The Apple Doesn’t “Feel” Far From the Tree: Mother–Child Socialization of Intergroup Empathy

Abstract

Like adults, children experience less empathy toward some groups compared with others. In this investigation, we propose that mothers differ in how much empathy they want their children to feel toward specific outgroups, depending on their political ideology. We suggest that how mothers want their children to feel (i.e., the motivation for their child’s empathy), in turn, is correlated with children’s actual experience of empathy toward the outgroup. Across four studies in the context of the Israeli-Palestinian conflict (N_Total = 734), the degree of empathy mothers wanted their children to experience in the intergroup context varied as a function of their political ideology. Mothers’ motivation for their child’s empathy toward the outgroup (but not in general) was further associated with how they chose to communicate messages to their children in a real-life context and how children actually felt toward the outgroup. We discuss implications for the socialization of intergroup empathy.

Intractable intergroup conflicts are prevalent worldwide. Such conflicts are often perpetuated by a psychological repertoire that includes attitudes, beliefs, and emotions, which are transmitted from one generation to another (Nasie et al., in press). Children learn this psychological repertoire partly through their parents (Degner & Dalege, 2013; Reifen Tagar et al., 2017). Empathy, namel10.1177_01461672211052120.xmly, the ability to take the perspective of another and experience compatible emotional reactions (Batson, 2009), may be particularly important to either de-escalate the conflict (Bar-Tal, 2011) or assist in its resolution (Halperin, 2016; Halperin & Reifen Tagar, 2017). Surprisingly, how mothers influence their children’s empathy in intergroup conflicts has been largely overlooked. Focusing on the Israeli-Palestinian conflict, this article examines motivation among mothers to arouse empathy in their children, the potential source of this motivation (i.e., intergroup goals), and how it is linked to their children’s empathy toward the outgroup.

Intergroup Empathy

Empathy emerges early in infancy (Davidov et al., 2013) and becomes more complex as the child develops (Taylor et al., 2013). People vary in how much empathy they feel toward targets from different groups. Intergroup empathy refers to the experience of empathy toward someone else based on group affiliation (Vanman, 2016), which may sometimes result in experiencing less empathy toward outgroup members (e.g., Cikara & Van Bavel, 2014). Because empathy often (but not always; Sierksma et al., 2015; Taylor & Glen, 2020) facilitates prosocial behavior, decreased empathy toward outgroup members may render people less likely to help outgroup members (Hein et al., 2010) or more likely to harm them (Cikara & Fiske, 2011).

Intergroup empathy has been examined both in nonconflict (or low-conflict) intergroup relations (e.g., Montalan et al., 2012) and in conflictual relations (e.g., Hasson et al., 2018, 2019; Porat et al., 2016). Compared with nonconflict intergroup relations, the context of conflict (and especially intractable conflict) has unique characteristics, such as how prolonged the conflict is, how central it is in people’s lives, and how hard it is to resolve (Bar-Tal, 2007). Societies in conflict adapt to it emotionally (Bar-Tal et al., 2007), for example, by decreasing empathy toward the outgroup to justify the ingroup’s actions (Bar-Tal, 2011). For societies in conflict, experiences of empathy toward the outgroup could carry direct costs, both psychologically (e.g., the ingroup’s self-image) and pragmatically (e.g., compromises, humanitarian aid; Halperin, 2016). Therefore, although not tested empirically to date, it is probable that biases in intergroup empathy are greater in intense conflicts (compared with nonconflict intergroup relations) and that such biases become evident earlier in child development (Nasie et al., 2016). Similar to adults, children experience decreased empathy toward outgroup members (Levy et al., 2016; Masten et al., 2010). Empathy could result from the child’s characteristics (e.g., Christov-Moore et al., 2014), parental influences, such as genes (e.g., Knafo et al., 2008), and emotion socialization (e.g., Farrant et al., 2012). In the intergroup context, we suggest that there may be other, unique factors that shape how much empathy a child feels toward outgroup members. Although certain aspects of intergroup cognition are apparent in early infancy (e.g., Pascalis & de Schonen, 1994), throughout childhood, children continue to learn valuable social knowledge. Such knowledge develops from toddlerhood throughout childhood (Nesdale et al., 2008) and could include, for example, ingroup norms (Charters et al., 2013) or the societal psychological repertoire (Teichman & Bar-Tal, 2008).

There are several ways this social knowledge is passed on to children (Bar-Tal et al., 2017; Nesdale, 2012; Warshel, 2007). Parents are considered the main source of knowledge for children about the outgroup (Bar-Tal, 1996). Based on their own attitudes regarding the conflict, parents expose and mediate reality to their children, and thus shape their children’s worldviews (Bar-Tal et al., 2017). We can assume that mothers shape their children’s levels of intergroup empathy as well. However, this has not been tested empirically, and we do not know what the psychological mechanisms that underlie such intergenerational socialization are. Children may learn how to feel by observing their mothers’ reactions, that is, through modeling and mimicry. Although mothers’ reactions may be spontaneous at times, they may also act in a way that communicates how they want their children to feel. We suggest that what mothers want their children to feel plays a key role in such a socialization process.

Individual Differences in Maternal Motivation for Intergroup Empathy

Emotions typically arise spontaneously, whether we want them to or not. However, people may be motivated to experience certain emotions (for review, see Tamir, 2016). What people want to feel guides how they regulate their own and others’ emotions (Tamir, 2016; Tamir et al., 2020). People want to feel emotions that help them achieve their goals (Tamir, 2016). Such goals can be hedonic (e.g., maximizing immediate pleasure) or instrumental (e.g., promoting desired social outcomes, regardless of the immediate pleasure). This motivation, in turn, shapes the direction in which people regulate their emotions (Tamir, 2009).

Adults differ in the extent to which they find empathy desirable (e.g., Tamir et al., 2016). They may be motivated to decrease empathy toward someone who is suffering to feel better, or they may be motivated to increase empathy toward another to promote social bonding (Zaki, 2014). In the intergroup context, people may be motivated to experience emotions that are in line with their intergroup goals, as captured by their political ideology (Porat et al., 2016). For example, rightist (vs. leftist) Jewish-Israeli adults may want to feel less empathy toward Palestinians, and thus downregulate their empathy toward that group (Porat et al., 2016).

Political ideology encapsulates core intergroup attitudes, beliefs, and values (Jost et al., 2009). Although socioeconomic-based ideology (as in the United States and Europe) may be associated with conflict-related phenomena, for societies involved in intractable conflicts, people’s political ideology is determined by their position regarding conflict-related issues rather than socioeconomic ones (see Orian Harel et al., 2020). In Israel, leftists (the “peace camp”) support more conciliatory policies regarding the Israeli-Palestinian conflict, and rightists (the “national / nationalistic camp”) support more hardline aggressive policies (Arian & Shamir, 2008; Schori-Eyal et al., 2019).

Similar to their motivation to regulate their own emotions, people are also motivated to regulate the emotions of others for either hedonic or instrumental reasons (e.g., Netzer et al., 2015). With respect to the former, for example, people may wish to make others feel better (e.g., Brewer & Kramer, 1985; Halevy et al., 2008) or worse (e.g., Bar-Tal et al., 2007; Plant & Devine, 2003), depending on the type of relationship with the other (whether the other is a friend or a foe). With respect to the latter, for example, people may be motivated to make others experience emotions that promote what they perceive as instrumental behavior (Netzer et al., 2015). We expected mothers to be motivated to have their children experience less empathy toward the outgroup, in line with their conflict-related intergroup goals (their political ideology). In this context, we expected rightist (vs. leftist) mothers to be motivated to have their children feel less empathy toward outgroup members.

The Current Investigation

We tested our hypotheses in a series of four studies in the context of the Israeli-Palestinian conflict. Participants (in all studies) were Jewish-Israeli mothers and their children (in Studies 1 and 3), within three age groups: 6 to 8 years (Study 1), 6 to 12 years (Study 2), and 9 to 12 years (Studies 3–4). Throughout the studies, we chose to use the social category “Arabs” as representing the outgroup for Jewish-Israelis in the context of the Israeli-Palestinian conflict (Bar-Tal, 1996; Birnbaum et al., 2010) so that children would be familiar with this category (Slone et al., 2000). Consistent with previous literature on children’s social learning (e.g., Nguyen et al., 2021), we sampled mothers, but not fathers. This is because, even today, the emotional aspect of parenting is more commonly attributed to mothers (Cheung et al., 2018; Moon & Hoffman, 2008). Compared with fathers, mothers talk more frequently (Aznar & Tenenbaum, 2013) and elaborately (Fivush et al., 2000; Zaman & Fivush, 2013) about emotions with their children.

Our studies targeted several interrelated research questions. In Study 1, we tested whether children of rightist versus leftist mothers differed in experiencing empathy toward the outgroup and in their general empathy tendencies (RQ1). In Study 2, we examined whether rightist and leftist mothers differed in the degree to which they wanted their children to feel empathy toward the outgroup, as opposed to empathy toward people, in general (RQ2). In Study 3, we tested whether the relationship between the mother’s political ideology and the child’s empathy toward the specific outgroup was mediated by the mother’s motivation for her child’s empathy toward that group (RQ3). In Study 4, we tested how the increased salience of the conflict influenced the relationship between the mother’s ideology and her motivation for her child’s intergroup empathy (RQ4). In addition to examining how mothers conveyed their motivation to their children in real-life situations, Study 4 also examined whether mothers’ communication strategies were in line with their motivation for their child’s empathy toward a victim in a story they read (RQ5).

Previous literature suggests that gender is linked to empathy (e.g., Schwenck et al., 2014) either because of gender differences in the experience of empathy (Baron-Cohen et al., 2005) or because of different norms about the acceptability or expectation to report empathy (Lennon & Eisenberg, 1987). Therefore, throughout all four studies, we included children’s age and gender as covariates.

Study 1

The goal of Study 1 was to first test whether differences in experienced empathy toward outgroup members (Arabs) existed among children of leftist and rightist mothers. Second, to test whether such differences reflected variance in maternal intergroup goals and not inherent differences in general empathy, we examined whether these children differed in their empathy expressions toward people in general (RQ1). We predicted that children would differ in how much intergroup empathy they experienced, as a function of their mothers’ political ideology, but not in their general empathy tendencies (H1).

Method

All relevant data files and syntax can be found at https://osf.io/mnrf2/.

Participants

One hundred three Jewish-Israeli dyads of mothers and their children (aged 6–8 years) were recruited using a “snowball” method. Participants completed a pen-and-paper questionnaire in their homes during a home visit by the research team. This study was conducted with no a priori power analysis to determine the sample size. A sensitivity analysis using G*Power 3.1 (Faul et al., 2009) indicated sensitivity to detect an effect size of 0.31 (power of 0.80, alpha of 0.05). Thirty-eight percent of the mothers self-identified as rightists, 18.4% as centrists, and 42.6% as leftists. The mean age of the mothers was 39.20 (SD = 4.75). The mean age of the children was 7.09 (SD = 0.80), and 52% were girls.

Procedure

After giving informed consent both for themselves and for their children, mothers completed the questionnaires. At the same time, an experimenter sat with the child in a separate room (to prevent any influence of the mother’s presence). The experimenter read the questions to the child and recorded the child’s responses.

Measures

The following measures were part of a bigger study (which also included pilot testing several measures), so only relevant measures to this project are reported.

Empathy toward Arabs

Children rated the degree of empathy they felt toward Arabs on a scale of 1 (to a very low extent) to 5 (to a very high extent). The children were asked, “To what extent would you feel sad if Arabs were sad?”¹

General empathic tendencies

To examine the child’s empathy abilities in general and based on past use of empathy-inducing pictures in measuring empathy in children (Decety et al., 2018), we chose nine pictures of children experiencing different emotions and presented them to the children. After recognizing the emotion, we asked the children how seeing this picture made them feel. If they reported feeling the same as the child in the picture, their answer was coded as a match (1), and if not, then as a mismatch (0). Then, all of their answers were coded to one measure (α = .89) so that each child had a general empathy tendency score.

Political ideology

Mothers were asked to report their political ideology on a scale of 1 (extreme right) to 7 (extreme left). Because we were interested in the difference between rightwing and leftwing ideologies, and perceived centrists as a distinct ideological group (see Hasson et al., 2018), we recoded this measure into three levels: right (including extreme right, right, and moderate right), center, and left (including moderate left, left, and extreme left).²

Sociodemographic measures

The mothers also completed a sociodemographic questionnaire, listing their age, religion, child’s age, and gender.

Results

Do children differ in their empathy toward the outgroup in accordance with their mothers’ political ideology?

First, we examined the relations between mothers’ political ideology and children’s empathy toward Arabs. An analysis of covariance (ANCOVA), F(2, 96) = 3.55, p = .03, partial η² = .07, revealed that, as predicted, children of leftist mothers reported significantly higher levels of empathy toward Arabs (M = 2.96, SD = 1.56) compared with children of rightist mothers (M = 2.26, SD = 1.27; p < .01; 95% confidence interval [CI] = [−1.40, −0.16]; d = 0.49; see Figure 1).

Figure 1.

Children’s empathy levels toward Arabs as a function of their mothers’ political ideology, Study 1.

Note. Brown points represent the means and their 95% confidence intervals as estimated by the model. Graphs generated by the ggstatsplot package in R (Patil, 2021).

Do children differ in their general empathy tendencies in accordance with their mothers’ political ideology?

In line with our hypothesis, when we ran the same ANCOVA with the child’s general empathic tendencies as the dependent variable, we did not find a significant effect, F(2, 97) = 1.11, p = .33, partial η² = .02; see Figure 2.

Figure 2.

Children’s general empathic tendencies as a function of their mothers’ political ideology, Study 1.

Note. Brown points represent the means and their 95% confidence intervals as estimated by the model. Graphs generated by the ggstatsplot package in R (Patil, 2021).

Discussion

Study 1 provided initial support for our hypothesis that children of rightist and leftist mothers differed in their empathy toward Arabs, such that children of leftist mothers reported that they felt more empathy compared with children of rightists. As predicted, we did not find an indication for the relations between the mothers’ political ideology and the children’s general empathy tendencies (H1). This suggests that differences in children’s intergroup empathy were contingent on their mothers’ intergroup goals rather than on inherent differences in empathy between rightists and leftists. Regardless of mothers’ ideology, all children reported that they felt low to medium levels of empathy toward Arabs. This is in line with the low levels of empathy people who live in areas of intractable conflicts tend to feel toward the outgroup (Bar-Tal, 2011).

This study’s main limitation was that linking mothers’ political ideology with children’s intergroup empathy did not necessarily indicate the existence of a motivational socialization process, as we have suggested. In addition, in this study, we encountered some children whose level of knowledge about the conflict or the outgroup was low: During the experiment, some of the children asked the experimenter questions about the social category “Arabs” or about the nature of the relationship between the ingroup or the outgroup. Because we wanted to present a direct link between the mother and the child, as well as exploring the mechanism underlying such a link, in the following studies, we chose to sample ages in which children and their parents can have a more meaningful and complex dialogue at home (9–12). Therefore, we made the age shift gradually, by adding the older age group to the younger one in Study 2 (thus sampling both 6- to 8-year-olds and 9- to 12-year-olds) and then sampling the older age group (9–12) in Studies 3 and 4.

As children mature, parents have more complex conversations with them, contributing to their knowledge (e.g., on ethnic socialization; Aboud & Amato, 2001). In Israel, parents of children aged 6 to 8 years, like those who participated in Study 1, vary in the degree to which they talk about the Israeli-Palestinian conflict with their children. Rightist parents have found it easier to talk to their children about the conflict at these ages, presenting them with a more categorical description of reality, which is easier to convey. Leftist parents, however, have tended to talk to their children less about the conflict at these ages because their messages are more complex and multidimensional, and therefore harder to convey (Bar-Tal, Reshef et al., 2012). Therefore, we gradually increased the target age group so that by Study 4, which focused on how mothers verbally communicated messages to their children, we could focus on age groups in which such complex discussions were more likely to occur.

With regard to this age shift, children’s empathy develops and changes throughout childhood. However, a significant aspect of empathy occurs even before the age of 6 (Eisenberg & Fabes, 1998) and remains relatively stable throughout elementary school (between the ages of 6–12; Côté et al., 2002). Moreover, the cognitive and the affective components of empathy develop separately (Knafo et al., 2008; Roth-Hanania et al., 2011). Although cognitive empathy continues to develop throughout childhood, the affective aspect of empathy (which we measured in children in this investigation) exists from infancy (Eisenberg et al., 2005; Tousignant et al., 2017; Van der Graaff et al., 2014). Because children’s affective empathy abilities remain relatively stable throughout elementary school, we started with children aged 6. Based on the existing literature, the affective aspect of empathy was unlikely to differ dramatically between the different age groups targeted in our studies.

Study 2

Study 2 had two main goals. First, we tested whether rightist and leftist mothers differed in the degree to which they wanted their children to feel empathy toward a specific outgroup (RQ2). Second, in contrast to Study 1, in which we measured general empathic tendencies and empathy toward the outgroup (using two different measures), in this study we directly compared empathy toward outgroup members to empathy toward people in general using equivalent measures. We predicted that leftist mothers would want their children to feel more empathy toward the outgroup compared with rightist mothers, whereas there would be no differences in their motivation for the child’s empathy toward people in general (H2). Furthermore, additional groups were added to make the focus on Arabs as the target outgroup less obvious.

Method

Participants

An a priori power analysis was conducted using G*Power 3.1 (Faul et al., 2009) to obtain a small effect size (Cohen’s f = 0.10) of a within-between interaction in a repeated-measures ANCOVA, and an alpha of .05, and an expected correlation among repeated measures of 0.5; a total sample of 246 participants was required to achieve a power of .80. Two hundred fifty Jewish-Israeli mothers of children aged 6 to 12 years were recruited through a survey company. (Three additional participants were removed for failing reading comprehension questions.) In terms of their political ideology, 42.2% of the participants identified as rightists, 23.1% as centrists, and 34.7% as leftists, with a mean age of 40.30 (SD = 6.05). The mothers were asked to report the gender and age of the child with whom they were responding—46.6% were girls; the mean age of the children was 8.93 (SD = 2.06).

Procedure and measures

After giving informed consent, mothers were presented with a definition of empathy to ensure a balanced and equivalent understanding of the construct (see full materials: https://osf.io/mnrf2/) and were asked three questions to make sure they had read and understood the definition.

Motivation for child’s empathy

To disguise our selected target groups and reduce potential demand characteristics, mothers were presented with 17 different groups from the Israeli society, people in general (humans) and animals, in counterbalanced random order.³ Mothers were asked to rate the extent of empathy they wished their child to feel toward each group on a scale of 0 to 100 (0 = no empathy at all, 100 = empathy to a great extent). These groups were used for this reason alone, and therefore, most of them are not part of the analysis. To make sure the mothers’ reports were as honest as possible, two steps were taken: (a) mothers’ answers were visually based, meaning the scale did not indicate the exact number they chose from 0 to 100, and (b) each group was presented to the mother on a different screen, thus not allowing her to visually compare her answers.

Political ideology

We used the same scale as in Study 1.

Sociodemographic measures

We used the same measures as in Study 1.

Results

Is mothers’ political ideology correlated with their motivation for child’s empathy?

See Table 1 for means and standard deviations. Using repeated-measures ANCOVA, we examined a within-between subjects interaction: the mothers’ political ideology (between subjects) × the target’s group identity (within subjects) predicting mothers’ motivation for their child’s empathy (see Figure 3 and Table 2). Results showed that rightist mothers wanted their children to feel significantly less empathy toward Arabs compared with leftist mothers (p < .001; 95% CI = [−48.56, −32.29]; d = 0.51). This effect varied as a function of empathy target group, as there were no significant differences in the levels of empathy toward people in general that mothers wanted their children to feel, across all political ideologies (all ps > .39; ds < 0.13). We found a main effect for the target’s group identity, as well as for political ideology.

Table 1.

Descriptive Statistics, Mothers’ Motivation for Child’s Empathy, Study 2.

Target	M	SD
Rightist mothers
People in general	86.04	18.74
Arabs	36.41	31.53
Centrist mothers
People in general	85.97	18.49
Arabs	52.62	30.86
Leftist mothers
People in general	88.10	14.43
Arabs	76.73	20.95

Figure 3.

Mothers’ motivation for their children’s empathy as a function of political ideology and the target’s group membership (humans/Arabs).

Note. Error bars reflect 95% confidence intervals. Graph generated by the interactions package in R (Long, 2019).

Table 2.

Repeated-Measures ANCOVA, Study 2.

Effect	F	df	p	Partial η2
Target	13.71	1	.00	.05
Political Ideology	30.76	2	.00	.20
Target × Political Ideology	41.72	2	.00	.25
Target × Child’s Gender	0.02	1	.90	.00
Target × Child’s Age	0.51	1	.48	.00
Error		245

Note. ANCOVA = analysis of covariance.

Additional analysis

Though not originally our intention, the measures provided the opportunity to test another question: Are rightists always motivated to have their children feel less empathy toward any outgroup compared with leftists, or perhaps we found this effect because Arabs, specifically, are an ideologically sensitive outgroup? If the latter is true, in another intergroup context in which both groups hold similar intergroup goals, we can expect a similar motivational pattern for both rightists and leftists.

Existing literature presents an ongoing debate regarding whether decreased empathy toward the outgroup, in which rightists feel less empathy toward the outgroup compared with leftists, is inherent or target-dependent (Hasson et al., 2018). Following this debate and the results of Study 1, we wanted to examine another interaction. As analyzed in the work of Hasson et al. (2018), we tested the same repeated-measures ANCOVA; however, this time our target groups were rightists and leftists. Rightist and leftist participants were analyzed as each other’s outgroup: We created two new scales: an ingroup scale and an outgroup scale. The ingroup scale captured mothers’ motivation for their child’s empathy toward their political ingroup (to what extent rightist mothers wanted their child to feel empathy toward rightists, and vice versa for leftist mothers). The outgroup scale measured mothers’ motivation for child’s empathy toward their political outgroup (to what extent rightist mothers wanted their children to feel empathy toward leftists, and vice versa for leftist mothers). We found a main effect for political ideology, F(1, 187) = 4.12, p = .04, partial η² = .02, but not for the target’s group identity, F(1,187) = 2.32, p = .13, partial η² = .01. We did not find a significant interaction, F(1, 187) = 3.55, p = .06, partial η² = .20, suggesting that when posed as each other’s ingroup and outgroup, rightist and leftist mothers did not differ in their motivation toward the political outgroup to a great extent.

Discussion

Study 2 showed that leftist and rightist mothers differed in their motivation for their child’s empathy toward the outgroup, but not toward people in general (H2). This suggests that in intractable conflicts, mothers who varied in their political ideology wanted their children to feel different levels of empathy toward outgroup members. Our additional analyses showed that this effect was context-dependent. Rightist and leftist mothers did not inherently prefer more or less empathy in their children. Instead, they differed with respect to how much empathy they wanted their children to feel toward outgroup members in the context of the conflict. The main limitation of Study 2 was that it did not link mothers’ motivation to their children’s actual experiences of empathy, so this was the goal of Study 3.

Study 3

The goal of Study 3 was to test whether the relationship between the mother’s political ideology and the child’s empathy was mediated by the mother’s motivation for her child’s empathy toward Arabs (RQ3). We predicted that leftist mothers would want their child to feel more empathy toward Arabs compared with rightist mothers and, in turn, that children of leftists would experience more empathy toward Arabs compared with children of rightists (H3).

Method

Participants

An a priori power analysis was conducted using G*Power 3.1 (Faul et al., 2009) to test an ANCOVA, with a medium effect size (0.30), based on the results of children’s empathy in Study 1, and an alpha of .05. Results showed that a total sample of 111 participants was required to achieve a power of .80. One hundred forty-five Jewish-Israeli mother–child dyads participated in this study. Another five dyads were excluded from the analyses (four because the mothers put the phone on speaker when the child was interviewed and might have influenced the child’s answers; one because the child self-identified as an outgroup member). In terms of political ideology, 43.4% of the mothers self-identified as rightists, 28.3% as centrists, and 28.3% as leftists. Of the children who participated in the study, 52.7% were girls; the mean age of the children was 10.43 (SD = 1.08). Participants were recruited either by a snowball procedure or by a survey company.

Procedure

Mothers completed an online questionnaire, while children were interviewed over the phone. After giving their informed consent, mothers read the definition of empathy used in Study 2 and answered reading comprehension questions. Then, they reported their motivation for their child’s intergroup empathy, their political ideology, and a sociodemographic questionnaire. Children reported their intergroup empathy experience.

Measures.⁴

Mothers’ motivation for the child’s empathy levels

As in Study 2.

Political ideology

As in Studies 1 and 2.

Sociodemographics

As in Studies 1 and 2.

Child’s self-reported empathy

Here too, we embedded our selected target groups for the children among six other social groups (selected from the 17 presented to mothers). Children were asked, “To what extent would you feel sad if X were sad?” on a scale of 1 (not at all) to 5 (very sad), similar to the question children were asked in Study 1. Given that children at these ages are not fully aware of all the social categories we presented to their mothers, we chose the six most salient social categories for children at this age in Israel (people in general, Arabs, Ethiopians, secular, religious, and disabled people).

Results

Main analysis: Is the relationship between mothers’ political ideology and children’s empathy toward the outgroup mediated by mothers’ motivation for their child’s empathy?

Means, standard deviations, and correlations are presented in Table 3. The relationship between the mother’s political ideology and the child’s levels of empathy toward Arabs was mediated by the mother’s motivation for her child’s empathy levels. Because we recoded political ideology as an order variable, we created two dummy variables for political ideology: D1, reflecting the comparison between rightists and centrists, and D2, reflecting the comparison between rightists and leftists. For this analysis, we used PROCESS Model 4 (Hayes, 2018) with 5,000 iterations. For D2 (rightists vs. leftists), we found a significant indirect effect of the mothers’ political ideology on their child’s empathy toward Arabs through the mothers’ motivation for the child’s empathy toward Arabs (effect = 0.33, SE = 0.15, 95% CI = [0.04, 0.63]; see Figure 4).

Table 3.

Means, Standard Deviations, and Correlations, Study 3.

Control variables		M	SD	1	2	3	4	5	6
Nonea	1. D1 (centrists = 1, rightists = 0)	0.28	0.45	1.00
	2. D2 (leftists = 1, rightists = 0)	0.28	0.45	–.40**	1.00
	3. Mother’s motivation for her child’s empathy toward Arabs	50.45	32.30	–.05	.49*	1.00
	4. Child’s empathy toward Arabs	2.32	1.22	.05	.14	.22**	1.00
	5. Child’s age	10.44	1.08	.05	–.04	–.16	.14	1.00
	6. Child’s gender	1.53	0.50	–.06	–.002	–.10	–.05	–.06	1.00
Child’s age and gender	1. D1 (centrists = 1, rightists = 0)			1.00
	2. D2 (leftists = 1, rightists = 0)			–.40**	1.00
	3. Mother’s motivation for her child’s empathy toward Arabs			–.04	.49**	1.00
	4. Child’s empathy toward Arabs			.04	.14	.25**	1.00

^aCells contain zero-order (Pearson) correlations.

^*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

Figure 4.

The relations between mother’s political ideology and her child’s empathy toward Arabs were mediated by her motivation for the child’s empathy toward them, Study 3.

Discussion

In Study 3, we found that the relationship between mothers’ political ideology and their child’s self-reported empathy levels toward the outgroup (Arabs) was mediated by how mothers wanted their child to feel toward the outgroup. The results of this study indicate that mothers’ motivation for their children’s empathy toward the outgroup informs the socialization of intergroup empathy in their children. So far, in Studies 1 to 3, we found differences in child intergroup empathy contingent on their mother’s ideology (H1; Studies 1 and 3), a link between the mother’s ideology and motivation for the child’s empathy (H2; Studies 2 and 3), and the mediating role of the mother’s motivation for the child’s empathy in the relations between the mother’s ideology and the child’s intergroup empathy experience (H3; Study 3). Next, we wanted to examine a boundary condition for this model in the form of the relevance of political ideology within the intergroup context.

Study 4

Intergroup goals are goals that a person adopts as a group member that concern the group’s future, identity, and values. These goals can take on different forms, and in contexts of conflict, they are often related to desired relationships with the outgroup. Certain aspects of political ideology could capture intergroup goals (Jost et al., 2009). In contexts of conflict, political ideology (as a reflection of intergroup goals) has been linked to motivation in emotion regulation (Hasson et al., 2018; Porat et al., 2016). Previous research has further suggested that the link between political ideology and emotion regulation might be particularly relevant in conflictual contexts. For instance, rightists and leftists differed in their emotional intensity when the context was conflict-related, but did not differ when it was non-conflict-related (Pliskin et al., 2018). Thus, we expected that the link between political ideology and motivation for child’s empathy toward outgroup members should be more salient in the context of the conflict (e.g., when harm to the outgroup was caused by the ingroup in the context of conflict; H4). Therefore, in Study 4, we tested whether the relationship between mothers’ ideology and their motivation for their child’s empathy was moderated by the salience of the conflict (RQ4). The second goal of Study 4 was to test the motivational socialization process in a way that was closer to the real-life process. In Studies 1 to 3, we examined mothers’ motivation for their children’s empathy toward Arabs as a group. Outgroup suffering, which is mainly introduced to the ingroup through the media, often takes the form of an individual victim or victims and not the group as a whole. This is noteworthy because people tend to experience more empathy toward an individual, compared with a group (Kogut & Ritov, 2005). Such media exposure offers mothers opportunities to socialize empathy toward the other. Some might suggest that in these cases, in real life, what matters is the parent’s reaction itself (Bandura, 1977) because one of the ways emotions such as empathy or prosocial behavior is socialized is through observation and imitation (Grusec et al., 2007; Hastings et al., 2007). For this reason, we wanted to examine whether mothers differed in their motivation toward individuals from the outgroup and how they communicated this motivation to their children (RQ5). We hypothesized that mothers would use emotion regulation tactics that were consistent with their motivation for their child’s empathy as a function of their ideology and the group identity of the target (H5). In intractable conflicts, mothers tend to talk to their children in a more explicit and emotional way, compared with a nonconflict situation (Aboud & Amato, 2001). Therefore, we could expect that mothers’ motivation for their child’s empathy would be more present in her communication to the child when the target was an outgroup member.

In Study 4, with the use of a single victim, we did not want the “general” group to remain unidentifiable so that we could control participants’ perception of the group affiliation. Therefore, in this study, we also examined mothers’ motivation for their child’s empathy toward the ingroup, in addition to “people in general” and the outgroup. Past research suggests that the empathic response toward the ingroup should be similar to participants’ ratings of unaffiliated targets (Cikara et al., 2014). Therefore, we also included the ingroup as a target, in an additional experimental condition.

Method

Participants

An a priori power analysis using G*Power 3.1 (Faul et al., 2009) showed that a total sample of 270 participants was required to achieve a small-medium effect size (0.19, based on the average effect size in the past two studies), to test an ANCOVA (fixed effects, special, main effect, and interactions), with an alpha of .05 and a power of .80. Two hundred sixty-three Jewish-Israeli mothers (to children aged 9–12) were recruited by a survey company. Another 61 participants were removed for failing reading comprehension questions. In terms of political ideology, 45.2% of the mothers self-identified as rightists, 28.7% as centrists, and 25.7% as leftists. The mothers reported the gender and age of their child—45.2% were girls; the mean age of the children was 10.32 (SD = 1.11).

Procedure

After giving their informed consent, mothers were presented with our definition of empathy and answered the reading comprehension questions about it (see full materials: https://osf.io/mnrf2/). Then, they reported their motivation for their child’s empathy toward different social groups, after which they all read a scenario (supposedly from a children’s online newspaper) about a grandmother whose house had burnt down. The story had four versions, one for each condition: In the first three conditions (“general,” “ingroup,” and “outgroup”), the victim’s house had burnt down due to a fault in the main heating system. The only difference between these three conditions was the group identity of the victim (in the general scenario, the victim was Greek; in the ingroup scenario, she was Jewish-Israeli; and in the outgroup scenario, she was Arab). In the fourth condition, the “conflict condition,” the grandmother was the same Arab grandmother as in the outgroup condition, but now participants were told that the fire had broken out due to an IDF (the Israeli Army) military action nearby (instead of a fault in the heating system). The general vignette was pilot tested for the extent to which the story induced empathy (see supplementary materials: https://osf.io/mnrf2/). Participants were randomly assigned to one of these four conditions. After reading the scenario, mothers were asked three reading comprehension questions (see full materials: https://osf.io/mnrf2/), reported the extent to which they would want their child to feel empathy toward the grandmother, as well as reporting in their own words how they would communicate this story to their child, followed by sociodemographic questions.

Measures

Maternal motivation for the child’s empathy levels

As in Studies 2 and 3.

Maternal motivation for the child’s empathy levels toward the grandmother

Mothers were asked, “To what extent do you want your child to feel empathy toward Athena/Rivka/Jamila (the names of the grandmothers described in the vignette) had she been exposed to the story?” (scale of 0–100).

Mothers’ use of empathy regulation tactics

Mothers’ open answers were coded for empathy regulation tactics to try to examine how they expressed their motivation for their child’s empathy in the way they communicated the story to the child. We wanted to examine whether the way the mother wanted her child to feel was clear from the way she expressed herself. Empathy can be upregulated or downregulated by tactics of situation selection, attention modulation, and appraisal (Zaki, 2014), as well as by viewing the other as different and far away (Batson & Ahmad, 2009). We built a coding system (see full materials: https://osf.io/mnrf2/) based on these tactics, for example: “To what extent does the mother focus on the grandmother’s suffering and pain?” and “To what extent does the mother describe the grandmother as different than herself and her child?” Texts were coded by two BA psychology students (interrater reliability of .64≤ α ≤.99 of 16 example texts) who were blind to mothers’ motivation and political ideology. The questions were then transformed to z-scores, the downregulation items were reverse coded so that for all items, higher values indicated attempts to upregulate the child’s empathy, and then all of these items were computed together to form an empathy regulation scale (α = .76). As a preliminary test, coders were also generally asked, “To what extent does this mother want her child to feel empathy toward the grandmother?” (on a scale of 1–7). This item was correlated with the mothers’ reported motivation (r = .13, p = .04), suggesting that, aside from their use of specific tactics, their motivation was apparent in the way they communicated.

Political ideology

As in Studies 1–3.

Sociodemographics

As in Studies 1–3.

Results

Preliminary analysis

Replication: Is mothers’ political ideology correlated with their motivation for their child’s empathy? (RQ2)

Using a repeated-measures ANCOVA, we examined the interaction between the mothers’ political ideology and the target (people in general vs. Arabs) on the mothers’ motivation for their child’s empathy (see Tables 4 and 5, and Figure 5). Results of the interaction analysis showed that leftist mothers wanted their children to feel significantly more empathy toward Arabs compared with rightist mothers (p < .001; d = 1.42). This effect varied as a function of the empathy-target group, as there were no significant differences in the levels of empathy that rightist (vs. leftist) mothers wanted their children to feel toward people in general (p = .45; d = 0.11; for analysis of centrist mothers, see Supplementary Materials: https://osf.io/mnrf2/). We did not find a significant main effect for the target, but we did find a main effect for political ideology, driven by differences in the outgroup condition.

Table 4.

Descriptive Statistics, Mothers’ Motivation for Child’s Empathy, Study 4.

Target	M	SD
Rightist mothers
People in general	90.42	14.02
Arabs	36.24	31.88
Centrist mothers
People in general	89.44	19.19
Arabs	56.47	31.41
Leftist mothers
People in general	88.85	14.49
Arabs	76.33	24.13

Table 5.

Repeated-Measures ANCOVA, Study 4.

Effect	F	df	p	Partial η2
Target	0.06	1	.81	.00
Political Ideology	23.46	2	.00	.16
Target × Political Ideology	43.28	2	.00	.25
Target × Child’s Gender	1.01	1	.32	.00
Target × Child’s Age	3.95	1	.05	.02
Error		254

Note. ANCOVA = Analysis of covariance.

Figure 5.

Mothers’ motivation for their child’s empathy as a function of the mother’s political ideology and the target’s group membership (people in general/Arabs), Study 4.

Note. Error bars reflect 95% confidence intervals. Graph generated by the interactions package in R (Long, 2019).

Main analysis: Does the conflict’s salience serve as a conditional boundary in the motivation for empathy toward the outgroup? (RQ4)

In an ANCOVA, we examined the interaction between the mothers’ political ideology × the condition (meaning the target’s group identity and cause of the fire; between subjects) on mothers’ motivation for their child’s empathy (see Tables 6 and 7, and Figure 6). Results showed that leftist mothers wanted their children to feel significantly more empathy toward the Arab grandmother in the conflict condition compared with rightist mothers (p < .001; 95% CI = [−36.93, −11.51]; d = 0.76). A similar pattern was found in the outgroup condition when leftist mothers wanted their children to feel significantly more empathy toward the Arab grandmother compared with rightist mothers (p = .04; 95% CI = [−25.19, −.32]; d = 0.91). This effect varied as a function of empathy target group, as rightist and leftist mothers did not differ in their motivation for their child’s empathy either in the ingroup condition (p = .78; 95% CI = [−10.61, 14.19]; d = 0.08) or in the general condition (p = .43; 95% CI = [−7.42, 17.24]; d = 0.32). We found a significant main effect for the condition and a trend for political ideology. Pairwise comparisons for the main effect of the condition showed that only the conflict-related condition was found to be significantly different than the general condition (M_Difference = 13.21, SE = 3.74, p < .001; 95% CI = [−20.57, −5.85]), the ingroup condition (M_Difference = 13.60, SE = 3.69, p < .001; 95% CI = [−20.87, −6.34]), and the outgroup condition (M_Difference = 10.66, SE = 3.72, p = .01; 95% CI = [−17.98, −3.33]). The outgroup condition was not found to be significantly different from the general condition nor the ingroup condition (ps ≥ .43).

Table 6.

Descriptive Statistics, Mothers’ Motivation for Child’s Empathy, Study 4.

Target	M	SD
Rightist mothers
General (Greek)	89.64	13.57
Ingroup (Jewish-Israeli)	89.54	21.40
Outgroup (Arab)—no conflict	81.83	22.09
Outgroup (Arab)—conflict	61.70	34.04
Centrist mothers
General (Greek)	91.40	13.52
Ingroup (Jewish-Israeli)	88.67	14.79
Outgroup (Arab)—no conflict	81.59	20.34
Outgroup (Arab)—conflict	77.44	23.95
Leftist mothers
General (Greek)	84.38	21.24
Ingroup (Jewish-Israeli)	88.00	18.46
Outgroup (Arab)—no conflict	94.20	6.21
Outgroup (Arab)—conflict	86.11	16.41

Table 7.

Univariate Analysis of Variance, Study 4.

Effect	F	df	P	Partial η2
Condition (Target)	6.03	3	.00	.07
Political ideology	2.97	2	.05	.02
Child’s age	2.25	1	.14	.01
Child’s gender	0.01	1	.94	.00
Condition × Political Ideology	2.60	6	.02	.06
Error		246

Figure 6.

Mothers’ motivation for their child’s empathy as a function of the mother’s political ideology and the target’s group membership (people in general/Jewish-Israeli/Arab/Arab—conflict-related), Study 4.

Note. Error bars reflect 95% confidence intervals. Graph generated by the interactions package in R (Long, 2019).

We then tested whether rightist and leftist mothers differed to a greater extent in the conflict condition compared with the outgroup condition, as predicted. We ran the analysis again, without the “general” or the “ingroup” conditions, to test the interaction only between the “outgroup” and the “conflict” conditions. An ANCOVA analysis revealed a nonsignificant interaction between the two conditions, F(2, 126) = 1.46, p = .24, partial η² = .02. Both the mothers’ political ideology, F(2, 126) = 6.87, p = .001, partial η² = 1.00, and the conditions, F(1, 126) = 5.79, p = .02, partial η² = .04, had significant main effects.

Despite the “outgroup” and the “conflict” conditions not yielding different effects, when looking at the first interaction (the one including all four conditions) from the alternative perspective, we found that leftist mothers showed no difference in the extent to which they wanted their child to feel empathy across all four conditions (all ps ≥ .17; all ds ≤ 0.65). In contrast, rightist mothers reported wanting their child to feel empathy to the same extent when the victim was either general, ingroup member, and even an outgroup member when she was not suffering due to the conflict (all ps ≥ .13; all ds ≤ 0.43). The salience of the conflict (in the form of the ingroup harming the outgroup member) was found to be a boundary condition, as the only condition in which rightist mothers reported wanting their child to feel less empathy toward the victim was in the conflict-related condition (ps compared with all other conditions <.001; ds ≤ 1.08), matching their context contingent intergroup goals (political ideology).⁵

How do mothers express their motivation? (RQ5)

Correlations are presented in Table 8. The more mothers wanted their children to feel empathy toward the victim (regardless of the condition), the more they used tactics aimed at increasing empathy when they communicated the story in their answer (r = .25, p < .001).

Table 8.

Correlations, Means, and Standard Deviations, Study 4.

Control variables		M	SD	1a	2	3	4	5
Noneb	1. Political ideologya	1.80	.82	1.00
	2. Motivation for child’s empathy toward the victim	84.02	21.67	.77	1.00
	3. Mothers’ use of empathy upregulation tactics	–.06	.52	.77	.24**	1.00
	4. Child’s age	10.32	1.11	–.03	–.09	.02	1.00
	5. Child’s gender	1.45	.50	–.09	.02	–.03	.05	1.00
Child’s age and gender	1. Political ideologya			1.00
	2. Motivation for child’s empathy toward the victim			.07	1.00
	3. Mothers’ use of empathy upregulation tactics			.02	.25**	1.00

^aCorrelations of political ideology contain Spearman’s correlations.

^bCells contain zero-order (Pearson) correlations (except for “political ideology”).

^*Correlation is significant at 0.05 level. **Correlation is significant at 0.01 level.

Moderated mediation model

We used PROCESS (Hayes, 2018) Model 8 with 5,000 iterations to examine whether the condition (the target’s group membership) moderated the indirect effect of mothers’ political ideology and their use of empathy upregulation tactics in their communication of the story to their child, via their motivation for their child’s empathy toward the target (see Figure 7). We created two dummy variables for mothers’ political ideology (D1 = rightists vs. centrists; D2 = rightists vs. leftists) and three more for the condition (Mod1 = general vs. ingroup; Mod2 = general vs. outgroup, Mod3 = general vs. conflict). For D2 (rightists vs. leftists), we found that the indirect effect of political ideology on mothers’ use of upregulating tactics via their motivation for their child’s empathy was significant when the target was an outgroup member who was harmed in the context of the conflict or not in the context of the conflict. Looking at the model, the interaction between political ideology and the condition (in the conflict-related conflict; D2 × Mod3) was significantly associated with mothers’ motivation for their child’s empathy toward the victim (Effect = 29.13, SE = 8.99, t = 3.24, p = .001), which was significantly correlated with mothers’ use of upregulation tactics (Effect = .004, SE = .002, t = 2.74, p = .01). In the nonconflict-related condition, although the indirect effect was significant, the interaction between political ideology and the condition (D2 × Mod2) was not significantly correlated with mothers’ motivation for their child’s empathy (Effect = 17.66, SE = 8.87, t = 1.99, p = .05). We did not find a significant indirect effect in the general condition in the ingroup condition (see Table 9). In addition, the index of moderation mediation was significant for both the outgroup (Index = .07, SE = .04, 95% CI [0.01, 0.17]) and the conflict conditions (Index = .12, SE = .06, 95% CI [0.03, 0.28]; for a regression table of the full analysis, see supplementary materials: https://osf.io/mnrf2/.

Figure 7.

Mother’s tactics for upregulating their child’s empathy as a function of political ideology and the mother’s motivation for the child’s empathy, Study 4.

Table 9.

Indirect Effects of Mothers’ Political Ideology (D1; Rightists vs. Leftists) via Their Motivation for Child’s Empathy, by Condition.

Condition	Effect	SE	LLCI (95%)	ULCI (95%)
General	–.02	.03	–0.08	0.03
Ingroup	–.01	.03	–0.07	0.05
Outgroup	.05	.03	0.01	0.11
Conflict	.10	.05	0.02	0.23

Note. LLCI = lower limit confidence interval; ULCI = upper limit confidence interval.

Discussion

In Study 4, we found that, consistent with findings in Studies 2 and 3, leftist and rightist mothers differed in their motivation for their child’s empathy only when the victim was Arab, namely, when the target of empathy was related to ideology-dependent intergroup goals (H4). We found that leftist mothers wanted their children to experience the same level of empathy across all four conditions. Furthermore, although rightist mothers wanted their children to feel empathy to the same extent toward all nonconflict-related victims (even when the victim was an outgroup member), they wanted their child to feel less empathy when the conflict was salient due to the ingroup’s responsibility for the victim’s suffering (the conflict-related condition). Rightist mothers’ decreased motivation for their child’s empathy toward the victim in the conflict-related condition may have resulted from their beliefs about the ingroup’s positive self-esteem or the justness of the ingroup goals, for example (Bar-Tal, Shravit et al., 2012). These results suggest that intergroup goals (as measured by political ideology) are associated with motivation for empathy only in the context in which they are most relevant—toward the outgroup. Consistent with previous work (Gaesser et al., 2020), these intergroup goals did not yield a different motivation toward the ingroup member compared with a “general” group affiliation that was irrelevant to the context of these intergroup goals.

In addition, we asked mothers to write how they would tell their child the story had the child asked them about it in two to three sentences. We found that mothers’ political ideology was correlated with their motivation for their child’s empathy, which in turn was associated with their use of regulation tactics in their communication of the story to their child (H5). The mediation model was moderated by the condition such that it was significant only when the victim was an outgroup member (both in the presence of the conflict and in its absence). The fact that the moderated mediation showed similar results for outgroup victims, whether the conflict was salient or not, might suggest that by mentioning the outgroup, the conflict’s salience increases in the participants’ perspective. Thus, the external salience we created in the “conflict” condition might not be required to find these effects. Future work should examine this using more sensitive tools that reflect mother–child interactions in a more natural way. This future examination should also include children to examine how mothers’ communication of such stories impacts their empathy toward outgroup members.

General Discussion

Across four studies, we explored the relationship between mothers’ political ideology (reflecting their intergroup goals), their motivation for their child’s empathy toward the outgroup (compared with empathy in general), and their children’s empathy toward the outgroup (and empathy in general). We first found that children differed in their empathy toward the outgroup according to their mother’s political ideology, but not in their general empathy (H1). Second, we found that mothers’ political ideology was associated with how much empathy they wanted their children to feel toward the outgroup but not associated with how much empathy they wanted their children to feel in general (H2). Third, we found that the relationship between mothers’ political ideology and children’s empathy toward the outgroup was mediated by mothers’ motivation for their child’s empathy toward the outgroup (H3). Fourth, we found that rightist and leftist mothers differed in their motivation for their child’s empathy both in response to the suffering of an outgroup member which was unrelated to the conflict and the suffering of an outgroup member caused in the context of the conflict. At the same time, we found that rightist mothers decreased their motivation for their child’s empathy toward the victim when the suffering was caused in the context of the conflict versus when it was unrelated to it (H4). Finally, we found that the motivation for the child’s empathy was associated with mothers’ intention to use emotion regulation tactics in communicating a story about an outgroup member’s suffering (H5). To the best of our knowledge, this investigation is one of the first to try to test how children learn how to feel in the context of intergroup conflicts, and the first one demonstrating that mothers’ intergroup goals (as demonstrated by political ideology) are associated with how their children feel toward the outgroup.

Implications

First, to the best of our knowledge, this investigation is the first to integrate research on the socialization of intergroup empathy. There is prior research on the socialization of emotion (e.g., Eisenberg, 2020; Halberstadt, 1986; Hussong et al., 2019; Rogers et al., 2016), the socialization of empathy (e.g., Brownell et al., 2013; Eisenberg & Fabes, 1998; Farrant et al., 2012; Hastings et al., 2007; Strayer & Roberts, 2004; Vinik et al., 2011), the socialization of conflicts (e.g., Aboud & Amato, 2001; Bar-Tal, Reshef et al., 2012, 2017; Nasie & Bar-Tal, in press; Nasie et al., 2016), and on intergroup empathy (e.g., Batson & Ahmad, 2009; Cikara et al., 2014; Hasson et al., 2018, 2019; Levy et al., 2016; Masten et al., 2010; Porat et al., 2016; Rosler et al., 2017; Vanman, 2016; Zaki, 2014). However, no research to date has studied the integration of all of these topics.

This investigation informs the study of intergroup empathy and the study of political ideology. First, most of the existing literature on intergroup empathy has studied decreased empathy toward the outgroup and ways to overcome it in adults (e.g., Cikara & Van Bavel, 2014). Our investigation mapped one possible way empathy toward the outgroup can be formed and developed from childhood. This may be helpful in creating interventions both for mothers and for children, and thus to intervene at an earlier age regarding the decreased empathy toward the outgroup. Second, our findings could also contribute to the ongoing debate in the literature on political ideology and empathy (Hasson et al., 2018). Our investigation showed that rightist and leftist mothers did not differ in their motivation for empathy toward other people in general. It also showed that children did not differ in their empathy toward people in general but only toward the outgroup.

This investigation also extends the prior research on motivation for intergroup emotions (Hasson et al., 2018; Porat et al., 2016), as well as motivation for interpersonal emotions (Netzer et al., 2015). To the best of our knowledge, this is the first time mothers’ motivation for their child’s emotion has been examined (as an interpersonal process), especially in the intergroup context. This investigation was the first step in showing that motivation for emotion plays an important part in the way intergroup emotional tendencies are formed and developed. The correlation between mothers’ motivation and the way they would have liked to express themselves in front of their child also suggested that this motivation for emotion might be directly linked to mothers’ behavior. This link should be further investigated (see “future directions”).

As for political psychology, previous studies of political psychology development mainly focused on intergroup empathy in children (Levy et al., 2016; Masten et al., 2010), how children acquire social intergroup knowledge (Degner & Dalege, 2013; Nasie et al., 2016), and the way intergroup goals may already differ in childhood (Reifen Tagar et al., 2014, 2017). This investigation may contribute to this literature by examining the socialization of intergroup emotions.

Many societies involved in intractable conflicts face similar sociopsychological processes (e.g., Bar-Tal, 2007). Therefore, there are reasons to expect that, in other intergroup contexts, mothers’ motivation for their child’s intergroup empathy may be similarly guided by their intergroup goals. These intergroup goals could differ from one context to another. In this investigation, we focused on political ideology as a reflection of intergroup goals due to its relevance to the Israeli-Palestinian conflict. However, different constructs may better capture intergroup goals in other intergroup contexts. In addition, the relevant ages in which the socialization of intergroup empathy is most pronounced may also differ across intergroup contexts, depending on when mothers talk to their children about intergroup relations.

Limitations and Future Directions

This investigation has several limitations. First, except for Study 4, all studies are correlational. The examined process of socialization of intergroup empathy in the context of the Israeli-Palestinian conflict does not allow us to strongly manipulate political ideology or motivation for emotion. Because the centrality of the conflict is important (Bar-Tal, Sharvit et al., 2012), as shown in Study 4, it is possible that examining this process for other intergroup goals, outside the context of the conflict, may yield different results. Nonetheless, it may be worth extending this investigation to a neutral intergroup context and manipulating mothers’ intergroup goals.

Second, this investigation left a major question unanswered: What does the socialization process look like (behaviorally) in practice? The findings of Study 4 provide only an initial glance at the process and do not illustrate a real-life mother–child interaction. Therefore, future studies should examine behavioral expressions of the mother’s motivation for her child’s empathy as well as the behavioral impact on the child’s empathy by observing mother–child interactions.

In sum, in this investigation, we found that in the context of the Israeli-Palestinian conflict, the degree of empathy mothers wanted their children to experience toward Arabs varied as a function of their political ideology. The extent to which mothers wanted their child to be empathetic toward Arabs (but not toward people in general) was linked to how mothers chose to communicate messages to their child in a real-life context as well as to how empathetic their children were toward Arabs.