Skip to main content
NCBI home page
Public Opinion Quarterly logoLink to Public Opinion Quarterly
. 2015 Dec 27;80(1):66–89. doi: 10.1093/poq/nfv050

Adolescent Determinants of Abortion Attitudes

Evidence from the Children of the National Longitudinal Survey of Youth

Julianna Pacheco 1,*, Rebecca Kreitzer 1
PMCID: PMC4884817  PMID: 27257307

Abstract

The stability of abortion opinions suggests that pre-adult factors influence these attitudes more than contemporaneous political events. Surprisingly, however, we know little about the origins of abortion opinions, no doubt because the majority of research focuses on cross-sectional analyses of patterns across cohorts. We use a developmental model that links familial and contextual factors during adolescence to abortion attitudes years later when respondents are between 21 and 38 years old. Findings show that religious adherence and maternal gender role values are significant predictors of adult abortion opinions, even after controlling for contemporaneous religious adherence and the respondents’ own views on gender roles. Adolescent religious adherence matters more than religious denomination for adult abortion attitudes. The results have important implications for future trends in abortion attitudes in light of declining religiosity among Americans.

Few issues are more important to the American political landscape than abortion. The abortion issue has led to an increase in political activism (Maxwell 2002) and significant changes in partisan loyalties among the masses (Carsey and Layman 2006). Abortion attitudes influence vote choice in presidential (Abramowitz 1995), Senate (Cook, Jelen, and Wilcox 1994b), and gubernatorial elections (Cook, Jelen, and Wilcox 1994a) and remain an important issue to voters. In 2012, according to a Pew Research Center poll, 45 percent of Americans cited abortion as either “one among many important issues” or “a critical issue facing the country” (see also Jelen and Wilcox [2003]).

Abortion attitudes also exhibit stability across the lifecycle and, by some accounts, as much stability as partisanship (Converse and Markus 1979; Wilcox and Norrander 2002). The stability of abortion opinions provides evidence of persistence, meaning that these opinions are acquired early in life and are unlikely to change throughout the lifecycle (e.g., Alwin 1994). Persistence in abortion attitudes suggests that pre-adult factors are particularly influential, perhaps more so than contemporaneous political events. We know relatively little, however, about the developmental origins of abortion opinions, no doubt because the majority of research focuses on repeated cross-sectional analyses.

The purpose of this paper is to explore the determinants of abortion attitudes using a developmental model that links familial and contextual factors during adolescence to abortion attitudes approximately 20 years later when respondents are between 21 and 38 years old. We use the National Longitudinal Survey of Youth 1979 (NLSY79) and the Children of the National Longitudinal Survey of Youth (NLSYA) to explore how parental religious socialization, maternal attitudes toward women’s roles, and the state female sociopolitical culture influence support for legal abortion. We also explore whether the socializing effects are conditional on the politicization of the home.

We find evidence that pre-adult factors influence abortion attitudes. Respondents who attended religious services frequently during adolescence are less likely to support legalized abortion in adulthood, even after controlling for their current level of religious adherence. Adolescent religious attendance appears to be more important than adolescent religious denomination. Maternal views toward gender roles affect respondent attitudes toward abortion separate from the respondents’ own views. There is suggestive evidence that the effects of adolescent religious attendance and maternal gender role values are magnified for respondents who grew up in highly political homes. Finally, we find little empirical support that adolescent state female sociopolitical culture influences abortion attitudes; instead, culture matters contemporaneously.

Our results provide additional evidence that abortion attitudes exhibit persistence since pre-adult factors remain statistically significant predictors in the face of contemporaneous controls. Scholars are encouraged to use a longitudinal research design to gain a fuller understanding about the individual-level determinants of abortion attitudes.

Persistence in Public Opinion

According to the traditional model of persistence that emerged in the 1960s and 1970s, certain political attitudes are acquired at an early age, persist into adulthood, and have a major influence over adult behavior and opinion (e.g., Hyman 1959; Miller and Sears 1986; Alwin 1994). While the traditional model has been called into question (Sears 1989), scholars generally acknowledge that pre-adult socialization plays a role in the formation of predispositions like partisanship (Green, Palmquist, and Schickler 2002) and ideology (Alwin and Krosnick 1991) that allow individuals to make sense of the political world throughout adulthood (Zaller 1992; Achen 2002). Pre-adult factors matter less (if at all) for issue preferences that are unstable across the lifespan, such as presidential approval, consumer sentiment, operational ideology, and spending preferences.

The dominant research design used to study the persistence of abortion attitudes is cohort analyses from repeated cross-sectional data. For instance, Cook, Jelen, and Wilcox (1993) use the cumulative General Social Survey to explore cohort differences in support for legal abortion. They find that those who came of age before the 1960s are less supportive of abortion than those who reached adulthood later (see also Wilcox and Norrander [2002]), suggesting that shared experiences during young adulthood contribute to generational differences in attitudes. Although cohort analyses are useful for certain analytic purposes (e.g., identifying period or generational differences), they do little to identify the pre-adult forces that influence the development of abortion attitudes.

The purpose of this paper is to use a longitudinal research design to explore the determinants of abortion. Longitudinal data allow for a developmental framework since many determinants are measured prior to the outcome variable. As a result, retrospective bias is avoided (Niemi, Katz, and Newman 1980). In addition, the analytical strategy we employ allows us to test whether pre-adult factors predict adult attitudes better than contemporaneous factors, which makes longitudinal data superior to cross-sectional data (Miller and Sears 1986). And, while longitudinal analyses are used frequently to study political behavior (Plutzer 2002) and partisanship (Jennings and Markus 1984), to our knowledge, we are the first to use this type of research design to study abortion attitudes.

Socializing Agents and the Development of Abortion Attitudes

We draw largely from the literature on political socialization to develop hypotheses. Socialization is the process by which specific beliefs, values, or norms are transmitted across generations and accepted by the next generation. We focus on two agents of socialization, including the home environment and the state political culture, that likely influence the development of abortion attitudes.

HOME ENVIRONMENT

Studies of political socialization show that the home environment, specifically the parents, plays an important role in the development of attitudes (e.g., Lewis-Beck et al. 2008; Jennings, Stoker, and Bower 2009). The traditional view suggests that parents influence the attitudes of children through the transmission of beliefs, norms, or values. Transmission can be explicit or implicit; for instance, parents may explicitly indoctrinate their children toward their own political beliefs or implicitly model certain behaviors or attitudes that encourage imitation among children. Successful transmission occurs when children accept the values or beliefs transmitted by parents and continue to hold or be influenced by those beliefs in adulthood. We argue that parents influence the development of abortion attitudes through the transmission of religious orientations and gender role values.

Religious orientations, including religiosity, values, and beliefs, are likely to be passed on from parents to child (Argyle 2000; Jennings, Stoker, and Bowers 2009)—a process called parental religious socialization. Parents impact the religious orientations of offspring through behavior, for instance by bringing children to church or baptizing them as members of a congregation (Dudley and Dudley 1986) as well as through child-rearing practices, beliefs, and behaviors (Alwin and Felson 2010). More important for the development of abortion attitudes, adolescents growing up in highly religious households are likely taught the ways in which religious values, such as a devotion to family life (Wilcox 2002), are attached to salient political issues (Ammann 2014). Given that religious orientations are highly predictive of “pro-life” abortion attitudes (Cook, Jelen, and Wilcox 1992; Jelen and Wilcox 2003), parents serve an important role in highlighting how their children’s religious values are aligned with the political aspects of the abortion issue. As a result, our expectation is that respondents who grew up in religious households are less supportive of legalized abortion in adulthood (H1).

Parents are also influential in teaching children about women’s equality and attitudes about gender roles (Min, Silverstein, and Lendon 2012). One way that parents influence gender role attitudes is through the allocation of housework (Blair 1992; Cunningham 2001). Parents also directly influence offspring attitudes via transmission of their own attitudes toward gender roles (Cunningham 2001; Min, Silverstein, and Lendon 2012). Scholars argue that debates about unborn life are similar to debates about the nature of sexuality, work, and family commitments in women’s lives (Luker 1984). This provides compelling theoretical reasons to expect that attitudes about women’s roles influence abortion attitudes (see also Petchesky 1990[1984]). Others find that attitudes toward gender roles are associated with abortion attitudes among elites and activists, but not the mass public (Cook, Jelen, and Wilcox 1992; Jelen, Damore, and Lamatsch 2002). Nonetheless, we expect that respondents whose parents have traditional attitudes toward gender roles are less supportive of legalized abortion in adulthood (H2).

POLITICAL CULTURE

Contextual factors, such as political competition, also affect political development (Gimpel, Lay, and Schuknecht 2003; Campbell 2006). Differences in adolescent political environments are influential in the development of partisanship (Wolak 2009), political knowledge (Wolak and McDevitt 2011), and turnout in young adulthood (Pacheco 2008). While debates over the role of political culture persist (Almond and Verba 1963; Elazar 1974; Abramowitz 1980), scholars acknowledge the importance of gender in understanding the general political culture of the nation (e.g., Diamond 1977; Hill 1981). Empirical evidence suggests that a politically salient female sociopolitical subculture exists whereby states differ in the extent to which they treat women and men equally (Atkeson 2003). This subculture has important political consequences. Women are more likely to run for higher-level political office and win in states with a more progressive female sociopolitical culture (Atkeson and Carrillo 2007; Windett 2011). This is particularly relevant for abortion policies (Thomas 1991; Berkman and O’Connor 1993) and, subsequently, the development of abortion attitudes among youth.

The female sociopolitical subculture is important not only for the policy process, but also for the existence of women candidates as role models to adolescent residents. The presence of women as candidates and officeholders stimulates political engagement (Karp and Banducci 2008) and political discussion (Atkeson 2003) among women. In addition, Campbell and Wolbrecht (2006) find that adolescent girls are more likely to indicate an intention to be politically active when exposed to women politicians in highly visible campaigns. Given the importance of the female sociopolitical subculture on the restrictiveness of abortion policy and the existence of women candidates, we expect that youths growing up in a progressive female sociopolitical culture are more supportive of legalized abortion in adulthood (H3).

POLITICIZATION OF THE HOME AS AN ENHANCEMENT TO SOCIALIZATION

Prior research suggests that familial circumstances influence the extent to which socialization is successful. According to social learning theory, the transmission of attitudes by parents and acceptance by offspring are enhanced when parents provide consistent cues on a frequent basis (Bandura 1969). Transmission of political attitudes is higher in more politicized homes where political discussion is frequent and parents are engaged in politics (Jennings, Stoker, and Bowers 2009). We expect similar dynamics to apply to the development of abortion attitudes; parental factors have a higher impact on abortion attitudes for respondents who grew up in highly politicized homes (H4). Children living with politically engaged parents are more likely to map religious orientations and gender role values to the abortion issue, while children living in households with low levels of parental politicization should be relatively more open to influence from other socializing agents.

Data and Methods

We test our hypotheses using the National Longitudinal Survey of Youth 1979 (NLSY79) 1 coupled with the Children of the National Longitudinal Survey of Youth (NLSYA). The NLSY79 is a nationally representative sample of young men and women first interviewed in 1979 when they were age 14–22. These individuals were interviewed annually through 1994 and biennially since then on a variety of topics, including attitudes toward women’s roles in the household. In 1986, a separate survey, the NLSYA, of children born to NLSY79 female respondents began. Biennially (since 1994), children age 15 and older completed a lengthy interview modeled after the NLSY79 questionnaire. With the use of proper weights, the child sample is nationally representative of the children of women who were born from 1957 to 1964 and who were living in the United States in 1978 (NLS Handbook). 2

In 2006, 2008, and 2010, the American National Election Studies (ANES) was invited to include questions on both the NLSY79 and NLSYA surveys. The collaboration between the NLSY surveys and the ANES presents a unique opportunity to study the development of political behaviors and attitudes. We measure abortion attitudes from a 2008 survey question, which leaves us with an initial sample of 4,067 respondents with mothers who participated in the NLSY79. Additional information about the respective samples is included in appendix table A2, and descriptive information on all variables is included in appendix table A3.

A major challenge with the NLSYA data is that the sample is not a unified cohort. As a result, respondents vary in terms of when they came of age, which creates extra complexity for measuring adolescent variables that impact political development. On average, respondents turned 14 in the year 1996. However, some turned 14 as early as 1984, while others turned 14 in 2001. While we know that adolescence is an important time for political development, there are disagreements about at what age political development occurs. For instance, scholars using the Jennings data measure socialization at age 18 (Jennings, Stoker, and Bowers 2009), while others measure socializing factors as early as eighth grade (Sandell and Plutzer 2005). We measure socializing factors as early as possible, realizing that because surveys are conducted biannually and not every question was asked on each survey, we sometimes measure factors at different times across respondents. Nonetheless, we are confident that the variables are capturing important socializing factors prior to adulthood.

DEPENDENT VARIABLE

We measure abortion attitudes using the following question asked on the 2008 survey:

“Please choose the one that is closest to your opinion about abortion. 3) It should be legal for a woman to get an abortion under all circumstances, 2) it should be legal for a woman to get an abortion only under some specific circumstances, or 1) it should be illegal for a woman to get an abortion under all circumstances?”

Higher values indicate less support for legalized abortion. About 20 percent of respondents answered that an abortion should be illegal in all circumstances, 51 percent gave the middle answer, and 28 percent supported abortion in all circumstances. 3

PARENTAL RELIGIOUS SOCIALIZATION

Our expectation is that kids who grew up in religious households are less supportive of legalized abortion as adults (H1). There are three components to understanding the relevance of religion to political attitudes: belonging, beliefs, and behavior. Belonging refers to religious tradition and is based on religious group identity or denominations (Kellstedt et al. 1996), beliefs refer to biblical views or religious adherence, and behavior is based on religious attendance. All three are predictive of abortion attitudes among adults (Cook, Jelen, and Wilcox 1992; Hoffmann and Miller 1997), although behavior matters the most (Evans 2002).

Ideally, we would use parental measures on all three components to empirically test how parental religious socialization influences the development of abortion attitudes. Unfortunately, data limitations exist. First, neither parents nor children were asked any questions that capture the religious beliefs component. Second, mothers were asked about their religious behavior and belonging much more infrequently than children. More specifically, mothers are asked about their religious orientations in 1979, 1982, and 2000; the 1979 and 1982 surveys occur too early to be relevant for the political development of offspring, while the 2000 survey occurs too late. Third, because only mothers are interviewed, we have no measure for paternal religious orientations and are likely missing important information about the religious orientations of the household if we were to use only the maternal answers.

We opt to measure parental religious socialization using questions asked of offspring during their childhood. We use a question that asks respondents how often they attended religious services in the past year, to which they can answer not at all, several times a year, about once a month, 2–3 times a month, about once a week, or more than once a week. To measure belonging, we use a question that asks respondents to select their present religion. We recoded this variable into four binary variables that include Protestant, Catholic, other (including Jewish), and none, with the “none” category being omitted. Both questions are asked in every survey; our measures capture the respondents’ answers in the earliest survey possible. For instance, at the onset, we use responses in 1988. If respondents are missing on that question, presumably because they were not old enough to answer it, we use the 1990 responses. If missing for both 1988 and 1990, we use the 1992 survey, and so on.

We note that maternal religious orientations are highly related to offspring orientations, suggesting that the offspring measures are picking up on parental religious socialization. Seventy-three percent of respondents who were Protestant in adolescence had mothers who also reported being Protestant, while 88 percent of respondents who were Catholic in adolescence had mothers who also reported being Catholic. The relevant numbers for the other and none categories are 23 and 26 percent, respectively. Similarly, the correlation between maternal and offspring religious behavior is .25. Finally, inferences regarding the role of adolescent religious belonging and behavior are nearly identical when maternal variables are substituted for the offspring measures.

GENDER ROLE VALUES

We use maternal responses to measure gender role values of the home, with the expectation that respondents who grew up with mothers with traditional gender role values are less supportive of legalized abortion in adulthood (H2). We combined answers on a series of six questions that measured different aspects of the roles women should take. The scale comprises a series of Likert scale responses to statements such as “A woman’s place is in the home, not the office or shop.” The full list of questions is in appendix table A4. Mothers are asked these series of questions in 1979, 1982, 1987, and 2004. A principal factor analysis of the questions indicated one significant dimension with an average reliability, as measured by Cronbach’s alpha of .80. We ultimately use the 1987 measure since this corresponds to the time frame when the majority of offspring were adolescents. 4 We control for the behavioral counterpart since maternal employment may increase support for legal abortion and other methods for regulating fertility, irrespective of gender role values. We create a measure of average maternal employment from a three-point measure of employment (unemployed, partially employed, fully employed) asked over many years. 5 This average measure is better than using the response in a single year, as women may enter and exit the workforce frequently during childbearing years. Average maternal employment is negatively correlated with traditional gender role values (r = –.15).

STATE FEMALE SOCIOPOLITICAL SUBCULTURES

We also expect the female sociopolitical subculture to influence the development of abortion attitudes such that youths growing up in a progressive female sociopolitical culture are more supportive of legalized abortion in adulthood (H3). We use Windett’s (2011) measure of female sociopolitical subculture matched to the state of residence reported by respondents. This measure takes into account ten state-level factors, including four separate indicators of the level of political representation of women at the state and Congressional levels, two variables that tap into broad political culture using Elazar’s (1974) typology, and four variables that measure the level of equality, including the educational attainment and workforce inclusion of women and the state’s participation in the ratification of the ERA. Higher values on this measure indicate a more progressive female sociopolitical subculture. 6

We matched respondents based on the state of residence reported in the earliest surveys possible. For instance, for the oldest respondents, we matched these variables based on the reported state of residence in 1994. 7 If respondents are missing on the state indicator variable in 1994, we use the reported state of residence in 1996. If respondents are missing on both state indicators in 1994 and 1996, we use the reported state of residence in 1998, and so on. 8

CONDITIONAL EFFECTS OF THE POLITICIZATION OF THE HOME

To test whether socialization is more successful for youths who grew up in a highly politicized home (H4), we construct an index of the politicization of the home that combines information from offspring about the frequency of political discussion within their families and the level of political interest of their mothers. Specifically, on the 2008 survey, respondents are asked, “When you were growing up, how often did you hear the adults in your household talking about politics?” to which answers range from extremely often (5) to never (1). Respondents are also asked how often their mother followed what was going on in politics, to which they could answer always (5) to never (1). 9 These two responses were combined to create a standardized variable, with higher values indicating a more politicized home; this scale has a high degree of reliability, as indicated by a Cronbach’s alpha of .66. To ease statistical interpretation, we create three groups roughly equal in size, corresponding to low, middle, and high home politicization. The most political group (33 percent) had scores that were less than or equal to –.30, the middle group (36 percent) had scores between –.30 and .27, and the least political group (31 percent) had scores greater than .27.

CONTROL VARIABLES

Our analyses control for a number of demographic variables that influence adult attitudes toward abortion including age, gender (1 = female), race (black, Hispanic, white is omitted category), employment status (1 = working 30 or more hours a week, 0 otherwise), whether the respondent is a parent, and partisan identification (Independent, Republican, Democrat is omitted category). Contemporaneous employment is important to include, particularly for women, since attitudes may change to reflect an individual’s economic situation, irrespective of gender role values (Plutzer 1988); for instance, women who work out of economic necessity may gradually adopt feminist attitudes that coincide with their current situation. 10 We also include educational attainment, measured as the highest grade of school that the respondent completed; this varies from first grade to eighth year of college or more. Finally, we test for whether pre-adult factors predict adult attitudes better than contemporaneous factors by including measures of religious belonging and behavior, respondent views toward traditional gender roles, and a measure of the female sociopolitical subculture based on their current state of residence. All of these variables are coded identically to the adolescent variables, but come from the 2008 survey. 11

In the results reported below, models are estimated using ordered logistic regression since the dependent variable is ordinal. 12 All estimates are weighted to account for the complex survey design, and standard errors are clustered by mother to account for the non-independence among siblings. To ease statistical interpretation, all variables are rescaled to range from 0 to 1. In the first model, we predict adult abortion attitudes using only control variables and the socializing factors that come from the home or the state political context. In the second model, we include the contemporaneous measures of religious behavior and belonging as well as gender role values; the third model includes contemporaneous measures of the female sociopolitical subculture. Comparisons between the first model and these two additional models provide information about whether respondents are influenced by pre-adult or contemporaneous factors. Finally, in the fourth model, we present a combined model with all of the covariates included. Results are shown in table 1.

Table 1.

Ordered Logistic Regression Predicting Support for Banning Abortion

Model 1 Model 2 Model 3 Model 4
(n = 2,076) (n = 1,760) (n = 1,841) (n = 1,754)
b (SE) b (SE) b (SE) b (SE)
Catholic, adolescent –0.37 (0.21) # –0.35 (0.24) –0.32 (0.56) # –0.36 (0.24)
Protestant, adolescent –0.02 (0.17) –0.09 (0.20) –0.34 (0.50) –0.07 (0.20)
Other religion, adolescent –0.37 (0.21) # –0.42 (0.23) # –0.63 (0.53) –0.41 (0.23) #
Religious attendance, adolescent 0.75 (0.16) ** 0.68 (0.18) ** 1.20 (0.34) ** 0.66 (0.18) **
Maternal average employment –0.22 (0.11) * –0.19 (0.12) –0.22 (0.19) * –0.19 (0.12) #
Maternal gender role values 1.07 (0.35) ** 0.80 (0.38) * 1.51 (0.64) * 0.79 (0.38) *
State female socio-political culture, adolescent –1.18 (0.30) ** –0.83 (0.32) ** 0.14 (0.47) 0.02 (0.47)
Catholic, adult 0.40 (0.21) # 0.38 (0.21) #
Protestant, adult 0.49 (0.18) ** 0.49 (0.18) **
Other religion, adult 0.34 (0.19) # 0.35 (0.19) #
Religious attendance, adult 1.28 (0.23) ** 1.25 (0.23) **
Respondent gender role values 1.03 (0.37) ** 1.02 (0.37) **
Employed 30 hours or more 0.21 (0.14) 0.30 (0.25) 0.21 (0.15)
State female socio-political culture, adult –1.78 (0.82) ** –0.90 (0.42) *
Female –0.24 (0.12) * –0.25 (0.13) # 0.10 (0.21) –0.26 (0.13) #
Age –0.02 (0.02) –0.03 (0.02) –0.05 (0.04) –0.02 (0.02)
Black 0.13 (0.12) 0.05 (0.14) 0.29 (0.23) 0.04 (0.14)
Hispanic 0.20 (0.15) 0.16 (0.15) 0.09 (0.29) 0.19 (0.15)
Education –1.18 (0.43) ** –1.71 (0.50) ** –0.29 (0.80) –1.67 (0.50) **
Respondent has child 0.16 (0.12) 0.17 (0.13) 0.02 (0.25) 0.16 (0.13)
Independent 0.11 (0.12) 0.23 (0.14) # 0.46 (0.24) # 0.21 (0.14)
Republican 0.72 (0.16) ** 0.64 (0.17) ** 0.96 (0.31) ** 0.63 (0.17) **
Political home 0.06 (0.31) –0.10 (0.36) –0.89 (0.91) –0.15 (0.35)
Cut 1, constant –2.30 (0.64) ** –1.66 (0.72) ** –2.57 (1.43) # –1.62 (0.72) *
Cut2, constant 0.14 (0.63) 0.95 (0.72) –0.20 (1.40) 1.00 (0.72)
Open in a new tab

Note.—Robust standard errors shown in parentheses. Estimates are weighted to account for the complex survey design.

#p < .10; *p < .05; **p < .01 with two-tailed significance tests.

Results

As shown in model 1, nearly all of the pre-adult factors influence adult abortion attitudes and our expectations are largely confirmed. Recall that we expect adolescents who grew up in religious households to be less supportive of legalized abortion than those who grew up in less religious households (H1). We included two measures for parental religious socialization in our model, based on the concepts of religious belonging (denomination) and religious behavior (frequency of attendance at religious services). Interestingly, it appears that the relationship between parental religious socialization and abortion attitudes in adulthood is more about behavior than belonging. Respondents who attended religious services frequently in adolescence are more likely to oppose legalized abortion compared to those who never attended religious services in adolescence. While two of the three religious denomination indicator variables are significant in our models, they are only marginally significant and for the Catholic category, in the opposite direction than expected. 13 Besides parental religious socialization, we also expected gender role attitudes to matter (H2). Our results suggest that respondents whose mothers had traditional gender role values are more opposed to legal abortion than those whose mothers had nontraditional gender role values.

Few of our inferences change when contemporaneous measures are added (model 2). Adolescent religious attendance and maternal traditional gender role values continue to influence abortion attitudes. The coefficients on the variables measuring contemporaneous Catholic and Protestant identity are significant and in the expected directions. Adult religious attendance and the respondents’ own views about traditional gender roles also matter.

We also expected the state political context to affect the development of abortion attitudes (H3). While the female sociopolitical culture measure in adolescence is statistically significant in model 1, when we add the adult cultural variable to the model (model 3), adolescent state culture no long matters. This is not particularly surprising, given the high correlation between adolescent and adult state cultures (r =.82). Yet, adolescent female sociopolitical culture remains insignificant even in models estimated on respondents who reported different states of residence in adolescence and adulthood and where the correlation between the two measures is much lower (r = .29).

Inferences are nearly identical using the full model specification, as shown in model 4. 14 Figures 1 and 2 present predicted probabilities that respondents support legalized abortion across levels of adolescent religious adherence and maternal general role values from model 4. The greatest differences are found among respondents who believe that abortion should always be legal or always be illegal. As shown in figure 1, the model predicts that adolescents who rarely attend religious services are 16 percentage points more likely to support legal abortion in all circumstances compared to those who attended religious services frequently. Figure 2 presents similar effects of maternal gender role values. There is an approximately 15-percentage-point difference in supporting legal abortion in all circumstances when mothers held very liberal or very conservative beliefs about gender roles. These substantive effects exceed those of partisanship and race 15 and are only slightly smaller than their contemporary counterparts. For instance, respondents who frequently attend religious services or who, in adulthood, hold the most traditional gender role values are 22.2 and 17.9 percentage points more likely to answer that abortion should be illegal in all circumstances compared to those who rarely attend religious services or who hold the least traditional gender role values, respectively.

Figure 1.

Predicted Probability of Supporting Legal Abortion, by Adolescent Religiosity.

Predictions are estimated using model 4 in table 1 and keeping all other variables constant.

Figure 1.

Open in a new tab

Figure 2.

Predicted Probability of Supporting Legal Abortion, by Maternal Gender Role Values.

Predictions are estimated using model 4 in table 1 and keeping all other variables constant.

Figure 2.

Open in a new tab

ARE EFFECTS UNIQUE TO ABORTION ATTITUDES?

Prior research suggests that pre-adult factors should matter most for attitudes that exhibit persistence, such as abortion attitudes, party identification, ideology, and racial attitudes, while pre-adult factors should be largely inconsequential for issue preferences that are unstable across the lifespan. While limited, the NLSYA does permit us to test these expectations by looking at the determinants of other outcomes, including party identification, attitudes toward success and hard work, attitudes toward the federal government providing financial help for unemployed citizens, attitudes toward the federal government making sure everyone has an equal opportunity to succeed, and political interest.

As shown in the supplementary data online, religious attendance during adolescence and maternal gender roles have no statistically significant relationship with any of the dependent variables aside from abortion. That is not to say that religion or gender role values do not matter; instead, these factors matter contemporaneously. The results suggest that current religiosity is important for party identification and respondents with more traditional gender role values are more politically interested. Religion and gender role values are largely inconsequential to other political preferences.

Conditional Effects Depending on Politicization of the Home

To test whether the effects of pre-adult factors are conditional on the amount of politicization in the homes (H4), we re-estimate model 4 with a split sample based on the composite index described above. 16

As shown in table 2, the evidence suggests that parental socializing factors matter more for respondents who grew up in households with lively political discussion and high political interest, which is consistent with previous research (Jennings, Stoker, and Bowers 2009). Among the respondents in the low political home group, maternal gender role values do not influence adult abortion attitudes and only the effect of adolescent religious service attendance is significant. Both parental socializing factors are statistically significant for respondents from highly politicized homes. Overall, then, there is moderate evidence that families marked by frequent political discussion and politically engaged parents are families fostering the transmission of values that have an impact on the development of abortion attitudes that then persist into adulthood.

Table 2.

Ordered Logistic Regression Predicting Support for Banning Abortion across Level of Home Politicization

Low-political home High-political home
(n = 602) (n = 512)
b (SE) b (SE)
Catholic, adolescent –0.48 (0.38) –0.91 (0.60)
Protestant, adolescent –0.14 (0.34) –0.46 (0.56)
Other religion, adolescent –0.69 (0.41) # –0.18 (0.58)
Religious attendance, adolescent 0.69 (0.31) * 0.97 (0.37) **
Maternal average employment –0.12 (0.20) –0.16 (0.21)
Maternal gender role values 0.49 (0.63) 1.14 (0.68) #
State female socio-political culture, adolescent 1.33 (0.86) 1.00 (0.95)
Catholic, adult 0.55 (0.32) # 0.45 (0.38)
Protestant, adult 0.82 (0.32) * 0.27 (0.36)
Other religion, adult 0.72 (0.31) * 0.03 (0.45)
Religious attendance, adult 1.06 (0.42) * 1.05 (0.43) *
Traditional gender role values, adult 0.85 (0.63) 1.32 (0.68) #
R employed 30 hours or more 0.51 (0.24) * 0.22 (0.23)
State female socio-political culture, adult –1.05 (0.74) –2.04 (0.85) *
Female –0.45 (0.25) # 0.16 (0.22)
Age –0.01 (0.04) –0.03 (0.04)
Black 0.31 (0.24) 0.32 (0.25)
Hispanic 0.16 (0.26) 0.18 (0.30)
Education –2.93 (0.79) ** –0.64 (0.82)
R has child 0.33 (0.24) 0.12 (0.26)
Independent –0.02 (0.22) 0.38 (0.26)
Republican 0.39 (0.29) 0.81 (0.31) **
Cut 1, constant –1.30 (1.25) –1.02 (1.45)
Cut 2, constant 1.35 (1.26) 1.41 (1.45)
Open in a new tab

Note.—Robust standard errors shown in parentheses. Estimates are weighted to account for the complex survey design. #p < 0.10; *p < 0.05; **p < 0.01 with two-tailed significance tests.

Conclusion

Forty years after the legalization of abortion in Roe v. Wade (1973), abortion remains an important political issue that is highly salient in elections at all levels of government and a quintessential litmus test for conservatism. Because of the stability of abortion attitudes over time, it is important for scholars to develop a better understanding of the early formation of these attitudes. We contribute to that understanding by using a longitudinal research design. Our analyses have uncovered some interesting patterns. We find that childhood religious attendance and maternal gender role values are significant predictors of adult abortion opinions, even after controlling for contemporaneous religious attendance and the respondents’ own views on gender roles. This suggests that abortion attitudes are transmitted across generations via religious practices and maternal values about gender equality.

LIMITATIONS AND SUPPLEMENTAL ANALYSES

Even though our results suggest that abortion attitudes are influenced by pre-adult factors, we emphasize the shortcomings of the NLSYA data. These include (1) the dearth of adolescent variables; (2) the three-point abortion question, which has less variation compared to a four-point abortion question; (3) the inability to differentiate among sects of religious denominations (e.g., mainline Protestants, Evangelicals); and (4) the absence of a variable that captures religious beliefs. We attempt to mitigate some of these shortcomings by replicating the analyses using the Jennings-Niemi (JN) Student Parent Socialization Study. As we note in the supplementary data online, inferences are similar across the two data sets, even given the differences in sampling, question wording, and so forth. In addition, preliminary evidence, shown in model 2 in table S4, suggests that religious beliefs also influence the development of abortion attitudes and, perhaps, even more so than religiosity or denomination. The supplemental analyses are generally supportive of our conclusions, giving us increased confidence in the evidence reported using the NLSYA data.

FUTURE RESEARCH

Our results suggest that gradual trends in religious commitment or gender role values impact abortion attitudes. Religious affiliation is at its lowest point since it started to be tracked in the United States. In 2012, for instance, 20 percent of respondents in the General Social Survey had no religious preference (despite only 8 percent being raised without religion), more than double the number reported in 1990 (Hout, Fischer, and Chaves 2013). The decline of religious commitment could have unforeseen consequences for abortion attitudes and abortion policy.

We also find that the effect of religious adherence and maternal gender role values are enhanced in highly politicized households where parents engage in political discussions with their children and are highly interested in politics. Increasing the political engagement of the electorate not only affects short-term factors, such as electoral results or political responsiveness, but also has lasting impacts on successive cohorts through the development of attitudes. We also note that our results provide additional evidence that the politically engaged are likely to link woman’s rights to the abortion issue (Cook, Jelen, and Wilcox 1992).

Finally, there is limited evidence that the state political context plays a role in socializing abortion attitudes. While others find that political contexts may impact the development of politically active citizens, we find that parental variables matter much more for the development of abortion attitudes. Even among adolescents with low levels of politicization within the home, we find little socializing effects of the state political culture, even though it is in those instances where agents other than the home may be particularly influential.

Additional research using longitudinal research designs would help clarify the relationships we find in this paper. We encourage scholars to explore the determinants of other abortion questions, such as those that ask about legalized abortion in specific circumstances, such as when the mother’s health is in danger versus when the pregnancy is a result of rape. Exploring the pre-adult determinants of other morality issues, such as gay marriage or the death penalty, that are value laden, connected to religious beliefs, and stable across the lifespan would help further illuminate the role that socialization plays on the development of politically salient attitudes.

Supplementary Data

Supplementary data are freely available online at http://poq.oxfordjournals.org/.

Supplementary Data
supp_80_1_66__index.html (689B, html)

Appendix

Table A1.

Response Rates for the Total Sample: NLSY 79

Year # of Responses Total deceased Rate
1979 12,686 0 –
1980 12,141 9 95.8
1981 12,195 29 96.3
1982 12,123 44 95.9
1983 12,221 57 96.8
1984 12,069 67 95.6
1985 10,941 79 94.5
1986 10,655 95 92.6
1987 10,485 110 91.2
1988 10,465 127 91.2
1989 10,605 141 92.5
1990 10,436 152 91.1
1991 9,182 144 91.9
1992 9,015 156 91.9
1993 9,011 177 92.1
1994 8,891 204 91.1
1996 8,636 243 88.8
1998 8,399 275 86.7
2000 8,033 313 83.2
2002 7,724 346 80.3
2004 7,661 399 80.1
2006 7,654 456 80.5
2008 7,757 503 82.0
2010 7,565 573 80.6
Open in a new tab

Note.—Response rate is defined as the percentage of base-year respondents remaining eligible and not known to be deceased who were interviewed in a given survey year. Obtained from the NLSY website: https://www.nlsinfo.org/content/cohorts/nlsy79/intro-to-the-sample/retention-reasons-noninterview. Interviews were completed with 89.7, 88.7, and 71.5 percent of those selected for the base year interview from the cross-section sample of youth, the oversampled group of economically disadvantaged sample, and the military sample, respectively. More information is available in the Technical Sampling Report (https://www.nlsinfo.org/sites/nlsinfo.org/files/attachments/130212/NLSY79%20Tech%20Samp%20Rpt.pdf). It should be noted that children of the poor and military samples were never considered eligible to be in the Young Adult sample.

Table A2.

NLSY79 Mother & Child Samples: 1979–2010

1979 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Female Respondents
    Interviewed 6283 5418 5312 4510 4535 4480 4361 4299 4113 3955 3984 3916 3975 3896
Mothers
    # Mothers Interviewed – 2922 3346 3088 3325 3464 3489 3533 3425 3315 3365 3311 3356 3280
Children
    Born a – 5255 6543 6427 7255 7862 8125 8395 8323 8100 8267 8131 8266 8099
    Interviewed b – 4971 6266 5803 6509 7089 7103 7067c 6417 7467d 7538e 7816f 7660g 6997h
 Under age 15 6109 5431 4924 3392 3229 2514 1972 1354 895
 Age 15 and over 980 1672 2143 3025 4238 5024 5844 6306 6102
Open in a new tab

Note.—Young Adult eligibility is discussed in the user guide section on sample design (https://www.nlsinfo.org/content/cohorts/nlsy79-children/intro-to-the-sample/sample-design). Children of the poor white and military samples were never considered eligible to be in the Young Adult sample. Young Adult aged children who have either never been assessed or whose mothers fall into the “do not refield” category are not interviewed. Children who are deceased or adopted out by the point in which they would end up into the Young Adult sample are never eligible. Sample sizes for all child survey years exclude the 441 female members of the military subsample dropped from interviewing in 1985 and the children born to these women. In addition, sample sizes for 1990 and later surveys exclude female members of the civilian economically disadvantaged, non-black/non-Hispanic subsample, whose children were not eligible for assessment or for interview as young adults. (Women in this oversample were interviewed in 1990, but their children were not included in that year due to budget constraints and in anticipation of the dropping of the sample the next round.) The exclusion of this sample after 1988 accounts for much of the drop in sample size between 1988 and 1990. Young adults age 21 and older were not fielded in 1998 but were returned to the eligible sample in 2000. In 2000, 38 percent of the black and Hispanic child and young adult oversamples were not fielded but were restored to the sample in 2002. Beginning in 2010, young adults over age 30 are only interviewed every other round.

aChildren born to interviewed mothers; this number includes deceased and non-resident children.

bA child interview was considered complete if an interviewer was able to directly assess a child, or to obtain mother-report assessment information on the child’s background and health. Child age is determined as of December 31 of the survey year.

cThis total includes 37 children (age 0–4) who were assessed or interviewed whose mothers were not interviewed.

dThis total includes 14 children (age 0–14) who were assessed or interviewed whose mothers were not interviewed and 257 young adult children whose mothers were not interviewed.

eThis total includes 13 children (age 0–14) who were assessed or interviewed whose mothers were not interviewed and 306 young adult children whose mothers were not interviewed.

fThis total includes 30 children (age 4–14) who were assessed or interviewed whose mothers were not interviewed and 452 young adult children whose mothers were not interviewed.

gThis total includes 7 children (age 4–14) who were assessed or interviewed whose mothers were not interviewed and 406 young adult children whose mothers were not interviewed.

hThis total includes 7 children (age 4–14) who were assessed or interviewed whose mothers were not interviewed and 283 young adult children whose mothers were not interviewed.

Table A3.

Mean, Standard Deviation, Minimum, and Maximum for All Variables

Obs Mean St Dev Min Max
Dependent variable
Abortion illegality 1254 1.93 0.70 1 3
Adolescent socialization variables
Catholic 3886 0.17 0.37 0 1
Protestant 3886 0.46 0.50 0 1
Other religion 3886 0.23 0.42 0 1
Religious attendance 1125 0.52 0.36 0 1
Maternal average employment 1017 2.13 0.54 1 3
Maternal traditional gender roles 1242 0.31 0.18 0 1
Female socio-political culture 1927 0.44 0.21 0 1
Adult religious orientation, gender role Attitudes, and state context
Catholic 8691 0.17 0.38 0 1
Protestant 8691 0.36 0.48 0 1
Other religion 8691 0.24 0.43 0 1
Religious attendance 1183 0.32 0.32 0 1
Traditional gender roles 1253 0.31 0.16 0 1
Employed 30 hours or more 2372 0.75 0.44 0 1
Female socio-political culture 3859 0.50 0.23 0 1
Adult contemporaneous variables
Female 1254 0.52 0.50 0 1
Age 1254 25.71 3.41 21 38
Black 1254 0.18 0.39 0 1
Hispanic 1254 0.08 0.28 0 1
Education 1254 0.56 0.13 0 1
Has child 2292 0.43 0.49 0 1
Independent 8691 0.39 0.48 0 1
Republican 8691 0.24 0.43 0 1
Political home 1235 0.31 0.20 0 1
Open in a new tab

Table A4.

Gender Role Statements Used to Create Composite Scale

We are interested in your opinion about the employment of wives. I will read a series of statements and after each one I would like to know whether you strongly agree, agree, disagree, or strongly disagree.
1. A woman’s place is in the home, not the office or shop.
2. A wife who carries out her full family responsibilities doesn’t have time for outside employment.
3. Employment of wives leads to more juvenile delinquency.
4. It is much better for everyone concerned if the man is the achiever outside the home and the woman takes care of the home and family.
5. Men should share the work around the house with women, such as doing the dishes, cleaning, and so forth.
6. Women are much happier if they stay at home and take care of the children.
Open in a new tab

Footnotes

1

Response rates are provided in appendix table A1.

2

The design of the NLSY79 survey includes three independent probability samples: (1) a cross-section sample designed to represent the non-institutional civilian segment of American young people age 14–21; (2) a supplemental sample designed to oversample minorities and economically disadvantaged youth; and (3) a military sample designed to represent the population age 17–21 who are serving in the military. All three samples were selected by standard area probability sampling methods. See the Technical Sample Report for more information (https://www.nlsinfo.org/sites/nlsinfo.org/files/attachments/130212/NLSY79%20Tech%20Samp%20Rpt.pdf).

3

Unfortunately, mothers are not asked a similar question.

4

If mothers are missing on the 1987 measure (N = 109), we use the 1982 measure if missing in 1987 (N = 109) and the 2004 measure (N = 37) if both the 1987 and 1982 measure are missing.

5

This question was asked in 1979, 1980, 1981, 1982, 1984, 1986, 1987, 1988, 1990, 1991, 1992, 1993, 1996, and 1998.

6

Detailed information on reliability is reported in Windett (2011).

7

This is the first year available with state of residence reported from youths.

8

We obtained permission from the Bureau of Labor Statistics to access the geocodes of the respondents. More details can be found on their website (http://www.bls.gov/nls/geocodeapp.htm). We used the state of residence in 1994 for 20 percent of respondents, 1996 for 16 percent of respondents, 1998 for 18 percent of respondents, 2000 for 18 percent of respondents, and 2002 for 27 percent of respondents.

9

If missing, we used responses from 2006.

10

The youths in the survey were not asked about their employment status, so we use a measure of full employment (defined as working 30+ hours) instead.

11

In the event that a respondent was missing on either of these variables in the 2008 survey, we used the 2006 survey.

12

The results are similar when estimated with a multinomial logit. A Brant test indicates that the parallel lines assumption is not violated.

13

An interaction variable between denomination and attendance failed to reach statistical significance.

14

Inferences are nearly identical when the religious denomination variables are omitted from the model. The coefficient on the religious attendance, adolescent variable is .54***.

15

The probability of saying that abortion should be illegal in all circumstances is 9.4 percentage points higher for Republicans relative to Democrats. Race is not associated with abortion attitudes.

16

A Chow test allows us to reject the null hypothesis that the difference between the coefficients in the low home politics model and the high home politics model is equal to zero, which provides some empirical support for estimating separate models. The interaction between the home politics indicator variable and adolescent religiosity is significant, indicating that adolescent religiosity is a stronger determinant of abortion attitudes among those from politicized homes (β = .59* with a one-tailed test). The interaction between home politics and maternal gender role values fails to reach statistical significance, albeit slightly so (β = .86; p-value is .12 with a one-tailed test).

References

  1. Abramowitz Alan. 1980. “A Comparison of Voting for US Senator and Representative.” American Political Science Review 74:633–40. [Google Scholar]
  2. ———. 1995. “It’s Abortion, Stupid: Policy Voting in the 1992 Presidential Election.” Journal of Politics 57:176–86. [Google Scholar]
  3. Achen Christopher H. 2002. “Parental Socialization and Rational Party Identification.” Political Behavior 24:151–69. [Google Scholar]
  4. Almond Gabriel A., Verba Sidney. 1963. The Civic Culture: Political Attitudes and Democracy in Five Nations. Sage. [Google Scholar]
  5. Alwin Duane F. 1994. “Aging, Personality, and Social Change: The Stability of Individual Differences over the Adult Life-Span.” In Life-Span Development and Behavior, edited by Featherman David L., Lerner Richard M., Perlmuter Marion, 135–85. Hillsdale, NJ: Lawrence Erlbaum. [Google Scholar]
  6. Alwin Duane F., Felson Jacob L. 2010. “Religion and Child Rearing.” In Religion, Families, and Health: Population Based Research in the United States, edited by Ellison Christopher G. and Hummer Robert A., 40–60. Piscataway, NJ: Rutgers University Press. [Google Scholar]
  7. Alwin Duane F., Krosnick Jon. 1991. “Aging, Cohorts, and the Stability of Sociopolitical Orientations over the Life Span.” American Journal of Sociology 97:169–96. [Google Scholar]
  8. Ammann Sky L. 2014. “Creating Partisan ‘Footprints’: The Influence of Parental Religious Socialization on Party Identification.” Social Science Quarterly 95:1360–1380. [Google Scholar]
  9. Argyle Michael. 2000. Psychology and Religion: An Introduction. New York: Routledge. [Google Scholar]
  10. Atkeson Lonna Rae. 2003. “Not All Cues are Created Equal: The Conditional Impact of Female Candidates on Political Engagement.” Journal of Politics 65:1040–1061. [Google Scholar]
  11. Atkeson Lonna R., Carrillo N. 2007. “More Is Better: The Impact of Female Representation on Citizen Attitudes Toward Government Responsiveness.” Gender and Politics 3:79–101. [Google Scholar]
  12. Bandura Albert. 1969. “Social-Learning Theory of Identificatory Processes.” In Handbook of Socialization Theory Research, edited by Goslin David A., 213–62. Chicago: Rand McNally. [Google Scholar]
  13. Berkman Michael B., O’Connor Robert E. 1993. “Do Women Legislators Matter? Female Legislators and State Abortion Policy.” American Politics Research 21:102–24. [DOI] [PubMed] [Google Scholar]
  14. Blair Sampson L. 1992. “The Sex-Typing of Children’s Household Labor: Parental Influence on Daughters’ and Sons’ Housework.” Youth and Society 24:178–203. [Google Scholar]
  15. Campbell David E. 2006. Why We Vote: How Schools and Communities Shape Our Civic Life. Princeton, NJ: Princeton University Press. [Google Scholar]
  16. Campbell David E., Wolbrecht Christina. 2006. “See Jane Run: Women Politicians as Role Models for Adolescents.” Journal of Politics 68:233–47. [Google Scholar]
  17. Carsey Thomas M., Layman Geoffrey C. 2006. “Changing Sides or Changing Minds? Party Conversion, Issue Conversion, and Partisan Change on the Abortion Issue.” American Journal of Political Science 50:464–77. [Google Scholar]
  18. Converse Philip E., Markus Gregory. 1979. “Plus ça Change…The New CPS Panel Study.” American Political Science Review 73:32–49. [Google Scholar]
  19. Cook Elizabeth Adell, Jelen Ted G., Wilcox Clyde. 1992. Between Two Absolutes: Public Opinion and the Politics of Abortion. Boulder, CO: Westview. [Google Scholar]
  20. ———. 1993. “Generational Differences in Attitudes toward Abortion.” American Politics Quarterly 21:31–53. [DOI] [PubMed] [Google Scholar]
  21. ———. 1994a. “Issue Voting in US Senate Elections: The Abortion Issue in 1990.” Congress and the Presidency 21:99–112. [Google Scholar]
  22. ———. 1994b. “Issue Voting in Gubernatorial Elections: Abortion and Post-Webster Politics.” Journal of Politics 56:187–99. [Google Scholar]
  23. Cunningham Mick. 2001. “The Influence of Parental Attitudes and Behaviors on Children’s Attitudes Toward Gender and Household Labor in Early Adulthood.” Journal of Marriage and Family 63:111–22. [Google Scholar]
  24. Diamond Irene. 1977. Sex Roles in the State House. New Haven, CT: Yale University Press. [Google Scholar]
  25. Dudley Roger L., Dudley Margaret G. 1986. “Transmission of Religious Values from Parents to Adolescents.” Review of Religious Research 28:3–15. [Google Scholar]
  26. Elazar Daniel. 1974. American Federalism: A View from the States. New York: Harper. [Google Scholar]
  27. Evans John H. 2002. “Polarization in Abortion Attitudes in US Religious Traditions, 1972–1998.” Sociological Forum 17:397–422. [Google Scholar]
  28. Gimpel James, Lay J. C., Schuknecht J. E. 2003. Cultivating Democracy: Civic Environments and Political Socialization in America. Washington, DC: Brookings Institution. [Google Scholar]
  29. Green Donald P., Palmquist Bradley, Schickler Eric. 2002. Partisan Hearts and Minds: Political Parties and the Social Identities of Voters. New Haven, CT: Yale University Press. [Google Scholar]
  30. Hill David. 1981. “Political Culture and Female Political Representation.” Journal of Politics 43:159–68. [Google Scholar]
  31. Hoffmann John P., Miller Alan S. 1997. “Social and Political Attitudes Among Religious Groups: Convergence and Divergence over Time.” Journal for the Scientific Study of Religion 36(1):52–70. [Google Scholar]
  32. Hout Michael, Fischer Claude S., Chaves Mark A. 2013. “More American Have No Religious Preference: Key Findings from the 2012 General Social Survey.” Institute for the Study of Societal Issues. Available at http://sociology.berkeley.edu/sites/default/files/faculty/fischer/Hout%20et%20al_No%20Relig%20Pref%202012_Release%20Mar%202013.pdf.
  33. Hyman Herbert H. 1959. Political Socialization. Glencoe, IL: Free Press. [Google Scholar]
  34. Jelen Ted G., Damore David F., Lamatsch Thomas. 2002. “Gender, Employment Status, and Abortion: A Longitudinal Analysis.” Sex Roles 47:321–30. [Google Scholar]
  35. Jelen Ted G., Wilcox Clyde. 2003. “Causes and Consequences of Public Attitudes toward Abortion: A Review and Research Agenda.” Political Research Quarterly 56:489–500. [Google Scholar]
  36. Jennings M. Kent, Markus Gregory B. 1984. “Partisan Orientations over the Long Haul: Results from the Three-Wave Political Socialization Panel Study.” American Political Science Review 78:1000–1018. [Google Scholar]
  37. Jennings M. Kent, Stoker Laura, Bowers Jake. 2009. “Politics across Generations: Family Transmission Reexamined.” Journal of Politics 71:782–99. [Google Scholar]
  38. Karp Jeffrey A., Banducci Susan A. 2008. “When Politics Is Not Just a Man’s Game: Women’s Representation and Political Engagement.” Electoral Studies 27:105–15. [Google Scholar]
  39. Kellstedt Lyman A., Green John C., Guth James L., Smidt Corwin E. 1996. “Grasping the Essentials: The Social Embodiment of Religion and Political Behavior.” In Religion and the Culture Wars: Dispatches from the Front, edited by Green John C., Guth James L., Smidt Corwin E., Kellstedt Lyman A., 174–92. New York: Rowman and Littlefield Publishers. [Google Scholar]
  40. Lewis-Beck M. S., Jacoby William G., Norpoth H., Weisberg H. F. 2008. The American Voter Revisited. Ann Arbor: University of Michigan Press. [Google Scholar]
  41. Luker Kristin. 1984. Abortion and the Politics of Motherhood. Berkeley: University of California Press. [Google Scholar]
  42. Maxwell Carol J. C. 2002. Pro-Life Activists in America: Meaning, Motivation, and Direct Action. New York: Cambridge University Press. [Google Scholar]
  43. Miller Steven D., Sears David O. 1986. “Stability and Change in Social Tolerance: A Test of the Persistence Hypothesis.” American Journal of Political Science 30:214–36. [Google Scholar]
  44. Min Joohong, Silverstein Merril, Lendon Jessica P. 2012. “Intergenerational Transmission of Values over the Family Life Course.” Advances in Life Course Research 17:112–20. [Google Scholar]
  45. Niemi Richard G., Katz R. S., Newman D. 1980. “Reconstructing Past Partisanship: The Failure of the Party Identification Recall Questions.” American Journal of Political Science 24:633–51. [Google Scholar]
  46. Pacheco Julianna Sandell. 2008. “Political Socialization in Context: The Effect of Political Competition on Youth Voter Turnout.” Political Behavior 30:415–36. [Google Scholar]
  47. Petchesky Rosalind Pollack. 1990[1984]. Abortion and Woman’s Choice: The State, Sexuality & Reproductive Freedom. Ann Arbor, MI: Northeastern University Press. [Google Scholar]
  48. Pew Research Poll. . 2013. “Roe v. Wade at 40: Most Oppose Overturning Abortion Decision.” http://www.pewforum.org/2013/01/16/roe-v-wade-at-40/#importance.
  49. Plutzer Eric. 1988. “Work Life, Family Life, and Women’s Support of Feminism.” American Sociological Review 53:640–49. [Google Scholar]
  50. ———. 2002. “Becoming a Habitual Voter: Inertia, Resources, and Growth in Young Adulthood.” American Political Science Review 96:41–56. [Google Scholar]
  51. Sandell Julianna, Plutzer Eric. 2005. “Families, Divorce, and Voter Turnout.” Political Behavior 27:133–62. [Google Scholar]
  52. Sears David O. 1989. “Whither Political Socialization Research? The Question of Persistence.” In Political Socialization, Citizenship Education, and Democracy, edited by Ichilov Orit, 69–97. New York: Teachers College Press. [Google Scholar]
  53. Thomas Sue. 1991. “The Impact of Women on State Legislative Policies.” Journal of Politics 53:958–76. [Google Scholar]
  54. Wilcox Bradford W, 2002. “Religion, Convention, and Paternal Involvement.” Journal of Marriage and Family 64:780–92. [Google Scholar]
  55. Wilcox Clyde, Norrander Barbara. 2002. “Of Mood Morals: The Dynamics of Opinion on Abortion and Gay Rights.” In Understanding Public Opinion, 2nd ed, edited by Norrander Barbara and Wilcox Clyde, 121–48. Washington, DC: CQ Press. [Google Scholar]
  56. Windett Jason Harold. 2011. “State Effects and the Emergence and Success of Female Gubernatorial Candidates.” State Politics & Policy Quarterly 11:460–82. [Google Scholar]
  57. Wolak Jennifer. 2009. “Explaining Change in Party Identification in Adolescence.” Electoral Studies 28:573–83. [Google Scholar]
  58. Wolak Jennifer, McDevitt Michael. 2011. “The Roots of the Gender Gap in Political Knowledge in Adolescence.” Political Behavior 33:505–33. [Google Scholar]
  59. Zaller John. 1992. The Nature and Origins of Mass Opinion. New York: Cambridge University Press. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_80_1_66__index.html (689B, html)
supp_nfv050_poq_14_0205_File004.pdf (550.1KB, pdf)

Articles from Public Opinion Quarterly are provided here courtesy of Oxford University Press

RESOURCES