Attitudes Toward Bicultural Policy Differentiating the Motivations and Justifi cations Underlying Individual Differences in Pakeha Opposition to Bicultural Policy: Replication and extension of a predictive model

Chris G. Sibley University of Auckland

Marc S. Wilson and Andrew Robertson University of Wellington

This study elaborated upon Sibley, Robertson, and Kirkwood’s (2005) a scale assessing support/opposition recently proposed model predicting individual differences in Pakeha (New for two different aspects of bicultural Zealanders of European descent) support/opposition for the symbolic and policy. The fi rst theme referred to the resource-specifi c aspects of bicultural policy. The theory integrates research symbolic principles of biculturalism, on the function of historical representations and collective guilt for historical defi ned as the degree to which people are injustices within the context of Duckitt’s (2001) model of the dual motivational supportive of the incorporation of Maori and cognitive processes underlying prejudice, and argues that the refutation culture and values into mainstream of responsibility for historical injustices functions as a legitimizing myth (primarily Pakeha) NZ culture and justifying social inequality between Maori and Pakeha. Consistent with national identity. The second theme Duckitt (2001), structural equation modeling indicated that social conformity referred to resource-specifi c aspects of predicted dangerous world beliefs, which in turn predicted Right-Wing bicultural policy, defi ned as the degree to Authoritarianism (RWA), whereas tough-mindedness predicted competitive which people are supportive of policies world beliefs, which in turn predicted Social Dominance Orientation (SDO). that aim to redistribute resources in SDO in turn predicted decreased levels of support for different aspects of favour of Maori on a categorical basis. bicultural policy, and as hypothesized, these effects were mediated by the Previous research using both student and refutation of responsibility for historical injustices. These fi ndings provide general population samples indicated further insight into the ideological attitudes thought to motivate (in this case that although the majority of Pakeha SDO), and the justifi cations thought to legitimize (in this case the refutation supported the symbolic principles of of historical responsibility) expressions of opposition toward different aspects bicultural policy (e.g., Maori language, of bicultural policy in the socio-political environment. The utility Marae greetings, the Haka, wearing of this theoretical framework for assessing both the processes underlying, bone carvings, etc), support for its and the content of, socially elaborated discourses legitimizing discriminatory resource-specific aspects (e.g., land attitudes in other domains and across other cultural contexts is discussed. claims, resource-allocations favouring Maori, affirmative action programs) was dramatically lower (Sibley & Liu, 2004; Sibley, Robertson, & Kirkwood, ew Zealand (NZ) is relatively and is considered by both Maori and 2005). unique on the world stage. Pakeha1 (New Zealanders of European Guided by such findings, we This is due in part to a political descent) to be the most important event N recently outlined a predictive model that system which formally recognizes in NZ’s history (Liu, 2005; Liu, Wilson, attempted to integrate Duckitt’s (2001; Maori (the indigenous peoples of NZ) McClure, & Higgins, 1999). Duckitt, Wagner, du Plessis, & Birum, and non-Maori New Zealanders as Over the last few years, research 2002) seminal research detailing the dual distinct but equal partners. This idea of has begun to map out the different cognitive and motivational processes biculturalism is enshrined in the Treaty themes underlying Pakeha discourse underlying individual differences in of Waitangi, signed in 1840, between and attitudes regarding Maori-Pakeha intergroup attitudes and prejudice with representatives of Maori and the British relations, concepts of biculturalism, research examining the function of colonial government. Once declared and related social policy. Elaborating representations of historical injustice a legal “nullity”, the Treaty began its upon this qualitative work (e.g., Nairn & and the experience of collective guilt for rehabilitation in the late 1960’s (Orange, McCreanor, 1990, 1991, Barclay & Liu, such injustices (Doosje, Branscombe, 2004). It is now regarded as one of the 2003), Sibley and Liu (2004) developed Spears, & Manstead, 1998; Liu & Hilton, legal foundations for NZ’s sovereignty,

New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 • 25 • C. Sibley, M. Wilson, A. Robertson in press). Preliminary fi ndings supported A brief overview of the model and superiority (i.e., those high in SDO) our model of the associations between According to Duckitt (2001; Duckitt to justify expressions of opposition the ideological attitudes motivating, and et al., 2002), individual differences toward bicultural policy. According to justifi cations legitimizing, individual in prejudice are born out of two Sidanius and Pratto (1999; Sidanius, differences in Pakeha expressions of complementary cognitive/motivational Levin, Federico & Pratto, 2001; Whitley, opposition/support for different aspects processes, which result in individual 1999) beliefs about responsibility for of bicultural policy (see Sibley et al., differences in two types of group- historical injustices should thus mediate 2005, for a detailed description of the based motivational goals. Duckitt’s the relationship between primary theory and reasoning underlying this (2001) model is presented in the left ideologically-based motivations (in this model). half of Figure 1. On the one hand, the case SDO) and discriminatory attitudes The present study elaborates upon personality trait of social conformity and outcomes, as indexed by measures our earlier research in this area in three predisposes the individual to perceive of support/opposition for different ways. Firstly, our earlier research has the social world as a dangerous and aspects of bicultural policy. Recent focused solely on the predictive utility threatening place, leading to higher research conducted in also of Social Dominance Orientation (SDO; levels of RWA. The personality trait supports the possibility that perceptions Pratto, Sidanius, Stallworth, & Malle, high in tough-mindedness, on the other of history may function as a legitimizing 1994) and Right-Wing Authoritarianism hand, predisposes the individual to myth in contexts where minority group (RWA; see Altemeyer, 1996, 1998). perceive the world as a competitive members have suffered from historical The present study provides a more jungle, leading to the endorsement injustices. Consistent with our own detailed assessment of the antecedents of SDO. High levels of SDO may research examining Pakeha beliefs in underlying individual differences in therefore be seen as an expression of the NZ context, Reid, Gunter, and Smith Pakeha opposition toward bicultural the motivational goal for group-based (2005) reported that collective guilt policy by including measures of dominance and superiority, whereas low for historical injustices experienced by personality and social worldview that levels refl ect goals of egalitarianism and Aboriginal- on the part of past Duckitt (2001) theorized underlie SDO altruistic social concern. High levels of generations of European-Australians and RWA. RWA, in contrast, are an expression of mediated the relationship between individual differences in European- Secondly, our earlier research in this the motivational goal for social control Australians’ levels of universalism (a area has assessed beliefs about Pakeha and security, whereas low levels refl ect value measure conceptually related to responsibility for historical injustices goals of independence and autonomy low levels of SDO) and compensatory using a combination of items referring (Duckitt et al., 2002). Thus, the dual attitudes. to beliefs about responsibility and process model provides insight into reparation (e.g., ‘We should not have to the processes underlying individual Overview and hypotheses differences in prejudice, and hence pay for the mistakes of our ancestors’) Elaborating upon our earlier work the conditions where one or the other, and items derived from Doosje et al’s in this area (i.e., Sibley et al., 2005; or the linear combination of these two (1998) scale referring to the experience Sibley & Liu, 2004), the current study ideological attitude dimensions (SDO, of guilt and regret for historical injustices tested a Structural Equation Model RWA) will predict discriminatory (e.g., ‘I feel guilty about the negative that replicated the pathways between attitudes. things NZ Europeans/Pakeha have done latent indicators of personality, social to Maori’). Although exploratory factor In NZ, it seems that general worldview and ideological attitude analysis indicated that items referring to political ideology, and hence the issues outlined in Duckitt’s (2001) dual these two concepts loaded on a single surrounding Maori-Pakeha intergroup process model. Applying Duckitt’s underlying factor (Sibley et al., 2005), relations tend to be anchored in issues (2001) model to the NZ context, we the present study uses items referring of equality-inequality and intergroup further hypothesized that SDO, but solely to beliefs about reparation in competition versus harmony, rather not RWA, would predict variation in order to avoid the possible confl ation than issues of danger and threat (see support/opposition for the symbolic and of these two constructs. for example, Liu, Wilson, McClure, & resource-specifi c aspects of bicultural Finally, the present research tests a Higgins, 1999). Accordingly, we expect policy. It was further expected that the Structural Equation Model using latent that opposition toward different aspects effects of SDO would be mediated by variables. The estimation of latent of bicultural policy will be predicted the recognition of historical injustices variables in Structural Equation Modeling primarily by the motivational goal for performed against Maori (these (SEM) has considerable advantages intergroup dominance and superiority predictions are outlined formally in over the use of the observed means as (indexed by SDO), rather than the Figure 1). In this sense, we argue that the it reduces the effects of measurement motivational goal for social control and refutation of responsibility for historical error, thus providing a more accurate security (indexed by RWA). injustices functions as a legitimizing appraisal of the associations between We further posit that the refutation myth that justifies expressions of constructs included in the model. Before of responsibility for historical injustices opposition toward different aspects of presenting these analyses, we fi rst offer may function as a hierarchy enhancing bicultural policy. a brief overview of previous theory and legitimizing myth allowing individuals research relevant to this investigation. motivated toward intergroup dominance

• 26 • New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 Attitudes Toward Bicultural Policy

Method is characteristic or uncharacteristic Right-Wing Authoritarianism (RWA) of YOUR PERSONALITY AND was measured using a shortened set of Participants and Procedure BEHAVIOR.” Adjectives were rated 16 balanced items from Altemeyer’s Participants were 213 undergraduate on a scale ranging from 1 (not at all (1996) scale (items: 8, 10, 12, 13, 15, students who participated for partial characteristic of my personality and 22, 23, 24, 25, 26, 28, 29, 30, 31, 32, course credit and who self-identifi ed behavior) to 7 (very characteristic and 34). The scale included items such as NZ European/Pakeha (the majority of my personality and behavior). as “The only way our country can get ethnic group in NZ). Participants (71 Higher scores indicated higher mean through the crisis ahead is to get back males and 142 females) ranged from levels of social conformity and tough- to our traditional values, put some 17-46 years of age (M = 19.77, SD = mindedness, respectively. tough leaders in power, and silence 4.04). Belief that the social world is the troublemakers spreading bad The measures used in this research a dangerous and threatening place ideas” (pro-trait), and “Our country were included in a larger series of and belief that the social world is a needs free thinkers who will have randomly ordered survey packets competitive place were each assessed the courage to defy traditional ways, unrelated to the current research that using 8 balanced Likert-type items even if this upsets many people” (con- trait). Items assessing SDO were rated were administered early in 2005. One from Duckitt et al’s (2002) scale3. on a scale ranging from 1 (strongly packet assessed (in order) personality Belief that the world is a dangerous negative) to 7 (strongly positive). Items (social conformity, tough-mindedness), place was assessed using items: 3, 4, assessing RWA were rated on a scale and attitudes toward bicultural policy 5, 6, 7, 8, 9, and 10 of Duckitt et al’s ranging from 1 (strongly disagree) and beliefs about historical injustices, (2002) scale. The scale included items to 7 (strongly agree). Con-trait items the other packet assessed (in order) such as “My knowledge and experience were reverse scored so that higher RWA, SDO, and social worldviews tells me that the social world we mean scores indicated higher levels (competitive and dangerous world). live in is basically a dangerous and of SDO and RWA, respectively. The Data from the two packets were unpredictable place, in which good, aforementioned scales have all been matched using confidential student decent and moral people’s values used extensively in the previous identifi cation numbers. The entire set and way of life are threatened and literature and have been shown to of survey packets took approximately disrupted by bad people” (pro-trait), display acceptable internal reliability 25 minutes to complete. Data from and “My knowledge and experience and construct validity (Altemeyer, people who identifi ed with an ethnicity tells me that the social world we 1998; Duckitt, 2001; Duckitt et al., other than NZ European/Pakeha (19 live in is basically a safe, stable and 2002; Pratto et al., 1994). Pacifi c Nations, 34 Asian, 27 Maori, 19 secure place in which most people European, 7 Indian, 1 unidentifi ed; n = are fundamentally good” (con-trait). Beliefs about responsibility for 107) were excluded from the sample as Belief that the world is a competitive historical injustices were assessed using the current investigation focused solely place was assessed using items: 1, three Likert-type items: “If our ancestors on self-identifi ed Pakeha participants’ 2, 4, 5, 8, 9, 10, and 11 of Duckitt et have acted unjustly in the past, then it is attitudes. al’s (2002) scale. The scale included our responsibility to see that those acts Materials items such as “It’s a dog-eat-dog world are corrected in the present” (pro-trait), “I believe that I should take part in the Personality dimensions of social where you have to be ruthless at times” efforts to help repair the damage caused conformity and tough-mindedness (pro-trait), and “The best way to lead to Maori by NZ Europeans/Pakeha in were each assessed using shortened 8- a group under one’s supervision is to the past” (pro-trait), and “We should item versions of Duckitt et al’s (2002) show them kindness, consideration, not have to pay for the mistakes of measures2. The scale assessing social and treat them as fellow workers, not as our ancestors” (con-trait). The second conformity contained the following inferiors” (con-trait). Items were rated of these items was adapted from an pro-trait adjectives: conventional, on a scale ranging from 1 (strongly item originally included in Doosje et respectful, moralistic, and obedient; disagree) to 7 (strongly agree). Con- al’s (1998) measure of collective guilt. and the following con-trait adjectives: trait items were reverse scored so that The other two items were developed by unorthodox, non-conforming, higher mean scores indicated higher Sibley et al. (2005). These items were unconventional, and rebellious. The levels of belief that the social world is rated on a scale ranging from 1 (strongly scale assessing tough-mindedness a dangerous and threatening place, and disagree) to 7 (strongly agree). Con-trait contained the following pro-trait belief that the world is a competitive items were reverse scored so that higher adjectives: unsympathetic, unfeeling, place, respectively. mean scores indicated higher levels of ruthless, and harsh; and the following Social Dominance Orientation the belief that Pakeha of the present con-trait adjectives: compassionate, (SDO) was measured using the are responsible for historical injustices generous, kind, and sympathetic. balanced 16-item SDO scale (Sidanius 6 experienced by Maori that were brought Consistent with Duckitt et al. (2002), & Pratto, 1999), and included Likert- about by European colonials. these items were administered with type items such as “Some groups of instructions to: “Please rate the people are simply inferior to other Attitudes toward the symbolic extent to which you feel each of groups” (pro-trait), and “No one group principles and resource-specifi c aspects the following descriptive adjectives should dominate in society” (con-trait). of bicultural policy were assessed using

New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 • 27 • C. Sibley, M. Wilson, A. Robertson

Sibley et al’s (2005) Pakeha Attitudes contained a balanced number of pro- Hu & Bentler (1999), the hypothesized toward Biculturalism Scale. Five and con-trait items (see Bandalos and model (shown by the solid lines items assessed support for symbolic Finney, 2001, for further discussion presented in Figure 1) approached principles of bicultural policy and fi ve of item parceling). As only three an acceptable level of fi t. Fit indices items assessed support for resource- items assessed beliefs about historical for the hypothesized model are specifi c aspects of bicultural policy. injustice, these three items (rather than presented in Table 2. However, post- The items included in this scale are item parcels) were entered as manifest hoc model modifi cation indices using presented in the Appendix. Items indicators of this construct. In all SEM the Lagrange multiplier identified were rated on a scale ranging from analyses, the three manifest indicators an additional unpredicted positive 1 (strongly disagree) to 7 (strongly created for a given scale were allowed path leading directly from social agree). Con-trait items were reverse to relate solely to the latent variable conformity to belief in a competitive scored so that higher mean scores assessing that particular construct. world. The revised model including indicated higher levels of support for Hu and Bentler (1999) argued this additional path is presented in these two aspects of bicultural policy. that it is important to consider both Figure 1. As summarized in Table 2, Descriptive statistics and internal the standardized Root Mean square the revised model performed well and reliabilities for all scales used in this Residual (sRMR; a residual-based yielded improved fi t indices that fell research are presented in Table 1. fi t index) and one or more index of well within the ranges recommended comparative fit, such as the Root by Hu and Bentler (1999), with the Results Mean Square Error of Approximation sole exception that the GFI remained somewhat lower than recommended. Correlations between self-identifi ed (RMSEA), Comparative Fit Index The revised model also yielded a lower Pakeha respondents’ personality, social (CFI), Non-Normed Fit Index (NNFI), Model CAIC. A Chi-square difference worldviews, SDO, RWA, beliefs about or Goodness of Fit Index (GFI), when test comparing the hypothesized and historical injustices, and support for considering the overall adequacy of a revised models indicated that the the symbolic principles and resource- model. sRMR and RMSEA and values revised model provided a signifi cantly specifi c aspects of bicultural policy are below .08 and .06, respectively, and better fi t, χ2 (1) = 22.31, p < .001. displayed in Table 1. CFI, NNFI, and GFI indices above .95 d.ff are considered indicative of good-fi tting All paths between latent variables In order to conduct SEM analyses models (Hu & Bentler, 1999). The reported in the revised model were using latent variables, we adopted Model Consistent Akaike Information statistically signifi cant, z’s > 1.96, with the partial disaggregation procedure Index (Model CAIC) is also reported the sole exception of the hypothesized outlined by Bagozzi and Heatherton for the purposes of comparison with direct effect of social conformity on (1994). Consistent with Duckitt et al. alternative models. dangerous world beliefs, z = 1.84, p = (2002), the items contained in each .07. The paths from each latent variable scale were parceled to form three The hypothesized and revised to its manifest indicators were also manifest indicators. Item parcels were models all highly signifi cant, β’s > .51, z’s > randomly selected, but where possible According to the criteria outlined by 6.97.

Table 1. Descriptive statistics and correlations between indices of self-identifi ed Pakeha respondents’ personality (social conformity, tough-mindedness), social worldviews (dangerous world, competitive world), SDO, RWA, beliefs about responsibility for historical injustices, and support for the symbolic principles and resource-specifi c aspects of bicultural policy.

Scale 1. 2. 3. 4. 5. 6. 7. 8. 9. 1. Social conformity - -.34* .16* .11 .41* .15* -.15* -.05 -.21* 2. Tough-mindedness - -.02 .37* -.07 .31* -.03 -.19* .05 3. Belief in a dangerous world - .23* .38* .18* -.16 -.16* -.17* 4. Belief in a competitive world - .22* .65* -.27* -.43* -.26* 5. Right-Wing Authoritarianism (RWA) - .35* -.13 -.21* -.15* 6. Social Dominance Orientation (SDO) - -.36* -.44* -.23* 7. Responsibility for historical injustices - .49* .56* 8. Symbolic principles of bicultural policy - .36* 9. Resource-specifi c aspects of bicultural policy -

M 4.46 2.49 3.86 2.63 2.87 2.75 3.21 5.10 2.97 SD .72 .67 .81 .74 .82 .85 1.27 1.21 1.24 α .75 .78 .71 .74 .87 .88 .74 .85 .85 Note. * = p < .05; n = 213 for all correlations. All scale scores ranged from 1 to 7. Higher levels of social conformity and tough-mindedness refl ect higher scores on those personality traits. Higher levels of belief in a dangerous and competitive world refl ect higher levels of these two worldviews. Higher levels of SDO and RWA refl ect higher levels of the motivation for intergroup dominance and superiority, and ingroup conformity and security, respectively. Higher levels of responsibility for historical injustices refl ect higher levels of the belief that Pakeha of the present are responsible for historical injustices experienced by Maori that were brought about by European colonials. Higher levels of biculturalism in principle and resource-specifi c biculturalism refl ect higher levels of support these two aspects of bicultural policy.

• 28 • New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 Attitudes Toward Bicultural Policy

The revised model (see Figure 1) world is a competitive place, β = .49, most parsimonious explanation of the performed well, and accounted for 43% z = 6.00, p < .01. Consistent with relationships between the variables of of the variance in support/opposition predictions, competitive world beliefs interest. This is a common step when for both the symbolic principles and predicted dangerous world beliefs. testing structural equation models, and resource-specifi c aspects of bicultural Consistent with analyses of other has been recommended by various policy. The personality and social NZ samples (Duckitt, 2001); SDO authors (e.g., Lee & Hershberger, 1990). worldview dimensions of Duckitt’s appeared to predict RWA rather than The alternative models considered here (2001) dual process model predicted vice-versa. The revised model further provided markedly worse fi t indices in 63% and 42% of the variance in suggested that social conformity had a all cases. SDO and RWA, respectively. The direct, positive effect on competitive Consistent with Duckitt et al. 2 proportions of predicted variance (R ’s) world beliefs in this data. (2002), we fi rst tested an alternative in all dependent measures included in As also shown in Figure 1, and model in which the placements of the model are displayed in Figure 1. consistent with Sibley et al. (2005), SDO and RWA were swapped with As can be seen in Figure 1, the paths SDO was directly negatively related to competitive and dangerous worldviews between personality, social worldview beliefs about the legitimacy of historical (Alternative model 1). All other paths and ideological attitude predicted by injustices and support for the symbolic in the model remained identical to Duckitt (2001) were all signifi cant and principles of bicultural policy. Beliefs those presented in Figure 1. Thus, the in the expected directions. As expected, about historical injustices were, in turn, model tested the possibility that social social conformity was negatively directly positively related to support worldviews functioned as legitimizing related to tough mindedness. Social for both the symbolic principles and myths akin to beliefs about historical conformity was also directly, although resource-specifi c aspects of bicultural injustices. As can be seen through relatively weakly, related to dangerous policy. Thus, the effect of SDO on comparison of the model fi t indices world beliefs, and dangerous world support/opposition for the resource- presented in Table 2, the revised model beliefs were in turn directly related specifi c aspects of bicultural policy was presented in Figure 1 outperformed this to RWA. Thus, in line with Duckitt’s indirect, being entirely mediated by alternative model. (2001) prediction, there was a trend beliefs about historical injustices, β = - A second alternative model in suggesting that social conformity .29, z = -5.03,-5.03, p < .01. Similarly, close to which beliefs about historical injustices exerted a weak indirect effect on RWA half (i.e., 40%) of the standardized total and attitudes toward bicultural policy that was mediated by dangerous world effect of SDO on support/opposition were included as additional worldview beliefs, β = .05, z = 1.65, p < .10, while for the symbolic principles of bicultural dimensions predicting SDO also being also directly related to RWA in policy was mediated by beliefs about provided poorer fi t indices, as shown its own right. Tough-mindedness was historical injustices, β = -.20, z = -4.20, in Table 2 (Alternative model 2). These directly related to competitive world p < .01. results were further supported by a chi- beliefs, which were in turn directly χ2 Testing alternative models square difference test, d.ff(2)(2) = 2200.42,00.42, related to SDO. Thus, the effects p < .001. of tough-mindedness on SDO were A series of plausible alternative models Finally, a third alternative model indirect, and entirely accounted for were also examined in order to evaluate in which attitudes toward different by intermediary beliefs that the social if the hypothesized model provided the

Figure 1. Structural equation model predicting individual differences in Pakeha students’ attitudes toward the symbolic principles and resource-specifi c aspects of bicultural policy with standardized path coeffi cients. (Note. For simplicity, manifest indicators and paths from latent to manifest indicators are not shown; + p < .07; * p < .05; standard errors displayed in brackets; dashed lines indicate unpredicted paths).

New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 • 29 • C. Sibley, M. Wilson, A. Robertson

Table 2. Model fi t indices for the hypothesized model, revised model, and alternative models.

χ2 (df) sRMR RMSEA NNFI CFI GFI Model CAIC Hypothesized model 407.18 (312) .08 .04 ± .01 .98 .98 .88 827.03 Revised model (shown in Figure 1) 384.87 (311) .06 .03 ± .01 .98 .99 .88 811.08 Alternative model 1 411.66 (311) .08 .04 ± .01 .98 .98 .87 837.86 Alternative model 2 585.29 (309) .14 .07 ± .01 .95 .96 .83 1024.22 Alternative model 3 415.19 (311) .07 .04 ± .01 .98 .98 .88 841.40 Note. sRMR = standardized Root Mean square Residual, RMSEA = Root Mean Square Error of Approximation, NNFI = Non-Normed Fit Index, CFI = Comparative Fit Index, GFI = Goodness of Fit Index, Model CAIC = Model Consistent Akaike Information Index. 90% confi dence intervals for the RMSEA are also shown (denoted by ± .01). aspects of bicultural policy functioned current study contributed to this growing decent people” (Duckitt et al., 2002, as legitimizing myths predicting literature by (a) operationalizing one p. 88). beliefs about historical injustices also aspect of the standard story (i.e., that The dual process model and provided poorer fit indices than the aspect based on the positioning of legitimizing myths model presented in Figure 1, as shown in responsibility for historical events), Table 2 (Alternative model 3).4 All three and (b) modeling its function as a The content of legitimizing myths used of these alternative models included legitimizing myth within the broader in a given domain should depend upon the additional non-hypothesized path contexts of Duckitt’s (2001) model of the social representations used to build leading from social conformity to the dual motivational and cognitive consensus and manage debate about competitive world beliefs.5 processes underlying prejudice and intergroup relations in that context. Thus, intergroup attitudes. the content of legitimizing myths used Summary of results to justify discriminatory attitudes may Our model replicated the causal In sum, the results indicated that a also differ according to whether such pathways predicted by Duckitt (2001). personality disposition high in social attitudes are driven by motivations for The model further indicated that conformity predisposed one to perceive intergroup competition and dominance political ideology surrounding ethnic the social world as both a dangerous and (SDO) or ingroup threat and security group relations in NZ, and particularly competitive place. Dangerous world (RWA). Maori-Pakeha intergroup relations, may beliefs led in turn to heightened levels be more heavily anchored in issues On the one hand, the content of of RWA. A personality disposition of equality-inequality and intergroup legitimizing myths stemming from high in tough-mindedness, in contrast, competition versus harmony (SDO), group-based motivations for dominance predisposed one to perceive the world as a rather than issues of danger and threat may be tailored toward justifying and competitive place, which in turn fostered to ingroup values (RWA). Consistent maintaining hierarchical relations heightened levels of SDO. Those high with this perspective, and in line with between groups. In the context of Maori- in SDO, in turn, tended to express lower Duckitt’s (2001) earlier fi ndings, our Pakeha relations, such myths may be levels of responsibility for historical SEM analyses indicated that SDO had a anchored in (a) notions of equality, and (b) injustices, which precipitated lower signifi cant effect on RWA, whereas RWA the positioning of ingroup responsibility levels of support for both the symbolic failed to exert a signifi cant reciprocal (or lack thereof) for contemporary and principles and resource-specifi c aspects effect on SDO.6 Previous research in historical disadvantages experienced by of bicultural policy (hence the negative South Africa, in contrast, has identifi ed minority group members. Consistent associations between SDO and support the opposite pattern of results, in which with this possibility, recent qualitative for bicultural policy). However, when RWA strongly predicts SDO (Duckitt work in this area indicates that Pakeha considered in a model also including et al., 2002). Given both these and draw upon two different themes when SDO, the motivational goal for social other similar findings (e.g., Duckitt, expressing opposition toward bicultural control and ingroup security indexed by 2001; Sibley et al., 2005), we argue policy. One of these themes appears to RWA did not directly predict opposition that representations of ethnic group be anchored in the notion that equality to bicultural policy in this sample. categorizations in NZ, and particularly should reflect individual merit. The Maori-Pakeha relations, are more likely construction of equality in this way Discussion to be characterized by discourses along allows one to oppose any entitlement, Examples of the discursive strategies the lines of “‘us,’ who are superior, provision or allocation of resources to or ‘standard stories’ used to legitimize strong, competent, and dominant (or particular ethnic groups on the basis or refute Pakeha responsibility for should be) and ‘them,’ who are inferior, that such allocations are unfair to other historical injustices are prevalent in both incompetent, and worthless”, than they individuals who do not belong to that the NZ media and in everyday Pakeha are by ingroup threat- and conformity- group. The other theme appears to be discourse (e.g., Barclay & Liu, 2003; oriented discourses along the lines based on (re-)positioning the relevance Kirkwood, Liu, & Weatherall, 2005; of “‘them,’ who are bad, dangerous, of key historical events (e.g., The McCreanor, 2005; Nairn & McCreanor, immoral, and deviant and who threaten Treaty of Waitangi) in order to refute 1990, 1991; Tuffi n, Praat, & Frewin, ‘us,’ who are normal, morally good, claims based on historical grievances 2004; Wetherell & Potter, 1992). The (Sibley, Liu, & Kirkwood, 2006). To

• 30 • New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 Attitudes Toward Bicultural Policy date, our empirical research in this area in our research may have occurred for and/or social control and conformity has focused on the latter of these two a similar reason, namely that, at the (RWA) to justify discriminatory possible legitimizing myths. Further time this research was conducted in attitudes in a given domain. We hope research is necessary in order to more early 2005, variation in one or more that future research may benefi t from clearly operationalize history- and unidentified external socio-political the theoretical framework outlined here, equality-based legitimizing myths factor(s) may have led to increased and we encourage researchers to further and assess their function in models normative perceptions of the social explore both the processes underlying, predicting attitudes toward bicultural world as a competitive place amongst and the content of, legitimizing myths policy in NZ. Research is also needed Pakeha students. For instance, at the in other domains of intergroup relation to further delineate the effects of beliefs time this data was collected Pakeha and across other cultural contexts. about historical injustices and collective higher in social conformity may have emotion for such injustices. been more likely to hold competitive References The content of legitimizing myths world beliefs because of recent Altemeyer, B. (1996). The authoritarian stemming from group-based motivations political discourse framing Maori- spectre. London, England: Harvard for ingroup security and conformity Pakeha relations as zero sum (e.g., the University Press. (i.e., RWA), in contrast, may be more emphasis placed on Maori land claims Altemeyer, B. (1998). The other explicitly tailored toward maintaining and related Treaty settlement issues by “authoritarian personality”. In M. P. ingroup norms and values. Such myths the National party). However, given Zanna (Ed.) Advances in experimental may center around notions of (ingroup) the number of unpredicted associations social psychology (Vol. 30. pp. 47-92). San Diego: Academic Press. morality and values, and the positioning involving social conformity that have of outgroup threats to ingroup values been observed both here and elsewhere, Altemeyer, B. (1999). To thine own self be untrue: Self-awareness in authoritarians. and way of life. For example, research such findings should be interpreted cautiously and, as Duckitt et al. (2002) North American Journal of Psychology, may fi nd that those high in RWA tend to 1, 157-164. justify discriminatory attitudes toward suggested, should be replicated using Bagozzi, R. P., & Heatherton, T. F. (1994). homosexuals using legitimizing myths alternate measures of social conformity before more substantive conclusions A general approach to representing grounded in biblical interpretations multifaceted personality constructs: of the immorality and sinfulness of may be formed. Application to state self-esteem. Structural homosexual practice. In sum, the current study sought Equation Modeling, 1, 35-67. The legitimizing myths used by to integrate Duckitt’s (2001) research Bandalos, D. L., & Finney, S. J. (2001). Item those high in SDO or RWA may also outlining the individual difference-based parceling issues in structural equation differ somewhat in function at the (personality) and contextual factors modeling. In G. A. Marcoulides & R. E. individual level. For instance, the (perceptions of the environment and Schumacker (Eds.), New developments and techniques in structural equation legitimizing myths used by people high social world) underlying discriminatory attitudes, with research detailing modeling (pp.(pp. 269-296).269-296). Mahwah,Mahwah, NJ:NJ: in RWA may allow those individuals to Lawrence Erlbaum Associates. endorse discriminatory beliefs without how those individuals who endorse Barclay, K. & Liu, J. H. (2003). Who gets experiencing cognitive dissonance in the resulting motivational goals of SDO and RWA draw upon socially voice? (Re)presentation of bicultural instances where such beliefs may be relations in New Zealand print media. elaborated legitimizing myths in order otherwise perceived as non-normative. New Zealand Journal of Psychology, We suspect that people high in SDO, in to justify expressions of opposition 32, 3-12. toward bicultural policy. This research contrast, may not fi nd it necessary to Doosje, B., Branscombe, N. R., Spears, R., justify their discriminatory attitudes to may be thought of as presenting a & Manstead, A. S. R. (1998). Guilty by themselves in this same way, and may be ‘static snapshot’ outlining the structural association: When one’s group has a less likely to experience dissonance in relations between the ideological negative history. Journal of Personality instances where their beliefs differ from attitudes theorized to motivate (namely and Social Psychology, 75, 872-886. perceived social norms (see Altemeyer, competitively-driven dominance and Duckitt, J. (2001). A dual-process cognitive- 1999; Wilson, 2003). superiority indexed by SDO), and motivational theory of ideology and the socially elaborated justifications prejudice. In M. P. Zanna, (Ed.), Advances Caveats and Conclusions theorized to legitimize (in this case the in experimental social psychology (Vol. An unpredicted direct effect leading refutation of responsibility for historical 33. pp. 41-113). New York: Academic from social conformity to belief in injustices) expressions of opposition Press. a competitive social world was also toward different aspects of bicultural Duckitt, J., Wagner, C., du Plessis, I., & identifi ed. Unpredicted direct effects policy. Thus, our research traces an Birum, I. (2002). The psychological bases between social conformity and negative outline that integrates theory and research of ideology and prejudice: Testing a dual process model. Journal of Personality attitudes toward various outgroups have on the processes thought to underlie and Social Psychology, 83, 75-93. also been observed in previous research, individual differences in prejudice, with and have typically been interpreted as research on the function and content Hu, L-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance refl ecting context dependent variation of more specific socially elaborated structure analysis: Conventional criteria in social norms against antiminority discourses used by those with a group- versus new alternatives. Structural prejudice (Duckitt, 2001; Duckitt et al., based motivational goal for intergroup Equation Modeling, 6, 1-55. 2002). The unpredicted effect observed dominance and superiority (SDO),

New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 • 31 • C. Sibley, M. Wilson, A. Robertson

Kirkwood, S., Liu, J. H., & Weatherall, A. towards the general principles and correlated with the full 14-item measure, (2005). Challenging the standard story resource-specifi c aspects of bicultural r(38) = .88, p < .001. The shortened of indigenous rights in Aotearoa/New policy. New Zealand Journal of 8-item measure of tough-mindedness Zealand. Journal of Community and Psychology, 33, 88-99. displayed a similarly high correlation with the full 24-item measure, r(38) = Applied Social Psychology, 15, 1-13. Sibley, C. G., Liu, J. H., & Kirkwood, S. .92, p < .001. These results indicate Lee, S., & Hershberger, S. (1990). A simple (2006). Toward a social representations that the shortened measures of social rule for generating equivalent models in theory of attitude change: The effect conformity and tough-mindedness used in covariance structure modeling. Multivariate of message framing on general versus this research provided relatively reliable Behavioral Research, 25, 313-334. specifi c attitudes toward equality and indicators that were consistent with entitlement. New Zealand Journal of Duckitt et al’s (2002) full measures of Liu, J. H. (2005). History and identity: these two constructs. Psychology, 35, 3-13. A system of checks and balances for 3 A validation sample of 40 undergraduate Aotearoa/New Zealand. In J. H. Liu, T. Sibley, C. G., Robertson, A., & Kirkwood, university students completed the McCreanor, T. McIntosh, & T. Teaiwa, S. (2005). Pakeha attitudes toward the full versions of Duckitt et al’s (2002) (Eds.), New Zealand identities: Departures symbolic and resource-specifi c aspects measures of belief that the social world is and Destinations, pp. 69-87. Wellington, of bicultural policy in New Zealand: a dangerous and threatening place, and NZ: Victoria University Press. The legitimizing role of collective guilt belief that the social world is a competitive place. A composite of the 8 items used Liu, J. H., & Hilton, D. (2005). How the for historical injustices. New Zealand Journal of Psychology, 34, 171-180. to assess dangerous worldview in the past weighs on the present: Social current research was highly positively representations of history and their role in Sidanius, J., Levin, S., Federico, C. M., & correlated with the full 10-item measure, . British Journal of Social Pratto, F. (2001). Legitimizing ideologies: r(38) = .96, p < .001. The shortened 8- Psychology, 44, 537-556. The social dominance approach. In J. T. item measure of competitive worldview displayed a similarly high correlation Liu, J. H., Wilson, M. W., McClure, J., Jost & B. Major (Eds.), The psychology of legitimacy: Emerging perspectives on with the full 20-item measure, r(38) = Higgins, T. R. (1999). Social identity .89, p < .001. These results indicate that ideology, justice, and intergroup relations and the perception of history: Cultural the shortened measures of dangerous representations of Aotearoa/New Zealand. (pp. 307-331). Cambridge: Cambridge and competitive worldviews used in this European Journal of Social Psychology, University Press. research provided relatively reliable 29, 1021-1047. Sidanius, J., & Pratto, F. (1999). Social indicators that were consistent with dominance: An intergroup theory of social Duckitt et al’s (2002) full measures of McCreanor, T. (2005). ‘Sticks and stones these two constructs. may break my bones…’: Talking Pakeha hierarchy and oppression. Cambridge: 4 Additional analyses were also conducted identities. In J. H. Liu, T. McCreanor, Cambridge University Press. examining latent indicators of only the T. McIntosh, & T. Teaiwa, (Eds.), New Tuffi n, K., Praat, A., & Frewin, K. (2004). two personality, two worldview, and two Zealand identities: Departures and Analysing a silent discourse: Sovereignty ideological attitude dimensions outlined Destinations, pp. 52-68. Wellington, NZ: and tino rangatiratanga in Aotearoa. in Duckitt’s (2001) dual process model Victoria University Press. New Zealand Journal of Psychology, 33, (excluding collective guilt and attitudes Nairn, R. G., & McCreanor, T. N. (1990). 100-108. toward bicultural policy). Some readers may be interested to know that this Insensitivity and hypersensitivity: An Wetherell, M., & Potter, J. (1992). Mapping model provided an acceptable fi t when imbalance in Pakeha accounts of racial the language of racism: Discourse and examined separately, χ2 (127, n = 213) confl ict. Journal of Language and Social the legitimization of exploitation. New = 170.38; sRMR = .08; RMSEA = .04; Psychology, 9, 293-308. York: Harvester Wheatsheaf. 90% confi dence interval = .02 < RMSEA Nairn, R. G., & McCreanor, T. N. (1991). Whitley, B. E. (1999). Right-Wing < .06; NNFI = .98; CFI = .98; GFI = .92; Model CAIC = 450.25. For the purposes Race talk and common sense: Patterns Authoritarianism, Social Dominance of comparison with Duckitt (2001; Duckitt in Pakeha discourse on Maori/Pakeha Orientation, and prejudice. Journal of et al., 2002), this model did not include the relations in New Zealand. Journal of Personality and Social Psychology, 77, additional unpredicted path from social Language and Social Psychology, 10, 126-134. conformity to competitive world beliefs. 245-262. Wilson, M. S. (2003). Social dominance and 5 Versions of these three alternative Orange, C. (2004). An Illustrated history of ethical ideology: The end justifi es the models which did not include a direct the Treaty of Waitangi (2nd eed).d). WWellington,ellington, means? Journal of Social Psychology, path leading from social conformity to NZ: Allen & Unwin. 143, 549-558. belief in a competitive world also provided comparably poorer fi t indices than those Pratto, F., Sidanius, J., Stallworth, L. M., & Notes provided by the hypothesized model Malle, B. F. (1994). Social Dominance 1 There is continued debate in New (which also did not include a direct path Orientation: A personality variable Zealand regarding the most appropriate leading from social conformity to belief in predicting social and political attitudes. term describing New Zealanders of a dangerous world). Journal of Personality and Social European descent. Although New Zealand 6 When these reciprocal effects were Psychology, 67, 741-763. European is the most popular term (Liu et included in the model, RWA non- al., 1999), Pakeha is the term that most signifi cantly predicted SDO, β = .09, z = Reid, S. A., Gunter, H. N., & Smith, J. R. strongly implies a relationship with Maori 1.28, whereas SDO remained a signifi cant (2005). Aboriginal self-determination and hence seems most appropriate for predictor of RWA, β = .21, z = 2.58. in Australia: The effects of minority- this paper. majority frames and target universalism 2 A validation sample of 40 undergraduate on majority collective guilt and university students completed the Keywords: compensation attitudes. Human full versions of Duckitt et al’s (2002) Bicultural policy, biculturalism, Communication Research, 31, 189-211. measures of social conformity and tough- legitimizing myths, history, social mindedness. A composite of the 8 items dominance theory, collective guilt, Sibley, C. G., & Liu, J. H. (2004). Attitudes used to assess social conformity in the Treaty of Waitangi, prejudice. towards biculturalism in New Zealand: current research was highly positively Social dominance and Pakeha attitudes

• 32 • New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 Attitudes Toward Bicultural Policy

Acknowledgement Appendix The Authors would like to thank John Duckitt for helpful comments Items and construct defi nitions for the Pakeha Attitudes toward on an earlier version of this Biculturalism Scale manuscript.

Attitudes toward the symbolic principles of bicultural policy Address for correspondence: Chris Sibley Defi ned as the degree to which people are supportive of the incorporation of Department of Psychology Maori values and culture into mainstream (primarily Pakeha) NZ culture and University of Auckland national identity. Private Bag 92019 Auckland, New Zealand 1. The New Zealand should be sung in both Maori and English. E-mail: [email protected] 2. New Zealand should be proud of its cultural diversity and embrace biculturalism. 3. Maori culture should stay where it belongs – with Maori – because it has nothing to do with the rest of New Zealand as a whole. (r) 4. Maori language should be taught in all New Zealand schools. 5. Maori culture should not be pushed on the rest of New Zealand. (r)

Attitudes toward the resource-specifi c aspects of bicultural policy

Defi ned as the degree to which people are supportive of policies that aim to redistribute resources in favour of Maori on a categorical basis.

1. We are all New Zealanders, and Maori should not get special allowances. (r) 2. It is only fair to provide additional resources and opportunities for under privileged minorities, such as Maori. 3. I feel that although Maori have had it rough in past years, they should still be treated the same as everyone else. (r) 4. It is racist to give one ethnic group special privileges, even if they are a minority. (r) 5. I fi nd the idea of giving priority or special privileges to one group appalling, minority or otherwise. (r)

Note. From Sibley et al. (2005). (r) item is reverse scored.

New Zealand Journal of Psychology Vol. 36, No. 1, March 2007 • 33 •