NEED FOR CLOSURE RELATIONS WITH AUTHORITARIANISM , CONSERVATIVE BELIEFS AND RACISM : THE IMPACT OF URGENCY AND PERMANENCE TENDENCIES

————— Both authors are affiliated with the Department of Developmental, Personality, and Social Psychology, Ghent University. This research was supported by a Ph.D. research grant awarded to Arne Roets (Grant 01D06105) under supervision of Alain Van Hiel at the Department of Developmental, Personality, and Social Psychology of Ghent University. The authors thank Ilse Cornelis for her constructive suggestions and proofreading of the manuscript. Correspondence should be addressed to Arne Roets or Alain Van Hiel, Department of Developmental, Personality and Social Psychology, Henri Dunantlaan 2, 9000, Gent. E-mail addresses: arne.roets@UGent.be; alain.vanhiel@UGent.be NEED FOR CLOSURE RELATIONS WITH AUTHORITARIANISM, CONSERVATIVE BELIEFS AND RACISM: THE IMPACT OF URGENCY AND PERMANENCE TENDENCIES


Introduction
Since the publication of Adorno, Frenkel-Brunswik, Levinson, and Stanford's (1950) The Authoritarian Personality, numerous studies have shown that authoritarianism is strongly related to political and ideological attitudes in general, and conservatism and racism in particular (for an overview, see Duckitt, 2001).Modern approaches of authoritarianism (e.g., Right-Wing Authoritarianism, RWA, Altemeyer, 1981;1996) define the concept as the covariation of pro-attitudes toward conventionalism, submission to ingroup leaders and authorities, and aggression toward outgroup members.High correlations were found between RWA and, for example, ethnic prejudice, nationalism, political and economic conservatism, and right-wing political party preferences (for an overview, see Altemeyer, 1981;1988;1996).Also, from the early days on, the hypothesis that authoritarianism is reflected in dysfunctional cognitive performance has attracted a lot of research attention.Classic studies that followed these lines reported that authoritarianism is related to cognitive rigidity (Rokeach, 1948), intolerance of ambiguity (Frenkel-Brunswik, 1949), and low levels of integrative complexity (for an overview, Van Hiel & Mervielde, 2003).
More recently, we witnessed the innovative introduction of Social Dominance Orientation (SDO) which has been defined as "a general attitudinal orientation toward intergroup relations, reflecting whether one generally prefers such relations to be equal, versus hierarchical" (Pratto, Sidanius, Stallworth, & Malle, 1994, p. 742).Many studies also showed that SDO is related to conservative beliefs, right-wing political preferences, and positive opinions about punitive policies and military programs (for an overview, see Pratto, 1999).Because these variables are related to authoritarianism as well, researchers have started to examine the predictive utility of both social attitudes (i.e., RWA and SDO) in explaining right-wing political attitudes and values (Duriez & Van Hiel, 2002;McFarland, 1998) and prejudice (e.g., Altemeyer, 1998;Van Hiel & Mervielde, 2002).

The relationship between social attitudes and cognitive functioning
Scholars recently showed a renewed interest in the motivational aspects of cognitive functioning, eliciting a vast amount of research on constructs like, for instance, the need for closure (Kruglanski & Webster, 1996), need for cognition (Cacioppo & Petty, 1982) and regulatory focus (Higgins, 1998).Especially the need for closure (NFC) has been intensively studied and will be the focus of the present manuscript.NFC has been introduced as an important concept within a theoretical framework of motivational aspects in decision-making (Kruglanski & Webster, 1996) and comprises two sequential strivings.First, strivings for urgency (i.e., the need for quick, unambiguous answers), promote behaviour in which people seize on an early available solution.Second, strivings for permanence (i.e., the need to consolidate previously acquired knowledge), may lead to behaviour in which people freeze on the answer just obtained.The seizing and freezing phenomena are as such manifestations of the underlying urgency and permanence strivings.Webster and Kruglanski (1994) also developed a Need for closure scale to assess individual differences in NFC.This instrument was conceived as a general measure of NFC comprising five facet scales which represent the various ways the concept manifests itself.

An integrative model of social attitudes and cognitive style explaining right-wing ideology and prejudice
Some studies (Chirumbolo, 2002;Cornelis & Van Hiel, 2006;Crowson et al., 2005;Van Hiel et al., 2004) tried to construct integrative models in which the effects of NFC on target variables, such as right-wing beliefs and prejudice, are transmitted through social attitudes.Van Hiel et al. (2004) have indeed shown that the NFC effects on conservatism and prejudice were fully mediated by RWA, as well as partially mediated by SDO.Moreover, separate analyses for the five NFC facet scales yielded striking differences.The relations between the conglomerate of four NFC facets scales -need for order, need for predictability, intolerance of ambiguity, and closed-mindednessand conservatism and racism were mediated by RWA and SDO.Interestingly, exactly these four scales have been identified as the Need for Simple Structure (NFSS), which according to Neuberg, Judice, and West (1997) probe permanence strivings.However, no such a mediation effect was obtained for Decisiveness, and this facet scale was not significantly correlated with social attitudes, nor with conservatism or racism (see also, Kossowska & Van Hiel, 2003).According to Neuberg et al. (1997), Decisiveness is the one facet scale that captures urgency strivings.

Is Decisiveness a measure of urgency strivings?
There has been some criticism on the Decisiveness scale focussing on what this facet scale exactly measures (see, Neuberg et al., 1997).This issue has been fiercely debated (see, Kruglanski, De Grada, Mannetti, Atash, & Webster, 1997) but until recently, no conclusive answer on the nature of the Decisiveness scale was reached.In the meanwhile, some authors have argued that because of the non-significant relationship between Decisiveness and social attitudes, urgency strivings are not related to ideological variables (see, Kossowska & Van Hiel, 2003;Van Hiel et al., 2004).Other authors who studied the relationship between NFC and social attitudes simply omitted the Decisiveness facet scale from their analyses (e.g., Pierro, Mannetti, De Grada, Livi, & Kruglanski, 2003;Chirumbolo 2002).
Recent developments however have advanced our understanding of Decisiveness, and therefore these new findings alter the interpretation of the previously reported non-significant relationship between Decisiveness and social attitudes.Roets, Van Hiel, and Cornelis (2006) found that Decisiveness and the Need for simple structure (NFSS) are not specific indicators of quick decision-making (seizing behaviour) and the preservation of previously made decisions (freezing behaviour) respectively, as has been assumed in the two-dimensional approach of Neuberg et al. (1997).In particular, both NFSS and Decisiveness had (equally) strong relations with tasks that measure seizing behaviour as well as with tasks that measure freezing behaviour.Since urgency tendencies are assumed to lead to seizing behaviour, these results seem to demonstrate that the Decisiveness scale is an adequate (although not exclusive) indicator of urgency tendencies.Therefore the lack of relationship between Decisiveness and right-wing beliefs may suggest the absence of true relationship between these variables.
However, a recent study by Roets and Van Hiel (in press) demonstrated that this latter assumption is not tenable, disclosing the poor validity of the Decisiveness scale as a measure of urgency strivings.In particular, the authors revealed that the actual behaviour of reaching a conclusion quickly is not only affected by the need to make quick decisions (i.e., urgency strivings) but also by the ability to make such decisions.Moreover, they demonstrated that, although the original Decisiveness scale does affect the swiftness of decision-making, as was demonstrated by Roets et al. (2006), it only represents the ability aspect and not the motivational aspect (i.e., the urgency needs) of seizing behaviour.This finding provided empirical evidence for the idea of Mannetti, Pierro, Kruglanski, Taris, and Bezinovic (2002) that many of the Decisiveness items refer to ability.To resolve the validity problem, Roets and Van Hiel developed a new Decisiveness scale that indeed proved to be related to urgency strivings.Put otherwise, their new Decisiveness scale reflects the intended need component, whereas the old scale did not.Moreover, this new scale was reported to be positively related to the other need for closure facet scales.
Hence, the absence of a relationship between Decisiveness and social attitudes reported in previous studies may only demonstrate that the ability to decide is unrelated to such attitudes.However, if a scale that measures the motivational need to seize (urgency strivings) had been used -like the new Decisiveness scale developed by Roets and Van Hiel -a significant correlation could have emerged.Hence, the main research question here is whether the new Decisiveness items are related to socio-political attitudes.

The present study
In the present study we included the original NFCS facet scales as well as the six new Decisiveness items (Roets & Van Hiel, in press) probing urgency strivings.We tested the relationship between NFC, social attitudes, conservatism and racism, as well as a mediation model in which the relationship between NFC and conservatism and racism is mediated by these social attitudes.In particular, we expected to find a significant relationship between the new Decisiveness scale on the one hand and conservatism and racism on the other hand, mediated by social attitudes, whereas no such relationship or mediation was expected for the original Decisiveness scale.

Participants
Data were collected in two samples of undergraduate social sciences students who completed the questionnaires in classroom sessions.The first sample (Sample 1) consisted of 164 participants (64% females) with an average age of 19.7 years (SD = .94).The second sample (Sample 2) consisted of 162 participants (82% females), having an average age of 19.8 years (SD = 3.0).

Measures
Sample 1 participants completed the 34-item NFCS (Webster & Kruglanski, 1994;translated by Cratylus, 1995).This scale includes items like: "I think that having clear rules and order at work is essential for success" and "I dislike unpredictable situations".Additionally, the 6 new Decisiveness items were administered probing the need for quick decisions (Roets & Van Hiel, in press; see appendix for the full scale).These questionnaires were rated on six-point scales anchored by 'Certainly disagree' (1) and 'Certainly agree' (6).An 11-item RWA scale (Altemeyer, 1981; translated by Meloen, 1991) and a 14-item SDO scale (Pratto et al., 1994;translated by Van Hiel & Duriez, 2002) were also administered and rated on five-point scales anchored by 'Certainly disagree' (1) and 'Certainly agree' (5).Representative items from the RWA scale are: "Obedience and respect are the most important virtues children should learn" and "Young people sometimes get rebellious ideas, but as they grow up they ought to get over them and settle down".The SDO scale contains items like: "Some groups of people are simply not the equals of others" and "To get ahead in life, it's sometimes necessary to step on others".
Sample 2 participants completed the 34-item Dutch version of the NFCS, the 6-item new Decisiveness scale, a 24-item RWA scale (Altemeyer, 1988; translated by Meloen, 1991) and a 14-item SDO scale.Additionally, the Middendorp cultural and economic conservatism scales (Dewitte, 1990) were administered in this sample.The 12-item cultural conservatism scale addresses beliefs and values about education, work ethic, position of women in society, abortion and euthanasia.Representative items of this scale are: "A woman is more suited to raise small children than a man" and "Working hard makes you a better person".The 12-item economic conservatism scale addresses issues such as the impact of trade unions, level of state intervention in regulating the economy, and income differences.Representative items of this scale are: "Differences between high and low incomes should remain as they are" and "Economic growth can only be realised when the government allows unrestricted private enterprise".Participants in Sample 2 also completed a 12-item (subtle) racism scale (see Pettigrew & Meertens, 1995;adapted by Van Hiel & Mervielde, 2005) which includes items like: "There are huge differences between immigrants and Belgian people with respect to their religious beliefs and practices" and "It is just a matter of some people not trying hard enough.If immigrants would only try harder they could be as well of as the Belgian people".
Cronbach's alphas and mean scores for all measures in both samples are reported in Table 1.The Need for Simple Structure (NFSS) was calculated by taking the mean of the items pertaining to the four relevant facet scores (preference for order, preference for predictability, discomfort with ambiguity and closed-mindedness) while a mean Decisiveness score was calculated from the Decisiveness subscale items.

Results
Table 2 clearly shows that whereas NFSS was significantly related to RWA and SDO in Sample 1, no relationship between the original Decisiveness scale and these variables was found.These findings corroborate previous results obtained by Kossowska and Van Hiel (2003) and Van Hiel et al. (2004).Conversely, the 6-item new Decisiveness scale, developed as an alternative for the old Decisiveness scale, shows a correlation pattern very similar to that of NFSS, and in line with Roets and Van Hiel (in press), both these NFC scales are strongly correlated.(Altemeyer, 1981), RWA in Sample 2: 24 items version (Altemeyer, 1988).In Sample 2, the pattern of correlations largely replicated the results obtained in Sample 1 (see also Table 2).NFSS and the new Decisiveness scale bore out significant correlations with RWA, but in this sample, the new Decisiveness scale was not correlated with SDO.Importantly, analogous to the first sample, the original Decisiveness scale was not significantly related to RWA and SDO.
RWA as well as SDO showed strong relationships with racism and cultural conservatism, whereas only SDO was significantly related to economic conservatism.NFSS showed significant correlations with all mediator and target variables, with the exception of economic conservatism.The new Decisiveness scale was related to all mediator and target variables, apart from the non-significant correlation with economic conservatism and SDO and a borderline correlation with racism.
In line with the model proposed by Van Hiel et al. (2004), we tested whether the effects of the original and new Decisiveness scale on the target variables were mediated by RWA and SDO, and whether this pattern of mediations was similar to the pattern obtained for NFSS.This model was only tested on Sample 2 data given that not all variables were administered in Sample 1.
Structural equation modeling with latent variables was performed using Lisrel 8.54 (Jöreskog & Sörbom, 1996a;1996b).We tested models with RWA and SDO as mediator variables, and racism and both cultural and economic conservatism as dependent variables for each of the three independent variables in parallel (see Figure 1).Instead of using separate items as indicators for the latent variables, we created three parcels of random items.Parceling is not without critique when carelessly or inappropriately applied (e.g., in examining the dimensional structure of an item set), but its usefulness is generally agreed on when investigating the nature of a set of constructs and their relationships (e.g., Little, Cunningham, Shashar, & Widaman, 2002), which was exactly the aim of the present study.When used to examine the relationships between constructs, parceling has the advantage of resulting in a smaller number of indicators per latent factor, often yielding stronger relationships between individual parcels and the latent factor.Moreover, through the use of parceling, the results are less likely to be influenced by method effects, and are more likely to meet the assumptions of normality (Marsh, Hau, Balla, & Grayson, 1998).To obtain a latent variable for NFSS, the four constituting facet scales were used as indicators.
Following recommendations by Hu and Bentler (1998) and MacCallum and Austin (2000), we examined the Root Mean Squared Error of Approximation (RMSEA, Steiger & Lind, 1980), the Standardised Root Mean Squared Residual (SRMR, Bentler, 1995) and Comparative Fit Index (CFI, Bentler, 1990) to evaluate the goodness of fit of the structural models.
These indicators have been shown to be most sensitive to models with misspecified factor loadings and factor (co)-variances.According to Hu and Bentler (1999), the combined cut-off values of .09for SRMR, .06 for RMSEA and .95 or more for CFI indicate good fit.
Total, direct and indirect effects were assessed using the Lisrel program, and in order to calculate the indirect effects via SDO and RWA, Sobel (1982) tests were computed on the path coefficients of the models.As can be seen in Table 3, NFSS and the new Decisiveness scale showed largely similar mediation effects.While both variables had no total effect on economic conservatism, strong overall effects on cultural conservatism and racism were obtained.The effects of both NFSS and the new Decisiveness scale on cultural conservatism were fully mediated by RWA.NFSS had an additional significant direct effect on racism as well as an indirect effect through both RWA and SDO, whereas the effect of new Decisiveness was fully mediated by RWA only.
In spite of some minor differences between the patterns of NFSS and the new Decisiveness scale, the present results yield an overall consistent pattern of relationships for these two independent variables.However, this overall pattern is totally dissimilar from the one obtained for the original Decisiveness scale.SEM analyses revealed no effects of Decisiveness on cultural or economic conservatism, and a negative total effect on racism that was not mediated by RWA or SDO.

Discussion
The present study yields two major findings.First, with respect to the NFSS and the original Decisiveness scale, we can conclude that NFSS is highly related to social attitudes (especially RWA) and cultural conservatism and racism.Furthermore, these relationships are largely mediated by Right-Wing Authoritarianism.The present results thus corroborate previous studies on the relationship between the need for closure and authoritarianism (e.g., Jost et al., 2003;Kemmelmeier, 1997;Kossowska & Van Hiel, 2003) as well as on the mediating role of RWA in the relationship between NFSS and rightwing attitudes (e.g., Chirumbolo, 2002;Cornelis & Van Hiel, 2006;Crowson et al., 2005;Van Hiel et al., 2004).The theoretical rationale that explains these well-replicated results has been summarised by Jost et al. (2003, p. 348) who argued that "contents that promise or support epistemic stability, clarity, order, and uniformity should be preferred by high-need-for-closure persons over contents that promise their epistemic opposites (i.e., instability, ambiguity, chaos, and diversity)".However, also in line with previous research, the original Decisiveness is generally unrelated to social attitudes, racism and conservatism.
Importantly, the second finding of the present study meaningfully extends the previous results by showing significant relations between the new Decisiveness scale and the target variables (social attitudes and racism and conservatism).Moreover, analogous to NFSS, the effects of the new Decisiveness scale on the target variables were transmitted through social attitudes, and the results therefore show that the mediation model proposed by Chirumbolo (2002), Cornelis and Van Hiel (2006), Crowson et al. (2005), andVan Hiel et al. (2004) is applicable to the present data.In sum, the new Decisiveness scale has much more in common with NFSS than the original Decisiveness scale.We discuss the implications of these results in the remainder and we also address some potential limitations of the present study.

The relationship between urgency tendencies and ideology
Previous studies (e.g., Kossowska & Van Hiel, 2003;Van Hiel et al., 2004) argued that the lack of a significant relationship between Decisiveness and right-wing ideology may reflect the absence of a true relationship between urgency needs and ideology.This explanation has been based on Neuberg et al. (1997) who asserted that Decisiveness captures individual differences in the urgency strivings and non-specific closure.That is, Decisiveness does not refer to specific contents, but would instead involve seizing on any possible solution.Kossowska and Van Hiel (2003) argued that making swift decisions does not account for the fact that information is permanently stored in memory and used in further evaluations.According to the authors, this may exactly be the reason why urgency strivings and the swift decision-making that is promoted by these strivings, are not the basis for political beliefs.
However, the present results cast serious doubts on the latter interpretation.Indeed, the interpretation of the results of previous studies may have been flawed because the old Decisiveness scale is not a measure probing into the intended motivation.The new Decisiveness scale on the other hand contains items referring to the need or motivation to obtain a quick answer, generating results dissimilar from those obtained with the original Decisiveness scale.If we accept that this new scale is a better measure of the desire to reach closure swiftly, as has been argued by Roets and Van Hiel (in press), the present research clearly shows that previous findings have been biased by a less-than-optimal operationalization.Instead, the present research clearly shows that urgency tendencies are related to right-wing ideology.
Then, what does the relationship between urgency strivings and rightwing ideology exactly mean?While people with higher urgency needs strongly prefer 'any' answer over no answer at all, another critical aspect of this striving is that the particular answer acquired should be conclusive and unambiguous and it should reduce uncertainty.The need for 'any' answer motivates people to seek for 'a quick and definite solution', not only in experimental tasks, but also in everyday life.NFC theory states that people high in urgency needs make decisions based on early, immediately available information and disregard additional information.Therefore, people high in the desire for urgency are more likely to consider the most salient and read-ily available information when they have to make a decision or judgment.Hence, their heightened need for immediate closure leads them to turn to those sources that can instantly provide an easily accessible set of clear-cut answers on the nature of society and social relationships.Obviously, ingrained conservative views on social issues supported by authorities, society and common cultural heritage are more readily available and salient compared to an innovative, alternative approach that is off the beaten track.
People high in urgency needs, craving to make up their mind as fast as possible are less likely to go beyond these most readily available, clear-cut answers, because what is already known will always be available before what has yet to be explored.Hence, they are most likely to decide or judge according to well-known schemes, because that is the easiest, fastest and most clear-cut way to satisfy their urgency needs.
Conservative ideologies thus do not only seem to satisfy epistemic, existential and ideological needs to preserve what is familiar (permanency), but also successfully satisfy the need to instantly resolve uncertainty (urgency).The present results therefore suggest that specific (directional) needs for closure and non-specific (non-directional) needs for closure, which are both assumed to influence belief formation by affecting information gathering and processing, are intertwined in a politically conservative direction.

Potential limitations of the present study
The samples used in this study consist of (mostly female) students which may raise questions regarding the robustness of the effects in the general population.Therefore, it can be argued that a more heterogeneous sample may be desirable to confirm these results in further research.However, the effects of NFSS and the original Decisiveness scale in the present study are very similar to the findings obtained in previous studies that also used heterogeneous adult samples (e.g., Van Hiel et al., 2004).Furthermore, in a test of an almost identical model as the one presented in the current study, Van Hiel et al. (2004) did not find differences between the adult and student samples.It is therefore very likely that the present findings, obtained in student samples, will largely apply to the general population.

Concluding remarks
As a general remark for future research on need for closure we want to stress the importance of making clear distinctions between the motivation to reach a decision quickly, the ability to do so, and the resulting behaviour of swift decision-making.The present studies demonstrated that the desire to decide swiftly affects ideological and political beliefs, whereas on the basis of previous studies it is suggested that the ability to make these quick decisions is unrelated to such attitudes and beliefs.We therefore argue that making a distinction between the desire to perform in a certain way (e.g., to decide quickly) and the ability to do so, is primordial for a better understanding of the effect of motivated cognitive functioning on third variables, including political attitudes and ideologies.
Figure 1.Model with RWA and SDO as mediator variables for effects of need for closure measures on racism and cultural and economic conservatism.Note.The model was tested separately with NFSS, the original Decisiveness scale and the new Decisiveness scale as independent measures of need for closure.

Table 1 .
Summary of Cronbach's alphas and test-retest reliabilities in Sample 1 and 2.

Table 2 .
Correlations between need for closure and socio-political attitudes.

Table 3 .
Total, indirect, and direct effects of NFSS, the original Decisiveness scale and the new Decisiveness scale via RWA and SDO on cultural conservatism, economic conservatism and racism.