Parenting and Politics : Exploring Early Moral Bases of Political Orientation

Based on Lakoff’s (2002) Strict Father and Nurturant Parent metaphors for political conservatism and liberalism respectively, two studies explored parenting styles, political ideology, and the moral orientations that might link the two. Restrictive parenting (by both mother and father) predicted political conservatism, and this path was mediated by a strong Social Order orientation (Study 1) reflecting, more broadly, an inhibition-based proscriptive morality (Study 2). Political liberalism was associated with a Social Justice orientation, but was not predicted by nurturant parenting in either study. Study 1 included mothers’ reports of their own parenting, and these were correlated with the students’ responses. Findings support a restrictive moral underpinning for conservatism, but raise questions about the assumed unique association between parental nurturance and political liberalism, which is addressed in the discussion.

In contrast, in the Nurturant Parent model children become responsible "through being cared for and respected, and through caring for others … Self-fulfillment and the nurturance of others are seen as inseparable" (Lakoff, 2002, p. 108).Here the child must learn how to cooperate, empathize with others, and maintain social ties, and this is done through a relationship of mutual respect, caring, and two-way communication with parents.Children are regarded as naturally good, and their inherent morality is cultivated through parental nurturance and understanding.Both the mother and father guide their children by example, and cooperation rather than competition is emphasized.Lakoff (2002) proposes that the Strict Father and Nurturant Parent models respectively underlie the distinct worldviews of conservatives and liberals.He therefore suggests that the framing of political issues should coincide with these family-based models; that is, for maximum effect conservative messages should be based in Strict Father metaphors, and liberal communications should reflect Nurturant Parent metaphors.Interestingly, Lakoff's framing recommendations have had a strong impact in the real world of politics (see, e.g., Bai, 2005).

Parenting and Political Orientation
The metaphors themselves also suggest that family of origin plays an important role in the development of political orientation.More specifically, family of origin provides strong experiential data for one's preferences and understanding of family, which are then more broadly applied to the realm of nation and politics; the conception of the family is extended to the nation, and strictness versus nurturance is translated into beliefs about optimal forms of governance.Lakoff's metaphors, then, suggest a causal relationship between actual parenting style and children's political orientation, a path that has received support in past work linking authoritarian parenting (i.e., obedience to authority, use of punishment) and political conservatism (e.g., Adorno, Frenkel-Brunswick, Levinson, & Sanford, 1950;Altemeyer, 1988;Duckitt, 2001;Tomkins, 1995;Wilson, 1973).Although the findings have not been entirely consistent (see, e.g., Altemeyer, 1981, on the role of social learning in adolescence), strong support for the authoritarian-conservative link was found in a recent longitudinal study by Fraley, Griffin, Belsky, and Roisman (2012).These researchers found that parents' authoritarianism when children were one month old predicted the children's conservative attitudes 17 years later.Fraley et al. (2012) also found a link between nurturance and political liberalism, a focus of far fewer empirical studies.Yet in recent research based on life-narrative interviews of politically active adults, McAdams, Albaugh, Farber, Daniels, Logan, and Olson (2008) found that although liberals identified lessons learned about empathy and openness, they were not more likely to describe authority figures as nurturant (McAdams et al., 2008).Conservatives, however, did regard authority figures as "strict enforcers of moral rules" (p.978) and identified important lessons in self-discipline, consistent with Lakoff's Strict Father metaphor.
Past research generally provides support for the conclusion that parenting practices promote political ideologies, with considerable evidence for the Strict Father-conservative link, and less evidence for the Nurturant Parent-liberal path.Interestingly, although Fraley et al.'s (2012) longitudinal study supports both paths from family of origin to political orientation, the authors nevertheless specifically note that their study does not speak to the specific mechanisms involved and "clarifying these pathways is an important direction for future research" (p.1429).Although Lakoff assumes that morality differs for his two family-based models, his focus is on distinct ways of parenting rather than morality per se, and there are no studies to date directly testing differences in moral orientation as mediators of the parenting-to-politics link.

Parenting and Politics 44
In the studies that follow we explored morality as a mechanism to account for the path from parenting to politics.
We focused on the two primary dimensions of parenting identified in socialization research: nurturance and restrictiveness (e.g., Baumrind, 1966Baumrind, , 1967)).Nurturance entails responsive, supportive parental attitudes and practices, including in particular a willingness to acknowledge children's experiences and feelings.In contrast, restrictiveness focuses on control, particularly attempts to control how the child behaves and feels through strict limits, threats, and punishment.These two dimensions clearly parallel the nurturant versus strict parental modes underlying Lakoff's family metaphors.We were interested in further investigating whether restrictive parenting (or fathering, à la Lakoff) is associated with (adult) children's political conservatism, and whether nurturant parenting, as suggested by Lakoff's metaphor, is associated with (adult) children's political liberalism.However, we also explored distinct moral orientations, derived from psychological work on self-regulation, as possible mechanisms for better understanding the path from parents' practices to children's politics.

Two Distinct Moral Orientations
More specifically, we investigated whether the possible paths from restrictiveness to conservatism and from nurturance to liberalism are mediated, respectively, by differences in prescriptive and proscriptive moral regulation (Janoff-Bulman, Sheikh, & Hepp, 2009; also see Janoff-Bulman & Carnes, 2013).The primary motivational distinction in psychology is the difference between approach versus avoidance regulation, with approach focused on positive outcomes and based in behavioral activation, and avoidance focused on negative outcomes and based in behavioral inhibition (Carver, 2006;Carver & Scheier, 1998;Gray, 1982Gray, , 1990;;Higgins, 1997Higgins, , 1998;; for reviews, see Carver &Scheier, 2008, andGable, Reis, &Elliot, 2003).Prescriptive and proscriptive morality reflect differences in approach and avoidance regulation in the moral domain; prescriptive morality is activation-based and focused on approaching positive outcomes, whereas proscriptive morality is inhibition-based and focused on avoiding negative outcomes (Janoff-Bulman et al., 2009).
Proscriptive morality is premised on inhibiting the "bad" (e.g., lying, cheating), whereas prescriptive morality involves enabling the "good" (e.g., helping, working hard prescriptive morality provides for well-being.In a series of studies Janoff-Bulman et al. (2009) found that proscriptive morality is focused on transgressions and is responsive to threat; it is mandatory and emphasizes blameworthiness.
In contrast, prescriptive morality is focused on good deeds and is not responsive to threat; it is more discretionary and emphasizes credit-worthiness.Overall, proscriptive moral regulation is condemnatory and strict, while prescriptive morality is commendatory and less strict.
Interestingly, each parenting dimension-restrictiveness and nurturance-seems to have a regulatory function that parallels distinct proscriptive versus prescriptive moral orientations (Sheikh & Janoff-Bulman, 2010).Parental restrictiveness is an inhibition-based, avoidant parenting strategy based in controlling the child through strong parental monitoring, rigid limits, and the use of threats and punishment.With its emphasis on strictness and control, restrictive parenting seems likely to socialize a proscriptive moral orientation-one that is sensitive to negative outcomes and focused on prohibitions and inhibiting bad behavior.The emphasis on parental responsiveness, affective praise, and encouragement in nurturant parenting seems likely to socialize a prescriptive orientation, based in doing the right thing, particularly helping others, out of love and caring rather than threats and fear of punishment.Parenting styles seem to provide distinct emphases on prohibitions and threat-sensitivity versus ac-tivation and greater openness, orientations associated respectively with proscriptive and prescriptive morality (Janoff-Bulman et al., 2009).

Political Orientation and Moral Regulation: Social Order Versus Social Justice
Yet how are these proposed moral distinctions related to political orientation?Why might they provide a way to better understand the path from parenting to politics, and more specifically from restrictive parenting to political conservatism and from nurturant parenting to political liberalism, as suggested by Lakoff's metaphors?Morality applies not only to how we regulate our own behavior and protect or provide for another person or persons, but also to the broader collective and choices about how a society protects or provides for its members; that is, morality is not solely about self-regulation, but about social regulation as well (see Janoff-Bulman, 2009;Janoff-Bulman & Carnes, 2013).The social regulation of morality is the domain of politics, for political orientation is associated with distinct policy preferences for protecting and providing for a society.
In past work Janoff-Bulman and colleagues (e.g., Janoff-Bulman, 2009;Janoff-Bulman & Carnes, 2013;Janoff-Bulman, Sheikh, & Baldacci, 2008) have linked the two moral-regulatory orientations to political orientation.More specifically, they have argued that conservatism is based in inhibition/avoidance motivation and proscriptive morality in particular; and liberalism is based in activation-based approach motivation and prescriptive morality i .
Conservatism is tied to a proscriptive moral orientation, which is focused on protecting self and others from harm through a concern with order, restraint, and the avoidance of negative outcomes; in contrast, liberalism is tied to a prescriptive orientation, which is focused on providing for the well-being of self and others through an interest in activating the "good" (Janoff-Bulman, 2009;Janoff-Bulman et al., 2008).From the perspective of politics, these two moral orientations are most evident in distinct group-based moralities: social order versus social justice (for a more detailed discussion, see Janoff-Bulman & Carnes, 2013).More specifically, proscriptive morality, with its emphasis on prohibitions and restraint, is evident in social regulations that focus on conformity, order, and the minimization of "deviance"; here group-based morality emphasizes social order concerns.This group-based morality follows from a motivation to protect the group from dangers and threats, from both outside and inside the group, including physical threats to safety and stability as well as psychological threats to identity.The importance of order and solidarity is evident in the emphasis on strict norm adherence, conformity behaviors and group loyalty, which serve to maximize group cohesion.
In the case of a prescriptive moral orientation, social regulations are more apt to reflect an emphasis on positive social actions and communal responsibility, and group-based morality is likely to focus on social justice concerns.
Here the moral concern is providing for the welfare of the larger community.Social justice activates group-based efforts to help, with a particular focus on equality-oriented distributional justice.Group bonds are strengthened through a shared sense of responsibility.

Parenting and Politics
In the two studies that follow, we explored the relationships among parenting style, politics, and moral orientation.
The United States provided the cultural context for this work, just as it is the context for Lakoff's political metaphors.
We hypothesized that parental restrictiveness would be positively associated with political conservatism, and this association would be mediated by a proscriptive moral orientation; and parental nurturance would be positively associated with political liberalism, and this association would be mediated by a prescriptive moral orientation.
Both studies explored moral orientation as a possible mechanism for better understanding the proposed parentingpolitics link popularized in Lakoff's work.Study one focused on Social Order versus Social Justice, the groupbased manifestations of the two types of morality, and Study 2 focused on more general indices of prescriptive versus proscriptive morality.

Study 1
Study 1 investigated the links between parents' restrictiveness and nurturance in a sample of college students in the U.S. and in particular explored distinct group-based moralities, moral orientations applied specifically to social regulation (proscriptive: Social Order; prescriptive: Social Justice), as possible mediators of the hypothesized parenting-politics link.To provide some validation of the undergraduates' retrospective reports, an assessment of parenting style was also sought from each student's mother regarding her own restrictive and nurturant parenting style.

Method Participants
A total of 191 students (156 women and 34 men) from the subject pool at a large public university in the northeast United States completed the study.Of these, 143 identified themselves as White, 25 as Asian, 9 as Latino/Hispanic, 8 as Black, and 5 as Other.A total of 103 mothers of these students also completed the parenting scales (see Procedure below).The mean age of participants was 20.21 years.
Parenting Questionnaire Rickel and Biasatti's (1982) Child Rearing Practices Report (CRPR) was used to assess two parenting styles: restrictiveness and nurturance.Based on Block's (1965) 91-item CRPR, this scale is the result of statistical analyses by Rickel and Biasatti that found two reliable factors based on 40 items reflecting restrictiveness and nurturance.
The CRPR is a respected measure that continues to be used in research ii .No items of the scale were removed for this study.For this research the scale was modified for the undergraduate sample to have children describe their parents rather than have parents describe themselves.Each item was assessed for a mother and a father figure separately, and presentation was counterbalanced.The original version of the scale (first person, from the parent's perspective) was administered to mothers of the undergraduate participants.We chose mothers to provide information on parenting style, based on the methodology of DeHart, Pelham, and Tennen (2006) in their study of college students' self-esteem and parenting style.Tapping one parent, and in particular the mother, was found to be particularly successful in their research.
Sample statements from the 22-item Restrictiveness scale included: "Thought a child should be seen and not heard"; "Used to control me by warning me of all the bad things that could happen to me"; "Encouraged me to keep control of my feelings at all times"; "Thought that scolding and criticism would make me improve"; " Did not allow me to question his/her decisions"; "Taught me that in one way or another punishment would find me when I was bad."Sample statements from the 18-item Nurturance scale included: "Gave me comfort and understanding when I was scared or upset"; "Talked it over and reasoned with me when I misbehaved"; "Expressed affection by hugging, kissing, and holding me"; "Encouraged me to talk about my troubles"; "Emphasized praising me when I was good more than punishing me when I was bad"; "Took into account my preferences when making plans for the family."All responses were made on 7-point scales (endpoints 1 = "not at all" and 7 = "very much so").

Political Orientation
Two items measured liberalism-conservatism within the contemporary U.S. political landscape.Respondents were asked to indicate where they would place themselves on two 7-point scales, one with endpoints 1 = "Very Liberal" and 7 = "Very Conservative" and the other with endpoints 1 = "Strong Democrat" and 7 = "Strong Republican."The items were averaged to provide a single measure of Political Orientation (α = .839),with higher numbers indicating greater political conservatism.It should be noted that if only the single liberal-conservative scale was used in analyses, the pattern of results was unchanged.
These items were drawn from Janoff-Bulman et al. (2008).The Social Order items were: "By bucking tradition and choosing new lifestyles, people are actually threatening the societal bonds that hold us together"; "In a decent society, people should not be completely free to make their own choices about how to live their own lives, but should attend to community standards"; and "People should not be completely free to express themselves through their own choice of lifestyle, even if they don't harm others."Examples of Social Justice items are: "In the healthiest societies those at the top feel responsible for providing better lives for those at the bottom"; "It's an obligation, not a matter of personal preference, to provide for groups worse off in society"; and "It's important for those who are better off in society to help provide resources for those who are worse off."Responses were made on 7-point scales with endpoints 1 = "strongly disagree" and 7 = "'strongly agree."

Procedure
After signing a consent form, participants completed two 40-item CRPR scales, one for mother figure and one for father figure (counterbalanced across participants), followed by the Social Order and Social Justice items and a brief demographics questionnaire that included the assessment of political orientation.They were then asked to address a large manila envelope to their mother at her home address so that she could complete the parenting questionnaire herself, regarding her own parenting style.We stressed the voluntary nature of addressing the envelope and having the questionnaire mailed to mothers.Only 12 students chose not to address the envelope.
Further, in the instructions to the mothers, they were assured that their participation was completely voluntary and there would be "no penalties or consequences of any kind" to their son or daughter if they chose not to participate in the study.
Parental Restrictiveness and Parental Nurturance were uncorrelated, r(191) = -.07,p =.337.Table 1 presents the means and standard deviations and Table 2 the correlations for all Study 1 variables.Father: b =.160, SE = .083,p =.056), suggesting that it is not father restrictiveness alone that is important, as suggested by Lakoff (2002), but rather parental restrictiveness of both mother and father, which are correlated in any case.Paralleling the null results for Parental Nurturance, neither Mother nor Father Nurturance was associated with political liberalism or political ideology more generally (Mother: b = .149,SE = .096,p =.109; Father: b = .080,SE =.062, p =.202).
iii Thus, the hypothesized link between parental restrictiveness and political conservatism was supported in this study, but the proposed link between parental nurturance and political liberalism was not supported.Father Nurturance: b = .046,SE = .069,p = .506).Although Social Order was associated with greater political conservatism, and Social Justice with greater political liberalism, only Social Order was associated with parenting style, and in particular with students' reports of their parents' restrictiveness; Social Justice was not associated with students' reports of their parents' nurturance.

Mediation Analysis
A multiple mediation analysis tested the direct effect of Parental Restrictiveness and the indirect effects of both Social Order and Social Justice on political liberalism/conservatism. Using the Hayes macro, we tested mediation by conducting bootstrapping analyses based on 10,000 bootstrap resamples.The bias-corrected 95% confidence interval for Social Order did not include zero (lower bound = .0294;upper bound = .2004),confirming that this indirect path was significant.In contrast, the bias-corrected 95% confidence interval for Social Justice included zero (lower bound = -.1043;upper bound = .0356),indicating that this indirect path was not significant.The direct effect of Parental Restrictiveness on political conservatism was not significant (lower bound = -.0430;upper bound = .3359).In other words, Social Order, but not Social Justice, mediated the relationship between Parental Restrictiveness and greater conservatism.
Given that political ideology was assessed on bipolar scales, greater parental restrictiveness predicted political conservatism, but conversely, lower parental restrictiveness was associated with greater political liberalism as well.It might be argued, then, that minimal parental restrictiveness may be a route to political liberalism, via low Social Order concerns.However, this does not account for the strong association between Social Justice and political liberalism found in this study.Perhaps in addition to a path involving low Social Order concerns, there is also a (low) restrictiveness-liberalism link mediated through Social Justice.However, not only was Parental Restrictiveness uncorrelated with Social Justice (r(191) = .06,p = .423),but in the mediation analysis the indirect effect of Social Justice on political orientation (i.e., liberalism/conservatism) was not significant.

Mothers' Responses
A total of 103 mothers of participants returned the questionnaires assessing their perceptions of their own parenting style.The sample of 103 mothers who returned the parenting questionnaire was compared to the group of 88 mothers for whom no information was available.The mothers who returned the questionnaire were reported to be both more nurturant and less restrictive by their college-aged children (Nurturance: Ms = 5.70 [SD = .78]vs. 5.35 [SD = .99],t(189) = -2.79,p = .006;Restrictiveness: Ms = 3.75 [SD = .84]vs. 4.14 [SD = .85],t(189) = 3.19, p = .002).
Interestingly, when comparing the responses of the 103 mothers who returned the questionnaire with the responses of their children (n = 103), mothers' self-reports were significantly higher on nurturance and lower on restrictiveness than the respective scores reported by their college-aged children (nurturance: Ms = 6.16 vs. 5.70, t(102) = -6.745,p < .001;restrictiveness: Ms = 3.45 vs. 3.75, t(102) = 2.89, p = .005).Although it is impossible to ascertain which responses are more accurate, it is interesting to note that the mothers' scores appeared to reflect some self-enhancement when compared to the students' scores.Despite these differences, there were nevertheless significant positive associations between the students' perceptions of their mother's parenting and the mothers' self-perceptions.The mothers' self-reported Nurturance (α = .810)was positively associated with students' reports of their mother's Nurturance, r(103) = .508,p < .001,and with Parental Nurturance more generally, r(103) = .405,p < .001.Similarly, mothers' self-reported Restrictiveness (α = .829)was positively associated with students' reports of their mother's Restrictiveness, r(103) = .257,p < .01,and with Parental Restrictiveness more generally, r(103) = .225,p < .05.The stronger correlations for nurturance than restrictiveness likely reflect the unambiguously positive evaluation of nurturance for both parents and children.
The perception of restrictiveness is likely to be more varied, with some regarding it positively and others less so; presumably those who value restrictiveness would regard this style as beneficial and socially appropriate.
The significant correlations between reported parenting styles of mothers and their college-aged children provide some support for the validity of the students' self-reports and are particularly noteworthy given the restricted nature of the sample of mothers who returned the questionnaire.Not only was the sample considerably smaller, but in terms of parenting style it was a less restrictive and more nurturant sample of mothers than the mothers represented in the larger student sample as a whole.Thus, it was not surprising that regression analyses did not find mothers' self-reported restrictiveness or nurturance associated with students' political orientation.Nevertheless, mothers' reports of their own restrictiveness marginally predicted students' Social Order scores, b = .229,SE =.128, p = .075,the most proximal link from parental restrictiveness on the path to conservatism.

Parenting and Politics
Past work has found that the proscriptive and prescriptive scenarios are uniquely correlated with Carver and White's (1994) measures of the Behavioral Inhibition System (i.e., avoidance orientation) and the Behavioral Activation System (i.e., approach orientation) respectively (Janoff-Bulman et al., 2009).
Participants rated the extent to which they felt the subject in each scenario should or should not perform the behavior, from 1 ("feel very strongly he/she should not") to 9 ("feel very strongly he/she should").Each scenario focuses on an everyday behavior; the proscriptive items involved inhibition of temptations and restraint of behavior in the interests of greater social conformity-in other words, behaviors the person presumably should not engage in to be considered moral.Examples include "excessive" gambling, going into greater debt to purchase an expensive TV, and wearing a skimpy dress to a funeral.The latter scenario is written as follows: "Sheila is going to a funeral, and it's an unusually hot day.She is thinking of wearing a skimpy, revealing dress to keep her relatively cool at the funeral."Prescriptive items involved activation of helping behaviors and self-reliance-that is, behaviors the person presumably should engage in to be considered moral.A sample prescriptive item is: "While on campus, Jay is approached by a student asking if he could volunteer two hours this weekend to help with a food drive for the local survival center.Jay doesn't have plans for the weekend.Jay is deciding whether to commit himself to helping with the food drive."Higher scores on the prescriptive items indicated a stronger prescriptive orientation.
To calculate the mean scores for proscriptive moral judgments so that higher scores indicated stronger proscriptive orientation, we subtracted participants' scores from 10 (see Janoff-Bulman et al., 2009).

Procedure
After signing a consent form, participants completed the 40-item CRPR separately for mother figure and father figure, followed by the Moralisms Scale and a brief demographics questionnaire that included the items assessing political orientation.
iv The greater the reported Parental Restrictiveness, the more conservative the student participant, but students' reports of Parental Nurturance were once again not associated with political liberalism.A multiple mediation analysis tested the direct effect of Parental Restrictiveness and the indirect effects of both Social Order and Social Justice on political liberalism/conservatism.We again tested mediation using the Hayes macro by conducting bootstrapping analyses based on 10,000 bootstrap resamples.The bias-corrected 95% confidence interval for Proscriptive Morality did not include zero (lower bound = .0108;upper bound = .0955),confirming that this indirect path was significant.In contrast, the bias-corrected 95% confidence interval for Prescriptive Morality included zero (lower bound = -.0404;upper bound = .0048),indicating that this indirect path was not significant).The direct effect of Parental Restrictiveness on political conservatism also was not significant (lower bound = -.0279;upper bound = .2909).In other words, the analysis revealed that Proscriptive Morality, but not Prescriptive Morality, mediated the relationship between Parental Restrictiveness and political conservatism.The Study 2 findings replicated the results of Study 1, although in this case the two types of morality were general measures of proscriptive and prescriptive moral regulation rather than group-based measures of these two orientations.Once again the predicted association between students' reports of their parents' nurturance and political liberalism was not supported.However, as hypothesized, students' reports of their parents' restrictiveness were positively associated with political conservatism, and this relationship was mediated by the students' proscriptive moral orientation.

General Discussion
As found in past work, in both studies parental restrictiveness was associated with political ideology, specifically greater conservatism.In this research we focused on morality to better understand this association.Proscriptive morality--with its emphasis on inhibition and "should nots" and its manifestation in group-based concerns with Social Order-emerged as a mechanism to help account for the path linking parenting and politics.
The restrictiveness-conservatism results provide partial support for Lakoff's (2002) models linking family-based morality to political orientation, although this path was not about the father alone, as suggested by Lakoff's Strict Father metaphor.Instead, both maternal and paternal restrictiveness were associated with offspring's conservative political ideology.In contrast to the restrictiveness-conservative path, no support was found for Lakoff's (2002) proposed Nurturant Parent-liberalism link.Although Social Justice was associated with political liberalism, parental nurturance was not associated with political ideology in either study.
Findings from Study 1 suggest that college students' self-reports of parental behavior are likely to be relatively accurate, or at least they are correlated with mothers' self-reports.Of course parents can be as biased as children in reporting about their own parenting.In fact when looking at only the reports of students whose mothers returned the questionnaire (103 of 191), it appears that the more self-enhancing responses may have been provided by mothers, who viewed their own parenting as more nurturant and less restrictive than did their children.Ultimately the question of who is more accurate-mothers or their college-aged children-remains open.However, consistent with our findings, Brewin, Andrews, and Gotlib (1993) reviewed research on retrospective reports of childhood and similarly concluded that they are "reasonably accurate" and further, that these accounts are stable over time and across psychopathological states such as depression.
Echoing past work, in this research Social Order was positively associated with political conservatism and Social Justice negatively associated with conservatism.(Conversely, Social Justice was positively associated with liberalism and Social Order negatively associated with liberalism.)These two group-based moralities largely reflect the two major dimensions of ideology-economic and social (see Feldman & Johnston, 2013; also see Duckitt, 2001;Jost et al., 2003).Social Justice involves a primary concern with economic issues (e.g., social welfare, universal health care), whereas Social Order reflects a primary concern with social issues (e.g., abortion, gay rights) (see Janoff-Bulman, 2009, for a review; also see Janoff-Bulman, Sheikh, & Baldacci, 2008).In these studies it was high Social Order, and not low Social Justice, that mediated the restrictiveness-parenting link, suggesting the primacy of the social dimension of political conservatism in the path from parenting to ideology.
There was no support for any path from parenting through Social Justice to political ideology, whether liberal or conservative.The conservatives in this sample were likely social conservatives; perhaps a sample of economic conservatives, who would presumably be low on Social Justice but unconcerned with social issues (see Feldman & Johnston, 2013) would be associated with a different set of parenting styles.In this case, however, we did not pre-select based on political ideology, and thus the labels reflected naturally occurring ideologies in this college student sample.
In considering the absence of support for a path from parental nurturance to political liberalism, it is informative to note that Parental Nurturance was relatively high in the undergraduate populations of both studies; this parenting dimension may be regarded as an essential component of good parenting, and students may therefore be motivated to perceive their own parents as nurturant.In addition, understanding, responsive parenting may be particularly well represented in a college student sample.There may thus have been insufficient variation on reported parental nurturance to provide support for a nurturance-liberalism link.
However, there may also be reason to re-evaluate the hypothesized relationship and reconsider the nature of the path from parenting to political liberalism.Our findings replicate those of McAdams et al. (2008), who used a very different research method.Based on detailed interviews of a politically active adult sample, these researchers found that their respondents' life narratives regarding morality provided support for a restrictiveness-conservatism link, but not a nurturance-liberalism path.Perhaps the nature of nurturant parenting itself should give us pause in thinking about the association between parental socialization and political ideology.Nurturance, involving parental support and responsiveness, is typically associated with greater empathy in the socialization literature (see review by Eisenberg, Spinrad, & Sadovsky, 2006).In considering political ideology and empathy, however, it is possible that the difference between liberals and conservatives may not lie in empathy per se, but rather in how far one casts one's "empathic net."From this perspective, we are not apt to see liberal-conservative differences in empathy and prosociality when considering close family and friends; both political groups would be expected to be caring and empathic here.Differences would be expected to arise as we move further away from similar others to societal groups quite different from our own; it is here that social justice concerns in particular play out.
One might suppose that empathy "taught" within the family would generalize to the larger social world.However, work on trust, a related construct, suggests this is not necessarily the case.More specifically, cross-cultural work has found that in Japan and China, for example, there is strong trust within the family, yet low trust of other society members more generally (see, e.g., Fukuyama, 1995;Uslaner, 2002;Yamagishi & Yamagishi, 1994).
In considering proposed liberal-conservative differences at this greater remove from close, familiar others, it may not be the availability of empathic responses in one's repertoire that matters, but rather acceptance of difference and the unfamiliar.This broader empathic net might be captured by the greater openness to experience associated with political liberalism (for reviews, see Jost et al., 2003;McCrae, 1996).Given that openness seems to be the opposite of restrictiveness, at first glance we might focus on low parental restrictiveness in understanding the link between parenting and political liberalism.Yet, as demonstrated by the Study 1 results, low parental restrictiveness may account for liberals' minimal focus on Social Order, but it does not account for liberals' strong Social Justice concerns.It seems important, then, to consider the meaning of restrictiveness as applied to parenting style.Here restrictiveness involves strict control and the use of negative outcomes (threats and punishment) to enforce proper behavior.An appropriate antonym for this type of restrictiveness would be leniency, which does not seem to describe the openness that would allow for greater inclusion of others.Rather, here openness would be defined by tolerance and broad-mindedness; these terms characterize the openness one would expect to be associated with Social Justice, rather than (low) Social Order concerns.These studies were not longitudinal; all findings are correlational, which limits any causal conclusions that can be drawn from this work.A further limitation of this research is that the two studies were based on college samples in the northeastern United States.More varied samples, in terms of economics, age, and culture are necessary to more fully address the important question of generalizability.In particular the university sample in this study was relatively liberal, and this might have impacted findings.As with Lakoff's (2002) work, the cultural context for this research was the United States.It is possible that the two-party system in the United States may exacerbate differences between liberalism and conservatism, particularly given the strong associations between the Republican Party and conservatism and between the Democratic Party and liberalism in contemporary politics.Cross cultural work on the links between parenting, morality, and politics remains a task for future investigations.

b
These two variables are based on about half the total sample (i.e., the 103 of 191 mothers who returned their survey).A multiple regression with students' reports of Parental Restrictiveness and Nurturance as predictor variables revealed that Parental Restrictiveness significantly predicted political orientation, b = .228,SE = .102,p =.026, whereas Parental Nurturance was not associated with political orientation, b = .135,SE = .088,p = .128.Students' reports of higher levels of parental restrictiveness were associated with greater political conservatism.Repeating the regression analysis separately for students' reports of mother and father parenting revealed the same pattern: Mother and Father Restrictiveness each predicted political conservatism (Mother: b = .205,SE = .096,p =.034;

Table 1
Means and Standard Deviations of Study 1 Variables a Higher scores indicate greater conservativsm.

Table 2
Correlations Among Study 1 Variables aHigher scores indicate greater conservatism.

Table 3
Means and Standard Deviations of Study 2 Variables