Indeed, only the parent’s party identification: parents the role of perception and adoption has impor- create the information environment in which tance beyond the transmission of party identi- children are embedded. Discussing politics fication. Religion and religiosity, social and leads to better-informed perceptions within a cultural identities, and social and political given information environment, but regardless attitudes, are just some examples of the social of parental communication, considerable indi- classifications that are affected by how indi- vidual differences emerge between children’s viduals perceive one another. Negative social perceptions. Children must detect and utilize attitudes, and larger cultural patterns that are the information provided to them and evaluate unwelcome, such as racism, sexism, and that information to determine their own politi- homophobia, are often perpetuated through cal identification. Some children accurately parent-child socialization. Exploring the role perceive their parents’ party identification, of the child in the socialization of these atti- regardless of what parents do, while others do tudes has the potential to contribute to their not. Likewise, some children are motivated to eradication. adopt what they perceive their parents’ identi- The data have many strengths: the combina- fication to be, whereas other children are moti- tion of the large number of families in the HLS, vated to identify with a different political party. the modernity and representativeness of the Children who experience greater social sup- NLSY, and the ability to compare the role of port are more strongly motivated to adopt the perception across three decades and diverse political affiliation they believe their parents to social contexts with differing levels of political have, but they may misperceive what their polarization is invaluable. Nevertheless, all parents believe and thus more strongly adopt data remain fallible, and there are limitations in the opposite values of their parents without the data and model. Model results may be dif- knowing it. Thus, social support and the moti- ferent across different political, sociological, or vation to adopt do not necessarily lead to true cultural contexts. For example, it is hard to say transmission. how well the model travels from the United We find children are also much more likely States to other comparative contexts that have to correctly perceive and adopt parental values less political polarization and different political when parents are concordant, consistent with systems. The same could be said with respect to past research (Bengtson 1975; Glass et al. time; these data cover an unusually large span 1986). There are two possible reasons for this for social scientific research but are still histori- effect. The first is that parent concordance cally situated. Data from the 1960s or some leads to a more consistent information envi- future time may produce different results due to ronment, thereby making it easier for children a host of factors, such as party polarization or to detect and utilize the relevant information changes in technology and communication that in the perception process. Another explana- affect the flow of information. tion, however, is that discordant parents avoid Looking beyond the parent-child relation- discussing politics because it is a source of ship by collecting data focusing on the larger tension, in which case children of discordant social milieu is a natural extension to the base- parents would receive less or inconsistent line model offered here and would address information from their parents. some of these limitations. Because the social Perceptual approaches to modeling social- milieu potentially alters the parent-child link ization have been slow to progress, in large by affecting both parents and children, incor- part, because the assumptions contained in porating specific manifestations of the macro unidirectional parent-driven approaches pre- and micro social context—including those of vent both theoretical and empirical inclusion the school, peer group, church, media, or 1170 American Sociological Review 80(6) political unit—may further inform the parent- somehow be influential when it occurs. For child perception-adoption model. Indeed, example, McDevitt and Chaffee (2002) show some environments may augment children’s that children can change the structure of fam- perceptual accuracy. Communities with high ily conversations about politics and prompt levels of political competition, for example, parents to increase their civic competence. may provide an information-rich environment Socio-emotional and cognitive aspects of where political divisions are deeply entrenched development are also probable mechanisms and more visible to children. In contrast, chil- in facilitating individual differences in the dren who attend schools where civics educa- perception and adoption of party identifica- tion is not emphasized may not learn about tion, and yet little is known about how exactly party politics, thereby reducing their ability to these differences factor into the transmission detect and categorize the cues parents provide of political values. Future research should about political affiliation. Other environments reexamine the effects of value salience and may also increase or decrease the rate at which family communication patterns on each step children adopt parents’ party identification. of the process, use more measures on the For example, when more facets of children’s quality of parent-child relationships (includ- social milieu are consistent with their parents’ ing parental bonding), and include educa- values, then children are not provided with a tional aspects, cultural features of value socializing agent that offers an alternative transmission, and other dispositional influ- viewpoint. Socially conservative parents who ences, such as children’s personality, emo- move into a more conservative community tional condition, and cognitive ability. and send their children to a socially conserva- The perception-adoption model can also be tive school, for instance, would reduce their extended to move beyond parent-child and children’s exposure to diverse ideas, poten- political socialization processes. For example, it tially dampening the possibility that the chil- can be used to characterize how norms, from dren opt not to adopt the parents’ values. larger social values like equality to specific These are only a few possibilities. taboos about how to dress, are circulated in an array of social networks, such as sibling and peer groups, office places, schools, and social Conclusions media. Current models of how beliefs, attitudes, Models that include children’s perceptions and behaviors spread within social networks are critical to understanding the transmission often rely on trait-to-trait analyses. However, if of political values. Considering a two-step the spread of ideas in a network is at all similar process of perception and adoption that to the socialization of political identification, underlies transmission may challenge the then perception and adoption have the potential majority of research that relies on trait-to-trait to further inform or complicate past findings on evidence, but in so doing it opens up count- the importance of social networks. Further less possibilities for future research. There is unpacking the perception-adoption model, and now reason to revisit the findings and data children’s role in this process, is an important from previous studies and include measures step in addressing these considerations. of cross-reporting in future work. The next step is to explore in detail bidirectional influ- Acknowledgments ences, and the environments, traits, and social We thank Lindon Eaves, Nicholas Martin, Andrew conditions that affect the perception and Heath, and Kenneth Kendler for granting access to the adoption process in the transmission of politi- Virginia 30,000 study. Data collection was supported by cal values. With respect to bidirectional influ- the National Institutes of Health and the National Insti- ence, parents’ party identifications are not tute of Child Health and Human Development (AA- fixed, and while party switching is an uncom- 06781 and MH-40828). We also thank the U.S. Bureau of Labor Statistics and the U.S. Department of Labor for mon phenomenon, it is not implausible that access to the National Longitudinal Surveys (https:// either adolescent or adult children could www.nlsinfo.org/). Ojeda and Hatemi 1171

Notes over the Life Span.” American Journal of Sociology 97(1):169–95. 1. The term “agency” has different theoretical founda- Atkin, Charles K., and Walter Gantz. 1978. “Television tions, meanings, and practical applications depend- News and Political Socialization.” Public Opinion ing on discipline (Hitlin and Elder 2007). We refer Quarterly 42(2):183–94. to agency as a sense of individual volition that can Baldassarri, Delia, and Andrew Gelman. 2008. “Partisans lead individuals, often unconsciously, to differen- without Constraint: Political Polarization and Trends tially perceive, produce, and reproduce their social in American Public Opinion.” American Journal of world. This broad definition surrounds some level Sociology 114(2):408–446. of personal causality in the evaluative, experien- Bao, Wan-Ning, Les B. Whitbeck, Danny R. Hoyt, and D. tial, and constructive dimensions of perception and Conger Rand. 1999. “Perceived Parental Acceptance behavior, within the context of social experience as a Moderator of Religious Transmission among (Emirbayer and Mische 1998). Adolescent Boys and Girls.” Journal of Marriage 2. Alternatively, these datasets could be used to com- and Family 61(2):362–74. pare temporal changes in the transmission process. Barrera, Manuel, Jr. 1986. “Distinctions between Social Corbetta, Tuorto, and Cavazza (2013), for example, Support Concepts, Measures, and Models.” American compare datasets from different decades to study Journal of Community Psychology 14(4):413–45. changes in ideological similarity between Italian Beck, Paul Allen, and M. Kent Jennings. 1991. “Family Tra- parents and children from 1875 to 2010. ditions, Political Periods, and the Development of Parti- 3. Frequencies are generated from a three-category san Orientations.” Journal of Politics 53(3):742–63. measure of PID (i.e., Republican, Independent, and Belsky, Jay. 1984. “The Determinants of Parenting: A Democrat). Thus, a child who claims a parent “usu- Process Model.” Child Development 55(1):83–96. ally supports Republicans” when the parent reports Bengtson, Vern L. 1975. “Generation and Family Effects that they “strongly support Republicans” would be in Value Socialization.” American Sociological classified as correctly perceiving the parent’s PID. Review 40(3):358–71. In this sense, the frequencies are a more generous Beutel, Ann M., and Margaret Mooney Marini. 1995. assessment of the rate of perception and adoption. “Gender and Values.” American Sociological Review 4. Treating the three-point scale in the NLSY as a cat- 60(3):436–48. egorical variable returns similar results. The largest Block, Jack, and Jeanne H. Block. 2006. “Nursery chi-square value (i.e., the strongest association) is School Personality and Political Orientation Two for perception, then adoption, and then trait-to-trait. Decades Later.” Journal of Research in Personality 5. It is possible that the three-point PID scale in the 40(5):734–49. NLSY data should be treated as categorical, in which Boehnke, Klaus. 2001. “Parent-Offspring Value Trans- case a multinomial logistic regression would be mission in a Societal Context: Suggestions for a appropriate. Results of a multinomial logistic regres- Utopian Research Design—with Empirical Under- sion are consistent with those reported in Table 4. pinnings.” Journal of Cross-Cultural Psychology 32(2):241–55. Boehnke, Klaus, Andreas Hadjar, and Dirk Baier. 2007. References “Parent-Child Value Similarity: The Role of Zeit- Abendschön, Simone. 2013. Growing into Politics: Con- geist.” Journal of Marriage and Family 69(3):778–92. texts and Timing of Political Socialisation. Essex, Brooks, Clem, and Jeff Manza. 1997. “Social Cleavages UK: ECPR Press. and Political Alignments: U.S. Presidential Elec- Achen, Christopher H. 2002. “Parental Socialization and tions, 1960 to 1992.” American Sociological Review Rational Party Identification.” Political Behavior 62(6):937–46. 24(2):151–70. Calarco, Jessica McCrory. 2014. “Coached for the Class- Acock, Alan C., and Vern L. Bengtson. 1980. “Socializa- room: Parents’ Cultural Transmission and Children’s tion and Attribution Processes: Actual versus Per- Reproduction of Educational Inequalities.” American ceived Similarity among Parents and Youth.” Journal Sociological Review 79(6):1015–37. of Marriage and Family 42(3):501–515. Campbell, Angus, Philip E. Converse, Warren E. Miller, Alford, John R., Peter K. Hatemi, John R. Hibbing, and Donald E. Stokes. 1960. The American Voter. Nicholas G. Martin, and Lindon J. Eaves. 2011. : Wiley. “The Politics of Mate Choice.” Journal of Politics Cashmore, Judith A., and Jacqueline J. Goodnow. 1985. 73(2):362–79. “Agreement between Generations: A Two-Process Alwin, Duane F., Ronald L. Cohen, and Theodore M. Approach.” Child Development 56(2):493–501. Newcomb. 1991. Political Attitudes over the Life Cavior, Norman, Karen Miller, and Stanley H. Cohen. Span: The Bennington Women after Fifty Years. Mad- 1975. “Physical Attractiveness, Attitude Similar- ison: University of Wisconsin Press. ity, and Length of Acquaintance as Contributors to Alwin, Duane F., and Jon A. Krosnick. 1991. “Aging, Interpersonal Attraction among Adolescents.” Social Cohorts, and the Stability of Sociopolitical Orientations Behavior and Personality 3(2):133–42. 1172 American Sociological Review 80(6)

Center for Human Resource Research (CHRR). 2004. Harris, Judith Rich. 1995. “Where Is the Child’s Envi- “NLSY79 Users Guide.” Columbus, OH: U.S. ronment? A Group Socialization Theory of Develop- Department of Labor, Bureau of Labor Statistics. ment.” Psychological Review 102(3):458–89. Corbetta, Piergiorgio, Dario Tuorto, and Nicoletta Haste, Helen, and Judith Torney-Purta. 1992. The Devel- Cavazza. 2013. “Parents and Children in the Political opment of Political Understanding. San Francisco: Socialisation Process: Changes in Italy over Thirty- Jossey Bass. Five Years.” Pp. 11–32 in Growing into Politics: Con- Haug, Lena. 2013. “A Picture Paints a Thousand Words: texts and Timing of Political Socialisation, edited by Children’s Drawings as a Medium to Study Early S. Abendschön. Essex, UK: ECPR Press. Political Socialisation.” Pp. 231–72 in Growing into Corsaro, William A., and Laura Fingerson. 2006. “Devel- Politics: Contexts and Timing of Political Socialisation, opment and Socialization in Childhood.” Pp. 125–56 edited by S. Abendschön. Essex, UK: ECPR Press. in Handbook of , edited by J. Dela- Haynes, Stephen E., and David Jacobs. 1994. “Macroeco- mater. New York: Springer. nomics, Economic Stratification, and Partisanship: A Dalton, Russell J. 1980. “Reassessing Parental Social- Longitudinal Analysis of Contingent Shifts in Politi- ization: Indicator Unreliability Versus Genera- cal Identification.” American Journal of Sociology tional Transfer.” American Political Science Review 100(1):70–103. 74(2):421–31. Hess, Robert D., and Judith Torney-Purta. 1967. The Dey, Eric L. 1997. “Undergraduate Political Attitudes: Development of Political Attitudes in Children. Chi- Peer Influence in Changing Social Contexts.” Journal cago: Aldine Publishing. of Higher Education 68(4):398–413. Hetherington, E. Mavis, David Reiss, and Robert Plomin. Easton, David, Jack Dennis, and Sylvia Easton. 1969. 2013. Separate Social Worlds of Siblings: The Impact Children in the Political System: Origins of Political of Nonshared Environment on Development. Hills- Legitimacy. New York: McGraw-Hill. dale, NJ: Routledge. Eaves, Lindon, Nicholas Martin, and Andrew Heath. Hitlin, Steven, and Glen H. Elder. 2007. “Time, Self, and 1990. “Religious Affiliation in Twins and Their the Curiously Abstract Concept of Agency.” Socio- Parents: Testing a Model of Cultural Inheritance.” logical Theory 25(2):170–91. Behavior Genetics 20(1):1–22. Hooghe, Marc, and Joris Boonen. 2013. “The Inter- Emirbayer, Mustafa, and Ann Mische. 1998. “What generational Transmission of Voting Intentions in a Is Agency?” American Journal of Sociology Multiparty Setting: An Analysis of Voting Intentions 103(4):962–1023. and Political Discussion among 15-Year-Old Adoles- Erikson, Erik H. 1968. Identity: Youth and Crisis. New cents and Their Parents in Belgium.” Youth & Society York: WW Norton & Company. 47(1):125–47. Fraley, R. Chris, Brian N. Griffin, Jay Belsky, and Glenn Huckfeldt, Robert. 1979. “Political Participation and the I. Roisman. 2012. “Developmental Antecedents of Neighborhood Social Context.” American Journal of Political Ideology: A Longitudinal Investigation Politcal Science 23(3):579–92. From Birth to Age 18 Years.” Psychological Science Huckfeldt, Robert, and John Sprague. 1991. “Discus- 23(11):1425–31. sant Effects on Vote Choice: Intimacy, Structure, and Franklin, Mark N., Thomas T. Mackie, and Henry Valen. Interdependence.” Journal of Politics 53(1):122–58. 1992. Electoral Change. New York: Cambridge Uni- Hyman, Herbert. 1959. Political Socialization. Glencoe, versity Press. IL: Free Press. Funder, David C. 1995. “On the Accuracy of Personal- Inglehart, Ronald. 1990. Culture Shift in Advanced Indus- ity Judgment: A Realistic Approach.” Psychology trial Societies. Princeton, NJ: Princeton University Review 102(4):652–70. Press. Gauchat, Gordon. 2012. “Politicization of Science in the Iyengar, Shanto, and Sean J. Westwood. 2015. “Fear and Public Sphere: A Study of Public Trust in the United Loathing across Party Lines: New Evidence on Group States, 1974 to 2010.” American Sociological Review Polarization.” American Journal of Politcal Science 77(2):167–87. 59(3):690–707. Glass, Jennifer, Vern L. Bengtson, and Charlotte Chorn Janowitz, Morris, and David R. Segal. 1967. “Social Dunham. 1986. “Attitude Similarity in Three-Gen- Cleavage and Party Affiliation: Germany, Great eration Families: Socialization, Status Inheritance, Britain, and the United States.” American Journal of or Reciprocal Influence?” American Sociological Sociology 72(6):601–618. Review 51(5):685–98. Jennings, M. Kent, and Gregory B. Markus. 1984. “Par- Greenstein, Fred. 1965. Children and Politics. New tisan Orientations over the Long Haul: Results from Haven, CT: Yale University Press. the Three-Wave Political Socialization Panel Study.” Grolnick, Wendy S., Richard M. Ryan, and Edward L. American Political Science Review 78(4):1000–1018. Deci. 1991. “Inner Resources for School Achieve- Jennings, M. Kent, and Richard G. Niemi. 1968. ment: Motivational Mediators of Children’s Per- “The Transmission of Political Values from Par- ceptions of Their Parents.” Journal of Educational ent to Child.” American Political Science Review Psychology 83(4):508–517. 62(1):169–84. Ojeda and Hatemi 1173

Jennings, M. Kent, and Richard G. Niemi.1981. Genera- Lewis-Beck, Michael S., William G. Jacoby, Helmut tions and Politics: A Panel Study of Young Adults and Norpoth, and Herbert F. Weisberg. 2008. The Ameri- Their Parents. Princeton, NJ: Princeton University can Voter Revisited. Ann Arbor: University of Michi- Press. gan Press. Jennings, Kent M., Laura Stoker, and Jake Bowers. 2009. Lopreato, Joseph. 1967. “Upward Social Mobility and “Politics across Generations: Family Transmission Political Orientation.” American Sociological Review Reexamined.” Journal of Politics 71(3):782–99. 32(4):586–92. Kalmijn, Matthijs. 1998. “Intermarriage and Homogamy: Maccoby, E. E., and J. A. Martin. 1983. “Socialization in Causes, Patterns, Trends.” Annual Review of Sociol- the Context of the Family: Parent-Child Interaction.” ogy 24:395–421. Pp. 1–101 in Handbook of Child Psychology: Social- Kaplan, Berton H., John C. Cassel, and Susan Gore. ization, Personality and Social Development, Vol. 4, 1977. “Social Support and Health.” Medical Care edited by E. M. Hetherington. New York: Wiley. 15(5):47–58. Manza, Jeff, and Clem Brooks. 1997. “The Religious Kessler, Ronald C., Kenneth S. Kendler, Andrew Heath, Factor in U.S. Presidential Elections, 1960–1992.” Michael C. Neale, and Lindon J. Eaves. 1992. “Social American Journal of Sociology 103(1):38–81. Support, Depressed Mood, and Adjustment to Stress: Mayall, Berry, and Helga Zeiher. 2003. Childhood in A Genetic Epidemiologic Investigation.” Journal of Generational Perspective. London, UK: Institute of Personality and Social Psychology 62(2):257–72. Education, University of London. Knafo, Ariel, and Avi Assor. 2007. “Motivation for McAllister, Ian, and Jonathan Kelley. 1985. “Party Iden- Agreement with Parental Values: Desirable When tification and Political Socialization: A Note on Aus- Autonomous, Problematic When Controlled.” Moti- tralia and Britain.” European Journal of Political vation and Emotion 31(3):232–45. Research 13(1):111–18. Knafo, Ariel, and Shalom H. Schwartz. 2003. “Parent- McAuley, William J., and Cheri L. Nutty. 1982. “Resi- ing and Adolescents’ Accuracy in Perceiving Parental dential Preferences and Moving Behavior: A Family Values.” Child Development 74(2):595–611. Life-Cycle Analysis.” Journal of Marriage and the Knafo, Ariel, and Shalom H. Schwartz. 2004. “Identity Family 44(2):301–309. Formation and Parent-Child Value Congruence in McDevitt, Michael. 2006. “The Partisan Child: Devel- Adolescence.” British Journal of Developmental Psy- opmental Provocation as a Model of Political Social- chology 22(3):439–58. ization.” International Journal of Public Opinion Knoke, David. 1972. “A Causal Model for the Political Research 18(1):67–88. Party Preferences of American Men.” American Soci- McDevitt, Michael, and Steven Chaffee. 2002. “From ological Review 37(6):679–89. Top-Down to Trickle-Up Influence: Revisiting Knoke, David, and Richard B. Felson. 1974. “Ethnic Assumptions about the Family in Political Socializa- Stratification and Political Cleavage in the United tion.” Political Communication 19(3):281–301. States, 1952–68.” American Journal of Sociology McLeod, Jack M. 2000. “Media and Civic Socialization 80(3):630–42. of Youth.” Journal of Adolescent Health 27(2):45–51. Knoke, David, and Michael Hout. 1974. “Social and McPherson, Miller, Lynn Smith-Lovin, and Matthew Demographic Factors in American Political Party E. Brashears. 2006. “Social Isolation in America: Affiliations, 1952–72.” American Sociological Changes in Core Discussion Networks over Two Review 39(5):700–713. Decades.” American Sociological Review 71(3):353– Kohn, Melvin L. 1983. “On the Transmission of Values in 75. the Family: A Preliminary Reformulation.” Research Merelman, Richard M. 1986. Making Something of Our- in Sociology of Education and Socialization 4(1):1–12. selves. Berkeley: University of California Press. Kohn, Melvin L., Kazimierz M. Slomczynski, and Car- Merriam, Charles Edward. 1931. The Making of Citizens. rie Schoenbach. 1986. “Social Stratification and the Chicago: University of Chicago Press. Transmission of Values in the Family: A Cross-national Middleton, Russell, and Snell Putney. 1963. “Politi- Assessment.” Sociological Forum 1(1):73–102. cal Expression of Adolescent Rebellion.” American Koleva, Spassena P., and Blanka Rip. 2009. “Attachment Journal of Sociology 68(5):527–35. Style and Political Ideology: A Review of Contradictory Miller, Richard B., and Jennifer Glass. 1989. “Parent- Findings.” Social Justice Research 22(2–3):241–58. Child Attitude Similarity across the Life Course.” Krueger, Brian S. 2006. “A Comparison of Conventional Journal of Marriage and Family 51(4):991–97. and Internet Political Mobilization.” American Poli- Morrow, Virginia. 2002. “Perspectives on Children’s tics Research 34(6):759–76. Agency within Families.” Pp. 109–129 in Handbook Langton, Kenneth P. 1969. Political Socialization. New of Dynamics in Parent-Child Relations, edited by L. York: Oxford. Kuczynski. Thousand Oaks, CA: Sage. Langton, Kenneth P., and M. Kent Jennings. 1968. Murray, Gregg R., and Matthew K. Mulvaney. 2012. “Political Socialization and the High School Civics “Parenting Styles, Socialization, and the Transmis- Curriculum in the United States.” American Political sion of Political Ideology and Partisanship.” Politics Science Review 62(3):852–67. & Policy 40(6):1106–30. 1174 American Sociological Review 80(6)

Myers, Scott M. 1996. “An Interactive Model of Reli- Transmission of Moral Values.” Journal of Moral giosity Inheritance: The Importance of Family Con- Education 26(2):211–21. text.” American Sociological Review 61(5):858–66. Tedin, Kent L. 1974. “The Influence of Parents on the Nasatir, David. 1968. “A Note on Contextual Effects and Political Attitudes of Adolescents.” American Politi- the Political Orientation of University Students.” cal Science Review 68(4):1579–92. American Sociological Review 33(2):210–19. Tedin, Kent L. 1980. “Assessing Peer and Parent Influ- Nelson, Joel I., and Irving Tallman. 1969. “Local-Cosmo- ence on Adolescent Political Attitudes.” American politan Perceptions of Political Conformity: A Speci- Journal of Politcal Science 24(1):136–54. fication of Parental Influence.” American Journal of Thompson, Kenneth H. 1971. “Upward Social Mobility Sociology 72(2):193–207. and Political Orientation: A Re-evaluation of the Evi- Nie, Norman H., Jane Junn, and Kenneth Stehlik-Barry. dence.” American Sociological Review 36(2):223–35. 1996. Education and Democratic Citizenship in Walker, Lawrence J., Karl H. Hennig, and Tobias America. Chicago: University of Chicago Press. Krettenauer. 2000. “Parent and Peer Contexts for Niemi, Richard G., and M. Kent Jennings. 1991. “Issues Children’s Moral Reasoning Development.” Child and Inheritance in the Formation of Party Identi- Development 71(4):1033–48. fication.” American Journal of Politcal Science Watson, D., E. C. Klohnen, A. Casillas, E. N. Simms, J. 35(4):970–88. Haig, and D. S. Berry. 2004. “Match Makers and Deal Orum, Anthony M., and Roberta S. Cohen. 1973. “The Breakers: Analyses of Assortative Mating in Newly- Development of Political Orientations among Black wed Couples.” Journal of Personality 72(5):1029–68. and White Children.” American Sociological Review Weiner, Terry S., and Bruce K. Eckland. 1979. “Edu- 38(1):62–74. cation and Political Party: The Effects of College Patterson, G. R., Barbara D. DeBaryshe, and Eliza- or Social Class?” American Journal of Sociology beth Ramsey. 1989. “A Developmental Perspective 84(4):911–28. on Antisocial Behavior.” American Psychologist Westholm, Anders. 1999. “The Perceptual Pathway: Trac- 44(2):329–35. ing the Mechanisms of Political Value Transfer across Piaget, Jean. 1965. The Moral Judgment of the Child. Generations.” Political Psychology 20(3):525–51. New York: Free Press. Whitbeck, Les B., and Viktor Gecas. 1988. “Value Attri- Pirolli, Peter, and Stuart Card. 1999. “Information Forag- butions and Value Transmission between Parents ing.” Psychological Review 106(4):643–75. and Children.” Journal of Marriage and Family Rubinson, Richard. 1986. “Class Formation, Politics, and 50(3):829–40. Institutions: Schooling in the United States.” Ameri- Wilson, Glenn D., and J. R. Patterson. 1968. “A New can Journal of Sociology 92(3):519–48. Measure of Conservatism.” British Journal of Social Saphir, Melissa Nichols, and Steven H. Chaffee. 2002. and Clinical Psychology 7(4):264–9. “Adolescents’ Contributions to Family Communi- Zuckerman, Alan S. 2005. The Social Logic of Politics: cation Patterns.” Human Communication Research Personal Networks as Contexts for Political Behav- 28(1):86–108. ior. Philadelphia, PA: Temple University Press. Schönpflug, Ute. 2001. “Intergenerational Transmission of Values: The Role of Transmission Belts.” Journal Christopher Ojeda is a PhD candidate in the Depart- of Cross-Cultural Psychology 32(2):174–85. ment of Political Science at the Pennsylvania State Uni- Sears, David O., and Nicholas A. Valentino. 1997. “Poli- versity. His research is focused on disparities in political tics Matters: Political Events as Catalysts for Preadult participation that result from the combination of poverty, Socialization.” American Political Science Review resources, parenting, and child development. 91(1):45–65. Stoker, Laura, and Kent M. Jennings. 2006. “Politi- Peter K. Hatemi is Professor of Political Science at the cal Similarity and Influence between Husbands and Pennsylvania State University, and research fellow at the Wives.” Pp. 421–23 in The Social Logic of Politics, United States Studies Centre, University of Sydney, and edited by A. Zuckerman. Philadelphia, PA: Temple the Virginia Institute for Psychiatric and Behavioral University Press. Genetics. He is primarily interested in understanding the Taris, Toon W., and Gun R. Semin. 1997. “Passing on the sources and dynamics of human preferences and decision Faith: How Mother-Child Communication Influences making in complex and varying environments.