According to the model of political ideology as motivated social cognition (Jost et al., 2003, 2009; Jost and Amodio, 2012), ideology meets epistemic, existential, and relational needs of certainty, security, and affiliation. In other words, a dispositional or situational heightened motivation to reduce uncertainty, threat or isolation drives the preference for stability over change and the acceptance of inequalities, which are the two core aspects of conservative ideology.

To start with epistemic motives, individual differences pertaining to the management of uncertainty have been shown to predict conservative positions. Indeed, individuals characterized by higher need for structure (Altemeyer, 1998), higher Need for Cognitive Closure (e.g., Kemmelmeier, 1997; Chirumbolo, 2002; Chirumbolo et al., 2004; Van Hiel et al., 2004; Jost et al., 2007; Leone and Chirumbolo, 2008; Hennes et al., 2012), and lower Need for Cognition (e.g., Sargent, 2004; Federico and Schneider, 2007) appear to also embrace more conservative views of the world. Not only dispositional differences, but also situational factors affect ideological positions. For instance, distractions or high cognitive load induce people to adopt more conservative attributional styles (Skitka et al., 2002).

To continue with security needs, conservatives have also been shown to be more sensitive to threat as compared to liberals. As far as existential fear is concerned, a line of research suggests that death anxiety tends to be associated with political conservatism, right-wing authoritarianism and system justifying beliefs (Wilson, 1973; Jost et al., 2003, 2007; Hennes et al., 2012). Experimental studies also showed that manipulations of mortality salience (e.g., Landau et al., 2004; Cohen et al., 2005; Nail et al., 2009) and terrorism salience (Ullrich and Cohrs, 2007) tend to shift people's political opinions and preferences toward conservatism (but see also Kosloff et al., 2010; Lambert et al., 2010 for different results). Besides fear, disgust is also associated to political ideology. Indeed, both correlational and experimental studies (Hodson and Costello, 2007; Inbar et al., 2009, 2012; Terrizzi et al., 2010; Helzer and Pizarro, 2011), indicated that individuals holding conservative opinions are more sensitive to disgust than liberals.

In a related vein, conservatives are more likely than liberals to perceive the world as a dangerous place (Altemeyer, 1998; Duckitt et al., 2002; Mirisola et al., 2014) and to interpret ambiguous facial stimuli as expressing threatening emotions (Vigil, 2010). Conservatives are also more cautious in exploring novel situations and give more weight to negative as compared to positive information in impression formation tasks (Shook and Fazio, 2009; Castelli and Carraro, 2011; Carraro et al., 2014). In addition, conservatives have been shown to display an automatic selective attention to negative stimuli (Carraro et al., 2011; for a review cf. Hibbing et al., 2014).

Finally, relational motives also seem to predict the adoption of conservative ideology (see Jost et al., 2008a). For example, right-wing and authoritarian people have been found to value conformity better than liberals (Feldman, 2003; Cavazza and Mucchi-Faina, 2008) and those endorsing conservative views also have a higher need to achieve a shared view of reality (Hennes et al., 2012). In addition, an experimental study (Jost et al., 2008a) showed that affiliation motives could actually affect political ideology. Indeed, students who had imagined an interaction with their more conservative parent subsequently displayed higher system-justifying beliefs than those who had imagined an interaction with their liberal parent.

Concluding, research involving adult respondents has overall demonstrated that conservatives are characterized by higher epistemic, existential and relational needs as compared to liberals. We thus expected that children of conservative parents would also be more likely to display higher need for certainty, security, and affiliation than children of liberal parents.

The investigation of the structure of socio-political and socio-cultural attitudes and values (for a review, cf. Duckitt, 2001) typically elicited two relatively orthogonal dimensions broadly corresponding to Right Wing Authoritarianism (RWA; Altemeyer, 1981, 1988) and Social Dominance Orientation (SDO; Sidanius and Pratto, 1993, 1999). Initially conceived as a personality dimension (Adorno et al., 1950; Altemeyer, 1981), RWA is now considered as a collection of generalized attitudes and beliefs (e.g., Duckitt, 2001; Duckitt et al., 2002; Van Hiel et al., 2004), likewise SDO, which was originally designed as such.

According to the DPM (Duckitt et al., 2002; Duckitt and Sibley, 2009), RWA and SDO scales capture two distinct and relatively independent ideological attitude dimensions, stemming from different situational factors, personality traits, and social worldviews. Individuals high on social conformity (vs. autonomy, or openness to experience in Big-Five terms) tend to believe that the social world is dangerous and threatening and thus to pursue the motivational goals of security and social control (vs. personal freedom and autonomy) expressed in high RWA scores. Therefore, RWA is also labeled as authoritarian submission. On the other hand, individuals high on tough-mindedness (vs. tender-mindedness, or agreeableness in Big-Five terms) tend to perceive the world as a ruthlessly competitive jungle and thus to pursue the motivational goals of power, superiority over others and dominance (vs. care, share and help) expressed in high SDO scores. Therefore, SDO is also labeled as authoritarian dominance.

In the present study, we measured both authoritarian submission (i.e., RWA; Altemeyer, 1981, 1988) and authoritarian dominance (i.e., SDO; Sidanius and Pratto, 1993, 1999) in order to achieve a more nuanced assessment of the links between parental ideology and children's socio-cognitive responses.

It is generally assumed that political ideology is to a large extent transmitted from parents to children (Adorno et al., 1950; Altemeyer, 1988). A large sample longitudinal set of studies including three generations (Jennings and Niemi, 1968; Jennings et al., 2009) confirmed that, besides the indirect parental influence exerted by placing their children in a given sociopolitical milieu, parents can also have a strong direct influence on children's political learning. This direct influence is likely to be exerted through both socialization and genetic transmission. As for the former process, for instance, Adorno et al. (1950) assumed that parenting styles are both antecedents and consequences of authoritarianism. Indeed, although there is only very weak evidence of an association between parents' ideology and parenting styles (Boshier and Izard, 1972), the latter have been shown to affect children's political ideology (e. g., Miklikowska and Hurme, 2011; Fraley et al., 2012). As for the genetic transmission, several studies (e.g., Eaves and Eysenck, 1974; Tesser, 1993; Olson et al., 2001; Bouchard et al., 2003; Alford et al., 2005; Kandler et al., 2012), comparing monozygotic and dizygotic adult twins, indicate that political ideology can also be genetically determined, with heritability explaining about 50% of the variance in relation to attitudes toward several political issues.

In addition, there are some clues of specific intergenerational transmission for the described two sides of the authoritarianism coin, SDO and RWA. Indeed, young adults' and adolescents' levels of RWA (Altemeyer, 1988; Peterson and Duncan, 1999; Duriez et al., 2008) and SDO (Duriez et al., 2008) have been found to correlate with their parents' responses on the correspondent measures. More importantly, in line with the 90,000 DPM (Duckitt et al., 2002), research indicated that authoritarian submission and dominance are rooted in different socialization experiences. Indeed, Duckitt (2001) has shown that punitive and strict parenting is associated to RWA (through social conformity and dangerous-world beliefs), while neglecting and unaffectionate parenting is associated to SDO (through tough-mindedness and competitive-jungle beliefs).

In a similar way, though from a different perspective, Duriez et al. (2008) found that these distinct dimensions of authoritarianism were specifically predicted by different parental goal promotion efforts that were linked, in turn, to parents' ideology. In particular, conservation (vs. openness to change) goal promotion mediated parent-child correlation in RWA, while extrinsic (vs. intrinsic) goal promotion mediated parent-child correlation in SDO. Focusing on the outcomes of socialization experiences, Weber and Federico (2007) also showed that anxious attachment styles predicted RWA while avoidant attachment predicted SDO, through the correspondent worldviews respectively (but see also Roccato, 2008; Lorito and Falgares, 2013).

It is worth noting that this research on parent-child resemblance in RWA and SDO, such as that on parent-child resemblance in other measures of political ideology (e.g., Jennings and Niemi, 1968; Acock and Bengtson, 1978; Jennings et al., 2009), involved young adults or adolescents as offspring. This is consistent with the general assumption that sociopolitical orientation crystallizes in young adulthood (e.g., Altemeyer, 1998). However, there is reason to predict that those epistemic (e.g., need for closure), existential (e.g., threat sensitivity), and relational (e.g., social conformity) needs that are known to be associated to ideology in adult populations may also be detected among children as a function of the ideological placement of their parents. To date, only two studies tested this hypothesis. In the former (Reifen Tagar et al., 2014), 3–4 year old children's deference to convention and authority was investigated as a function of parents' authoritarian and social conformity values. As predicted, Reifen Tagar et al. (2014) found higher sensitivity to adults' conventionality and status, when deciding whether to trust them in a labeling objects game, among children of parents who were high (vs. low) in authoritarianism and social conformity values. In another study, Dennis et al. (2015) examined 5–7 year old children's neurocognitive responses to conflict as a function of parental ideology and emotional context. Their results showed larger N2 amplitudes (indicating enhanced recruitment of cognitive resources to detect and resolve conflict) in threatening conditions among children of liberal parents as compared to children whose parents embrace more conservative or moderate political views. These results show that psychosocial and neural differences associated with parents' political views can be detected at a very young age, long before the development of sophisticated political beliefs. These fascinating recent lines of research (Reifen Tagar et al., 2014; Dennis et al., 2015), suggesting that a psychological predisposition to embrace certain political views is already present at pre-school age, prompted us to further investigate the differential effects of parents' authoritarianism dimensions on their very young children.

Drawing on the previously described research findings, we here tested the general hypotheses that children of conservative parents would display greater needs for certainty, security, and affiliation as compared to children of liberal parents. In operational terms, given that conservative adults, compared with liberal ones, are more attentive to threatening stimuli (e.g., Carraro et al., 2011), exhibit a stronger implicit preference for order vs. chaos (Jost et al., 2008b) and have more positive attitudes toward social conformity (e.g., Cavazza and Mucchi-Faina, 2008), we expected that children of conservative (vs. liberal) parents would score higher on measures of negativity/threat bias (assessed as proxies of existential needs), implicit attitudes toward order vs. chaos (as a proxy of epistemic needs), and social conformity to peers (as a proxy of relational needs). Since it is not clear whether conservatives are biased toward negative stimuli in general or only in relation to threatening stimuli (Lilienfeld and Latzman, 2014), we included measures that enabled us to assess attentional biases elicited by both threatening stimuli and negative stimuli that are not expected to trigger any threat.

Moreover, besides these general hypotheses, we also explored how the distinct dimensions of parental authoritarianism (i.e., SDO and RWA) might be differentially associated to the various dependent measures administered to the children. To the best of our knowledge, past research is not very informative about the differential effect of the socio-cognitive motives underlying ideology Jost et al., 2003 on authoritarian submission and dominance (but see Van Hiel et al., 2004 for an exception). These constructs are often addressed by separate literatures, which have not yet been fully integrated, but specific hypotheses can be put forward. First, we expected that parental RWA, as rooted in a view of the world as a dangerous place, would predict children's sensitivity to threat and could predict their sensitivity to negativity. Second, we expected that parental SDO, as based on a perception of the world as a competitive jungle, might lead to a higher need for order in children, foreshadowing a preference for hierarchy as a form of regularity and arrangement. In addition, following Van Hiel et al. (2004) who showed that adults' need for simple structure predicted not only their SDO but also their RWA score, we hypothesized that parental RWA would also be associated with children's need for order. Finally, as for the relational needs of children, we expected that whereas parental authoritarian submission, valuing obedience and conventionalism, should increase children's tendency to conform, authoritarian dominance, characterized by a need of prevailing and being superior, might reduce children's social conformity to peers.

Finally, in the present study we separately investigated the impact of mothers' and fathers' ideology on children's responses. Indeed, having an independent assessment of ideology from both parents is crucial on order to estimate the role of each parental figure while controlling for the eventual influence of the other parent. The scarcity of previous studies prevents us from formulating straightforward hypotheses, although some research findings suggest that mothers might be in general more influential than fathers (Acock and Bengtson, 1978; Castelli et al., 2009).

This study was carried out in accordance with the recommendations of the ethical committee for psychological research at the University of Padova. We recruited participant families in 4 kindergartens in a medium sized town in Northern Italy. One-hundred and six children¹ (50 males) aged 4–6 years (M = 5.18, SD = 0.56) were individually interviewed in a quiet room by a researcher after having obtained written permission from the parents in accordance with the Declaration of Helsinki. Children were asked to perform a series of tasks (presented as games) on a computer laptop. In order to simplify the task, we used a keyboard with three different colored labels on the 3 response keys (blue for the left key, yellow for the central key and red for the right key). At least one parent for each child filled in a questionnaire at home and returned it in a closed envelope. Overall, we obtained responses from 85 fathers and 102 mothers, aged 24–62 years (M = 41.17, SD = 5.52), for a total of 83 complete triads. As for the educational level of the involved parents, 40.2% of them declared a senior high-school leaving accreditation and 41.8% had a university level degree.

Parents completed the Italian version (Roccato et al., 2009) of the Right Wing Authoritarianism scale by Funke (2005; 12 items) and an Italian short version (Di Stefano and Roccato, 2005; 8 items) of the Social Dominance Orientation Scale (Pratto et al., 1994). Responses were provided on 5-point Likert scales ranging from “completely disagree” to “completely agree.” We computed average scores for each measure for both mothers (RWA α = 0.63; SDO α = 0.81) and fathers (RWA α = 0.64; SDO α = 0.87).

As measures of authoritarianism could be inflated by social desirability bias (Taylor, 1961), we followed other authors (e.g., Duriez and Van Hiel, 2002; Ekehammar et al., 2004) in administering the Short Marlowe-Crowne Social Desirability Scale (Reynolds, 1982; 13 item). Participants answered on 5-point Likert scales ranging from “completely false” to “completely true.” An average score was computed for each parent (mothers α = 0.62; fathers α = 0.74).

Results of linear regression analyses showed that fathers' SDO scores were predictive of children's IAT scores, so that children displayed a stronger preference for order as their fathers' level of authoritarian dominance increased. No effect was found in relation to mothers' SDO, as well as in relation to the RWA scores of both mothers and fathers.

As for the social conformity task, the RWA score of mothers (but not that of fathers) significantly and positively predicted children's responses, while the SDO score of fathers (but not that of mothers) negatively predicted them. In other words, children were more conformist to allegedly erroneous peers as their mothers' level of authoritarian submission increased and their fathers' level of authoritarian dominance decreased.

Finally, the RWA score of the mothers was a positive and significant predictor of children's threat bias, whereas the RWA of the fathers was unrelated to children's responses in the task (see Table 3). Interestingly, no effect emerged in relation to SDO confirming, as predicted, that authoritarian submission, as captured by responses to the RWA scale, is a stronger predictor, as compared to SDO, of the tendency to carefully monitor for the presence of threatening stimuli in the environment. Neither SDO nor RWA scores were related to responses in the Dot-probe task in which negative but not threatening stimuli were presented. This further suggests that mothers' RWA is likely to specifically influence children's security needs, as indexed by their tendency to carefully monitor for stimuli that are not only negative but also able to elicit threat.

As previously remarked, this study only represents an initial exploration of the influence of parents' authoritarianism dimensions on children's socio-cognitive responses and thus several limitations must be acknowledged. First, one might appropriately observe that RWA scale is tridimensional and thus is not a “pure” measure of submission (Duckitt and Bizumic, 2013). However, it has to be noted that we used a brief version of the RWA scale, with only 12 items (Funke, 2005; Roccato et al., 2009). This scale has been found to have a three-dimensional factorial structure (defined by authoritarian submission, authoritarian aggression, and conventionalism), but the authors of the Italian validation (Roccato et al., 2009) highlighted the poor reliability of the three substantive factors and the very strong correlations among them, thus warning against mechanically considering the brief version of the scale as three-dimensional. Therefore, we also decided to calculate a single score for the RWA scale, observing that the reliability of the overall scale was relatively low, but still acceptable (i.e., α = 0.63 for mothers and α = 0.64 for fathers), whereas the reliability of the subscales was very low.

In addition, it has to be acknowledged that despite the significant results reported and discussed above, in the current study several effects were non-significant. Therefore, even the single significant effects that have been presented here should be placed in the context of an overall fuzzy picture. For this reason, future research recruiting a wider sample is necessary in order to further investigate and better understand the relations between parents' SDO and RWA and their children socio-cognitive motives. Moreover, the current mixed results may also indicate that the relations between the studied variables are very complex and somehow relatively weak and, as a consequence, unstable. It is likely that other variables do play an important role and thus future studies may include some moderators while analyzing these relations. Our position is that the mixed findings presented in the present paper, although suggestive, should be taken with extreme caution being in no way conclusive about an actual link between parents' ideology and children's socio-cognitive responses. Nevertheless, both the significant and non-significant results obtained in the current study may represent important additional elements for a comprehensive meta-analytic approach to the issue that might hopefully provide more conclusive responses.

Finally, it has to be noted that this is a cross-sectional correlational study, thus causal relations cannot be determined. Although mutual influence cannot be excluded (Miklikowska, 2016), we can reasonably assume that parental ideological beliefs are more likely to affect children's responses, rather than the reverse, at least in the case of very young respondents, such as the preschoolers involved in our study. However, longitudinal research is necessary in order to provide more straightforward results not only about the casual relations but also about the development stability of these embryonic differences. Some studies have already shown that some features detected in children (e.g., early temperament) may be correlated with political ideology in adulthood (e.g., Block and Block, 2006; Fraley et al., 2012) suggesting that children's behaviors can actually be predictive of later more sophisticated political attitudes.

Notwithstanding these limitations, the present work adds to the recent and so far limited literature suggesting the existence, at a very young age, of ideology-based differences in children of conservative and liberal parents, also providing some hints about the possible different and specific influences of the two parents and of the two different aspects of authoritarianism.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.