The willingness of a society to act on behalf of Indigenous Australians and refugees: the role of contact, intergroup anxiety, prejudice, and support for legislative change

Authors


Correspondence concerning this article should be addressed to Anne Pedersen, School of Psychology and Exercise Science, Murdoch University, Perth, WA, 6150, Australia. E-mail: A.Pedersen@murdoch.edu.au

Abstract

The relationship between contact and prejudice against Indigenous Australians and refugees was explored. Using path analysis, increased quality of contact significantly decreased prejudice toward Indigenous Australians, both directly and indirectly through intergroup anxiety; while increased quantity of contact reduced prejudice via a direct pathway. Decreased levels of prejudice toward Indigenous Australians led to increases in support for legislation, which led to increases in willingness to act. Similar results were found for the refugee analysis, except that there was no relationship between quantity of contact and other variables. Qualitative analyses revealed the importance of context, the nature of experience and indirect experience, and societal factors. Our results indicate the power of contact, as well as other structural, interpersonal, and personal factors.

Australia has a long history of immigration, which has resulted in a checkered history of relations between different cultural groups. Now, Australia typically projects itself as an accepting and tolerant multicultural society (Department of Immigration and Citizenship [DIAC], 2003). However, research has found that prejudice and racism are pervasive in Australian society (Dunn, Forrest, Burnley, & McDonald, 2004). It is also evident that some minority groups experience more prejudice, racism, discrimination, and consequent inequities than do others (Human Rights and Equal Opportunity Commission [HREOC], 2001).

Indigenous Australians, in particular, experience much prejudice (Paradies & Cunningham, 2009), as well as disparities in health, housing, employment, and other social advantages (Paradies, 2006). Research also indicates that awareness of prejudice in the community negatively affects the physical and mental health of Indigenous Australians (Paradies, 2006) and is significantly related to anger, suicidal behavior, incarceration rates, and prison experiences of Indigenous Australians (Larson, Gillies, Howard, & Coffin, 2007). Asylum seekers also experience a great deal of negativity (Pedersen, Attwell, & Heveli, 2005), as do refugees (Tilbury & Colic-Peisker, 2006).

Research focusing on different ethnic groups and immigrants has also found prejudice against people of Asian, Arab, and African descent (e.g., Islam & Jahjah, 2001); these regions being predominant areas of origin of many refugees (HREOC, 2005). Perceived discrimination in the community by refugees has likewise been found to be significantly associated with negative well-being (Werkuyten & Nekuee, 1999) and to increase the difficulties of settling and integrating into Australian society (Davidson, Murray, & Schweitzer, 2008). Prejudice not only has negative consequences for specific minority groups, but is detrimental for a society generally and leads to a lack of social cohesion (Pedersen, Walker, & Wise, 2005). If the occurrence of prejudice and its detrimental consequences are to be reduced in Australia, it is important to understand what is behind prejudiced attitudes.

Intergroup contact

One of the most widely researched methods of reducing prejudice in the psychological literature is the contact hypothesis. The contact hypothesis proposes that having contact with a member of another out-group will decrease prejudice toward that group as a whole and will assist in positive future intergroup interactions, as long as some key conditions are met (Allport, 1954).

Allport's (1954) original contact hypothesis held that four prerequisite conditions should be present for intergroup contact to reduce prejudice: (a) group members should have equal status within the contact situation; (b) there should be common goals in place between the groups; (c) there should be intergroup cooperation and minimal competition for the attainment of the common goals; and (d) positive intergroup contact should be sanctioned by authorities, law, or norms. An excellent meta-analysis of 515 empirical studies on the effects of intergroup contact from the 1940s to 2000 found that the majority of studies (94%) showed increased contact had a significant association with decreased prejudice (Pettigrew & Tropp, 2006). Although stronger effects were found for intergroup contact that did meet Allport's prerequisite conditions, the fact that almost all contact was found to reduce prejudice led the authors to conclude that these conditions may be facilitating, rather than essential for reducing prejudice. Nevertheless, research in Australia has reported mixed results on the effects of intergroup contact. It would appear that Pettigrew and Tropp's results are not as clear-cut in this context (Pedersen, 2009). Pedersen reported that the quality of contact is consistently more reliable in correlating with positive attitudes, compared with simple quantity of contact.

Much contact research is based solely on quantitative methods that use predetermined measures of contact and prejudice/attitudes. Some researchers have argued for a reorientation of intergroup contact research that allows participants to express their own views about their experiences with other groups, and focuses on how people themselves view and manage their contact with other groups (Dixon, Durrheim, & Tredoux, 2005). Dixon and Durrheim (2005) used discursive analyses of interviews with White South African beachgoers to show how opposition to desegregation was justified by relying on subtle racial stereotypes and explaining segregation as a normal part of nature. They argued that opposition to desegregation was generated and reproduced by embedded social practices that would not have been exposed through quantitative measures of prejudice or racism.

Connolly (2000) similarly asserted that qualitative analyses allow for an exploration of the impact of the broader social and political environment of contact and intergroup attitudes. Connolly's ethnographic case study of a cross-community scheme arranged for Protestant and Catholic children in Northern Ireland identified the importance of the culture of violence of the locality and wider political events in shaping attitudes. This exploration helped explain why contact with the other group at an arranged cross-community disco did not improve attitudes, confirming negative stereotypes and beliefs, rather than combating them.

Intergroup anxiety

Stephan and Stephan (1985) proposed that intergroup anxiety accounts for many of the unique characteristics of intergroup contact and stems from the anticipation of negative consequences from contact with an out-group member. These anticipated consequences involve fearing negative psychological or behavioral outcomes during contact. They also argued that the amount and type of contact experienced with another group member influences the level of intergroup anxiety felt by that person.

Extensive empirical evidence has now established that intergroup anxiety is central to experiences of intergroup contact (e.g., Brown & Hewstone, 2005; Pettigrew & Tropp, 2008). However, contact has had varied effects on intergroup anxiety; both negative (decrease in anxiety with increases in contact; e.g., Stephan et al., 2002) and positive (increase in anxiety with increases in contact; e.g., Mendes, Blascovich, Lickel, & Hunter, 2002). In a review of 23 studies, Paolini, Hewstone, Voci, Harwood and Cairns (2006) concluded that because of differences in research design, studies reporting detrimental effects of contact on intergroup anxiety were actually indicative of the anxiety-provoking effects of singular contact experiences (where participants had little prior contact). Instead, studies reporting beneficial effects of contact were cross-sectional, correlational studies looking at contact retrospectively, showing the cumulative effects of contact on reducing anxiety over time. This suggests that although contact with an out-group member may initially produce intergroup anxiety, increased contact over time lessens this effect and reduces anxiety.

The consequences of intergroup anxiety on attitudes toward out-groups has similarly been widely documented (e.g., Paolini et al., 2006). In particular, it has been shown that increased levels of intergroup anxiety are significantly associated with increased prejudice levels (Islam & Hewstone, 1993), and reductions in intergroup anxiety have been found to be associated with decreased levels of prejudice (e.g., Paolini, Hewstone, Cairns, & Voci, 2004). Recent work in Australia has shown intergroup anxiety to fully mediate the relationship between cross-group friendship with Aboriginal Australians and old-fashioned racism, issue avoidance, and active avoidance of Aboriginal Australians (Barlow, Louis, & Hewstone, 2009).

Support for change and willingness to act

As noted, prejudice and discrimination contribute to the inequities and disadvantages suffered by Indigenous Australians and refugees. Yet, both groups require significant social change for their circumstances to be improved. Changing prejudiced attitudes toward these groups is a positive step in intergroup relations, but if this does not further promote social action, both groups will continue to be disadvantaged, compared to the majority of Australian society.

Since the early 1970s, a large amount of research has been conducted in the United States to assess the impact of racism on political responses by White Americans toward racial policies predominantly concerning African Americans (for a review, see Sears, 1988). It has generally been found that low levels of prejudice are significantly associated with increased support for racial policies, such as welfare reform and affirmative action, and that high levels of prejudice have been associated with opposition to these reforms (Sidanius & Pratto, 1999). In Australia, it has also been found that as measures of symbolic racism toward Indigenous Australians increased, opposition to government redress in the form of a federal government apology increased (Leach, Iyer, & Pedersen, 2007). These findings show that reduced prejudice may improve support for policies and legislation believed to benefit disadvantaged groups.

However, it is also important to distinguish between abstract support for relevant policies and actual willingness to take action to make these policies or legislation a reality (Leach, Iyer, & Pedersen, 2006). Good intentions may result in nothing if people are not willing to change or take action in the pursuit of the goal they support (Gollwitzer, 1999). Willingness to undertake action has also been found to be a better predictor of real action than more general measures of support for government policies or social justice goals (e.g., Ajzen, 1991). There is very little research that has considered the distinction between support and willingness to act and investigated the association between the two and other variables (e.g., prejudice). According to Ajzen and Fishbein's 1980) theory of reasoned action, the stronger an individual's intention to perform a given behavior, the more likely should be the performance of that behavior. In terms of social change, it is argued that the stronger an individual's support for specific legislation or goals, the more likely he or she will be to take action in support of that change.

Supporting this argument, Leach et al. (2006) found a significant path from modern prejudice to the goal of material compensation in employment and education for Indigenous Australians (i.e., affirmative action), which, in turn, influenced willingness to act to support or not support this goal. Essentially, as prejudice increased, support for the goal decreased, which resulted in decreased willingness to act. While there is a small amount of research regarding Indigenous Australians, none exists with respect to refugees in Australia. Thus, it is important to determine whether people's support for policy/legislative changes actually translates into willingness to engage in action for this support.

Overview of the present study

The present study has two main aims, both primarily concerned with how contact affects attitudes toward Indigenous Australians and refugees. The first aim is to quantitatively investigate how quantity and quality of contact affect intergroup anxiety, prejudice toward the target groups, support for relevant legislative change, and willingness to act on that support. In line with most previous work (e.g., Pettigrew & Tropp, 2006), it is predicted that increased quantity and quality of contact with the target groups will significantly decrease prejudice directly and indirectly through intergroup anxiety. It is also predicted that decreased prejudice will increase support for relevant legislative change, which will increase willingness to act in support of that change (Leach et al., 2006).

The second aim is to qualitatively explore how participants view their experiences with the target groups and how they express why and how this affected their attitudes. Given the limited research on contact with Indigenous Australians and refugees and the differing results mentioned previously, it is hoped that gaining comments from participants about their experiences will help explain and expand knowledge of how contact works in relation to these groups.

Our study is the first known research to investigate the connection between contact, intergroup anxiety, prejudice, legislation support, and willingness to act. As shown previously, there is research showing the relationship between these variables individually. However, there is no known work investigating legislation support and willingness to act as a function of contact, intergroup anxiety, and prejudice. Investigation of the two target groups is, likewise, a novel contribution to the literature. Both groups are minorities within Australian society, where it has been found that the broader community reports unacceptable levels of prejudice toward both groups. However, the history of these groups (and thereby the construction of prejudice) is very different. Using both groups allows for comparison of the aforementioned variables for different groups.

Furthermore, this study uniquely employs both qualitative and quantitative analyses. All methods have strengths and weaknesses (Cohen, 2007). It was envisaged that the qualitative data would reveal factors about contact and prejudice in relation to these target groups that could not be exposed through quantitative prejudice measures alone. Utilizing both qualitative and quantitative analyses, therefore, would allow the strengths of both to produce a clearer, more detailed picture of how contact and experiences affect attitudes toward Indigenous Australians and refugees. Identifying these factors could also provide recommendations for future research for the contact field.

Method

Procedure/participants

In June 2009, four suburbs were randomly selected within the Perth metropolitan area. A questionnaire was delivered with an accompanying letter and a postage-paid return envelope to 1,000 households: 250 to each of the four suburbs. Half of the accompanying letters requested an adult male to complete the questionnaire if possible, while the other half requested an adult female. A reminder letter was delivered 2 weeks later to the same 1,000 households requesting completion and return of the questionnaire. A total of 114 questionnaires were returned, resulting in a response rate of 11.4%.

Construction of the Questionnaire

The questionnaire consisted of five primary sections, including sociodemographics; qualitative questions; and quantitative measures for contact, intergroup anxiety, prejudice, and legislation/policy support. The main qualitative question was placed before the quantitative measures for both target groups.

Sociodemographics

Respondents were asked to specify their age (in years) and their gender (1 = male, 2 = female). They were then asked to provide their ethnic/cultural background.

Contact

Contact quantity and contact quality were assessed by a general contact measure of six items adapted from Prestwich, Kenworthy, Wilson, and Kwan-Tat (2008). The following items measured contact quantity:

  1. How much contact have you had with Indigenous Australians/refugees?
  2. Thinking of social contact, how much contact have you had with Indigenous Australians/refugees in the following ways, on average?
    1. at work, schools, meetings, or events
    2. just chatting to people
    3. over all social situations

Two items measured contact quality (“Do you find the contact is … ?” which was followed by the adjectives pleasant/unpleasant and positive/negative). All items were rated on a 7-point Likert-type scale ranging from 0 to 6. Higher scores indicate increased contact (both quality and quantity).

Intergroup anxiety

Respondents were asked to indicate the degree to which they experienced 13 emotions (e.g., apprehensive) when interacting with or thinking about interacting with Indigenous Australians or refugees. This was adapted by Barlow et al. (2009) from Stephan and Stephan (1985). The emotions were rated on a 7-point scale with higher scores indicating higher levels of intergroup anxiety toward Indigenous Australians or refugees.

Prejudice toward Indigenous Australians and refugees

Prejudice toward Indigenous Australians and refugees was measured by a six-item semantic-differential scale taken from Wright, Aron, McLaughlin-Volpe, and Ropp (1997). Higher scores indicate higher levels of prejudice.

Legislation support

One question adapted from Leach et al. (2006) assessed support for government legislation concerning equal opportunity/affirmative action. Respondents were asked whether they thought the federal government should implement equal-opportunity legislation in relation to Indigenous Australians or refugees. Responses were rated on a 7-point scale with higher scores indicating higher levels of support.

Willingness to act

Immediately following the question regarding legislation support, respondents were asked to indicate regarding 10 items/actions what they would be “willing to do to support their beliefs” (adapted from Leach et al., 2006). The responses were rated on a 7-point scale with higher scores indicating higher levels of willingness to act in support of the participant's position.

Qualitative questions

The qualitative section stated “Respondents in the past have told us that an important reason why they feel the way they do about different racial, cultural, or ethnic groups partly involves their experiences. Does your experience with Indigenous Australians/refugees affect your views?” This item was answered Yes or No. If participants answered Yes, they were given 9 lines to briefly explain what they thought about the question in their own words. A definition of Indigenous Australians (Aborignal and Torres Strait Islanders) and refugees (any person who seeks refuge in Australia and is granted full humanitarian refugee status) was also stated for clarity. In addition, the participants were given 10 lines at the conclusion of the questionnaire to make any other comments about the issues in the questionnaire that would help explain their views better.

Content analysis

We conducted a content analysis on the corpus of the qualitative data. From careful reading of the data, themes were generated inductively, using a form of open coding, as suggested by Glaser and Strauss (1980). After a first reading of the data, four major themes were identified for Indigenous Australians (i.e., context, nature of experience, indirect experience, extraneous issues), and three were identified for refugees (i.e., extraneous issues, nature of experience, context of experience). The reliability of the analysis was assessed by employing an independent rater to code the data using the coding guides developed by Landis and Koch (1977). Interrater reliability was satisfactory. Of the 10 variables (including themes and subthemes for both target groups), 2 had an almost perfect match (0.81–1.00) and 8 had a substantial match (0.61–0.80).

Indigenous Australians

Context

Context was divided into two subthemes, which we describe in order of prevalence. First, the location where people had experienced contact with Indigenous Australians was noted by some respondents. Coding of this theme was then based on the work of social geographers Forrest and Dunn (2009) concerning location of contact with out-groups (0 = didn't mention; 1 = work; 2 = education; 3 = housing; 4 = police/criminal activity; 5 = shopping/restaurants; 6 = sport/public places; 7 = community organizations; 8 = family/friends; 9 = other, e.g., traveling). Interrater reliability was substantial (.70). The second subtheme noted by some respondents was specific comments about the importance of context and impact of behavior on how they felt about Indigenous Australians. Interrater reliability for the specified importance of context was substantial (.63; 0 = didn't mention importance of context; 1 = mentioned importance of context).

Nature of experience

The nature of the experience could be positive experiences, negative experiences, or a combination of both (0 = didn't write about nature of experience; 1 = positive experience or experiences; 2 = negative experience or experiences; 3 = both positive and negative experiences). Interrater reliability was substantial (.72).

Indirect experience

Some respondents commented on the role of indirect experience (things they hadn't actually experienced themselves, but had heard about through other means) on their own and others' views about Indigenous Australians. Once this theme became apparent, coding was guided by the indirect experiential-schematic function measure used by Griffiths and Pedersen (2009; 0 = didn't mention indirect experience; 1 = mentioned things others have told them about their personal experiences; 2 = mentioned comments from others about Indigenous issues, e.g., political opinions; 3 = mentioned media). Interrater reliability was substantial (.80).

Extraneous issues (special treatment)

Although it did not specifically relate to experience with Indigenous Australians, some respondents made comments about the perception that Indigenous Australians receive special treatment by the government (0 = didn't mention special treatment; 1 = mentioned special treatment). Interrater reliability was almost perfect (.88).

Refugees1

Extraneous issues

Some respondents also made comments in relation to refugees that were not specifically or sometimes even remotely related to their experiences with refugees. Three specific subthemes were found: need for integration, confusion, and false beliefs. The items were coded as follows: 0 = no extraneous issues mentioned; 1 = mentioned need for integration; 2 = confused refugee with immigrant or asylum seeker; and 3 = mentioned false beliefs. The need for integration involved remarks about how refugees either should or did not integrate into Australian society, with interrater reliability being substantial (.68). Confusion came from comments that showed there was obvious confusion as to what a refugee actually is, and interrater reliability was substantial (.63). Finally, some respondents made comments that displayed false beliefs—as defined by Pedersen and colleagues (Pedersen, Attwell et al., 2005; Pedersen, Watt, & Hansen, 2006)—about asylum seekers in Australia. Interrater reliability for false beliefs was substantial (.76).

Nature of experience

Respondents noted that their experiences with refugees were positive, negative, or both (0 = didn't write about nature of experience; 1 = positive experience or experiences; 2 = negative experience or experiences; 3 = both positive and negative experiences). Interrater reliability was substantial (.63).

Context of experience

Location of contact with refugees was also noted by some respondents and was coded using the same approach as for Indigenous Australians (0 = didn't mention; 1 = work; 2 = education; 3 = housing; 4 = police/criminal activity; 5 = shopping/restaurants; 6 = sport/public places; 7 = community organizations; 8 = family/friends; 9 = other, e.g., traveling). Interrater reliability was almost perfect (.82).

Results

Sample description

The mean age of the sample was 48.5 years (SD = 15.9, range = 19–87 years), and there were more female respondents (61%) than male respondents (39%). The sample was fairly well educated, with 68% of respondents having participated in or completed at least one post-school qualification.

Compared with the population and census figures provided by the Australian Bureau of Statistics (2007, 2009) in which 59.4% of Western Australian respondents reported at least one non-school qualification and 49.4% of Perth respondents were male, this study's sample was more educated than the general population and had more female respondents. There was a slight leaning for participants to support right-wing politics (37%), with another 33% of respondents reporting being “centre,” and 29% leaning toward left-wing views. In terms of ethnic background, most respondents reported being Caucasian/European (85%), with 3 respondents reporting an African background, 4 reporting an Asian background, 1 reporting an Indian background, and 1 reporting a Middle Eastern background.

Quantitative results

Prior to analysis, missing data were replaced using the SPSS (Version 17.0) expectation maximization (EM) procedure (Piggot, 2000; Schafer & Graham, 2002). Missing data replacement ensured that the analyses of the Australian data and the analyses of the refugee data were conducted on precisely the same sample of respondents (N = 114).

Cronbach's alphas across the five multi-item measures ranged from .89 to .97 for the Indigenous Australian data, and from .91 to .96 for the refugee data (see Table 1). Means and standard deviations for each scale and the intercorrelations among the scales are presented in Table 1. Respondents reported moderate amounts of contact and moderately positive interactions with Indigenous Australians, although they were somewhat anxious about these interactions. In contrast, respondents reported less contact with refugees, but the quality of that contact tended to be more positive. Respondents were also less anxious about their interactions with refugees. Prejudice scores indicated relatively favorable attitudes toward both target groups. Level of support for the equal-opportunity legislation and degree of willingness to act on that support were moderate for both groups.

Table 1. Summary of Means, Intercorrelations, and Reliabilities for the Study Variables as a Function of Target Group
Variable12345678MSDRel.
  1. Note.N = 114. Sex: 1 = male, 2 = female. Intercorrelations for the Indigenous Australian data are presented above the diagonal, and intercorrelations for the refugee data are presented below the diagonal. Means, standard deviations, and reliabilities for the Indigenous Australian data are presented in the vertical columns; means, standard deviations, and reliabilities for the refugee data are presented in the horizontal rows.
  2. *p < .05. **p < .01.
1. Age−.096−.088.078−.200*−.109.108.11248.515.9
2. Sex−.096−.057.177−.035−.270**.083.037
3. Quantity of contact−.006−.005.309**−.326**−.093−.045.1272.571.45.89
4. Quality of contact−.022.232*−.356**−.666**−.652**.457**.1773.011.41.92
5. Anxiety−.047−.110−.352**−.745**.643**−.469**−.267**4.031.51.97
6. Prejudice.116−.245**−.220*−.761**.660**−.380**−.08745.221.1.91
7. Legislation support.038.148.106.413**−.339**−.459**.421**3.721.90
8. Willingness to act.069.062.179.079−.057−.104.361**3.591.35.89
M48.51.793.783.0140.03.463.48
SD15.91.731.141.0921.91.791.39
Rel..97.93.94.93.91

The Indigenous Australian data yielded a number of significant correlations. As expected, respondents who reported higher levels of contact quantity and quality with Indigenous Australians reported lower levels of intergroup anxiety toward Indigenous Australians. There was no significant relationship between contact quantity and prejudice, but increased contact quality was associated with lower levels of prejudice. Respondents with lower levels of intergroup anxiety also reported lower levels of prejudice, and decreased prejudice was associated with increased support for legislation, which, in turn, was associated with increased willingness to act.

A similar, but not identical pattern of significant correlations was observed in the refugee data. Higher levels of contact quantity and quality were associated with decreased levels of intergroup anxiety. Higher levels of contact quality were associated with lower levels of prejudice and, in contrast to the Indigenous Australian data, higher levels of contact quantity were also associated with lower levels of prejudice. Decreased intergroup anxiety scores were again associated with decreased prejudice, which was significantly associated with increased support for legislation. Furthermore, respondents who were more supportive of the equal-opportunity legislation also reported increased willingness to act.

We used LISREL (Version 8.54; Jöreskog & Sörbom, 2003) to conduct two maximum likelihood path analyses. Testing the model requires the estimation of 14 free parameters. For reliable estimates, Kline (2005) recommends 10 participants per parameter (N = 140). He suggests that five participants per parameter (N = 70) would represent the absolute minimum. The current sample size of 114 lies between these two values. The first analysis tested the fit of the proposed path model to the Indigenous Australian data, while the second analysis tested its fit to the refugee data. The fit statistics are presented in Table 2. All four fit indexes indicate that the proposed model provides an excellent fit for the refugee data, and three of the four fit indexes indicate a good fit for the Indigenous Australian data. It is concluded that the models fit sufficiently well for a meaningful interpretation of the path coefficients.

Table 2. Fit Indexes for the Indigenous and Refugee Data
Target groupCFIaNFIbSRMRcRMSEAd
  1. Note.N = 114. CFI = comparative fit index; NFI = normed fit index; SRMR = standardized root mean square residual; RMSEA = root mean square error of approximation; CI = confidence interval.
  2. aA value greater than or equal to .90 indicates a good fit (Hu & Bentler, 1999). bA value greater than or equal to .90 indicates a good fit (Hu & Bentler, 1999). cA value close to .08 indicates a good fit (Hu & Bentler, 1999). dA value less than or equal to .08, or a CI that encompasses this value, indicates a good fit (Jaccard & Wan, 1996).
Indigenous.906.910.082.174 (90% CI = .114 − .240)
Refugee1.000.976.041.000 (90% CI = .000 − .113)

Path coefficients for the Indigenous Australian model are presented in Figure 1, while path coefficients for the refugee model are presented in Figure 2. The direct, indirect, and total effects for the mediator part of the model are reported in Table 3. The pathway from contact quantity to intergroup anxiety failed to reach significance in the Indigenous Australian model (p = .07). All other pathways in the Indigenous Australian model were significant. The pathway from contact quantity to intergroup anxiety was also nonsignificant in the refugee model (p = .139). The refugee model had a second nonsignificant pathway from contact quantity to prejudice (p = .211). All other pathways in the refugee model were significant. The percentages of variance in each endogenous variable explained by the Indigenous Australian model were 46% (intergroup anxiety), 53% (prejudice), 14.4% (legislation support), and 17.7% (willingness to act). For the refugee model, these percentages were 56.4% (intergroup anxiety), 60.4% (prejudice), 21.1% (legislation support), and 13% (willingness to act).

Figure 1.

Path model for the Indigenous Australian data with standardized beta weights as path coefficients. *p < .05. ***p < .001.

Figure 2.

Path model for the refugee data with standardized beta weights as path coefficients. *p < .05. ***p < .001.

Table 3. Direct, Indirect, and Total Effects of Contact Quantity and Quality on Prejudice
 DirectIndirectTotal
  1. *p < .05. ***p < .001.
Indigenous Australians   
Contact Quantity→Prejudice−.175*−.055.120
Contact Quality→Prejudice−.431***−.258***−.689***
Refugees   
Contact Quantity→Prejudice.081−.022.059
Contact Quality→Prejudice−.622***−.160*−.782***

The absence of the pathway from contact quantity to intergroup anxiety in both models prevents contact quantity indirectly influencing prejudice via intergroup anxiety. The hypothesis that contact quantity would have such an indirect influence on prejudice, therefore, is not supported for either target group. It was further hypothesized that contact quantity would have a direct impact on prejudice for both target groups. This hypothesis was supported for the Indigenous Australians, where increased contact significantly reduced prejudice. In contrast, contact quantity had no direct impact on prejudice for the refugees.

It was further hypothesized that increased quality of contact would significantly decrease prejudice directly and indirectly through intergroup anxiety. The direct effect is apparent in both models. The indirect effect also appears to be present in both models, as evidenced by significant pathways from contact quality to intergroup anxiety and from intergroup anxiety to prejudice. The significance of these pathways, however, does not guarantee a significant indirect effect (e.g., Paolini et al., 2004). The LISREL estimates for the indirect pathways from contact quality through intergroup anxiety to prejudice were −.258 (p < .001) and −.160 (p = .015) for the Indigenous Australian model and the refugee model, respectively. Both estimates were significant and negative, providing support for the hypothesis that quality of contact significantly decreases prejudice through intergroup anxiety. Intergroup anxiety, therefore, partially mediates the impact of contact quality on prejudice.

Finally, it was hypothesized that decreased levels of prejudice would increase support for legislation, which, in turn, would increase willingness to act on that support. In both models, the direction and significance of the pathways from prejudice to legislation support and from legislation support to willingness to act are consistent with this hypothesis.

Qualitative results

Of the total sample, 79% (n = 90) made comments in relation to Indigenous Australians, while 72% (n = 82) made comments in relation to refugees. The percentages noted for each theme (displayed in Table 4) are based on the total qualitative data (i.e., the number of respondents who made comments, rather than total respondents).

Table 4. Themes: Indigenous Australians and Refugees
ThemeIndigenous AustraliansRefugees
  1. Note. Separate percentages under themes do not always equal total percentage, as some respondents mentioned more than one code.
Context   74%   28%
Location of experience  65%   N/A 
Location not mentioned35%   72%   
Work 31%   15%  
Education 6%   2%  
Housing 9%   4%  
Police/criminal activity 11%   0%  
Shopping/restaurants 3%   2%  
Sport/public places 22%   4%  
Community organisations 1%   5%  
Friends/family 12%   7%  
Other (e.g., travel) 5%   0%  
Specified importance of context  20%   N/A 
Nature of experience   63%   34%
Nature of experience not mentioned37%   66%   
Positive experiences 11%   17%  
Negative experiences 28%   4%  
Positive and negative experiences 24%   13%  
Indirect experience   18%  N/A 
Indirect experience not mentioned82%       
Things other have told about their experiences 6%      
Comments from others on Indigenous issues 2%      
Media 11%      
Extraneous issues   12%   44%
Extraneous issues not mentioned88%   56%   
Special treatment 12%   N/A  
Need for integration N/A   20%  
Confusion N/A   17%  
False beliefs N/A   18%  

Indigenous Australians

Context

The most common theme in relation to Indigenous Australians was context (see Table 4). Three quarters of respondents (74%) mentioned either one or both of the context subthemes of location and importance of context. The two subthemes highlight the major finding that context matters in terms of both contact and attitudes. Research across disciplines has found prejudiced attitudes to differ across different geographical areas (e.g., Pedersen, Griffiths, Contos, Bishop, & Walker, 2000).

The first subtheme of location of experiences indicates that specific locations of contact also produce different types of experiences and, arguably, different attitudes. Two thirds of the respondents (65%) who made comments in relation to Indigenous Australians mentioned the location where they had experienced contact. The two most common locations (i.e., work, public places) generally generated opposing experiences, with experiences at work often being related as positive (e.g., “I work in the public health sector and have positive experiences dealing with Indigenous Australians”), while experiences in public places tended to be described as negative (e.g., “They cause trouble in the streets and display much antisocial behavior and often escape punishment”). There was even a respondent who specifically compared the two locations in this way:

I have two very different experiences. In the workplace, experiences are not much different [from] all other people of all other backgrounds. I generally deal with professionals or paraprofessionals. In other situations, it is usually quite different; for example, being robbed by Aboriginal people in the street.

The key difference between the descriptions of experiences in these locations appears to be the quality of the experience. Those locations or contexts that allow for better quality of contact and typically encourage the development of relationships display generally positive experiences. Locations where the contact is brief (public places) or is salient because of issues such as antisocial behavior would generally not allow for improved quality of contact to occur. Using the data from the present study and other Australian quantitative data (e.g., Pedersen, 2009), it would seem that contact at locations involving better quality would then produce more positive attitudes. It also highlights the importance of positive contextual factors in contact, as first stipulated by Allport (1954). The second subtheme (i.e., importance of context) was noted by 20% of respondents; for example:

If in the past I have felt uneasy or apprehensive in the company of Indigenous Australians, this has more to do with contextual factors—for example, it was dark; they were male; they were intoxicated and, therefore, unpredictable; I was alone—than with any sense of cultural/religious difference.

Again, this supports the argument made in relation to location of experiences: that different contexts produce different experiences and, consequently, different attitudes. Positively, it shows that some people do not necessarily generalize their experiences with some Indigenous Australians to all Indigenous Australians as a whole.

Nature of experience

The second strongest theme was the expression by respondents of the nature of their experiences with Indigenous Australians. As is indicative from the previous theme, almost two thirds of respondents (63%) talked about the nature of their experiences, relating that they had positive, negative, or a variety of experiences with Indigenous Australians. Comments about negative experiences were the most dominant, mentioned by 28% of the qualitative sample. For example,

I've mainly had negative experiences with Indigenous Australians as neighbors, being abused, broken into, and have experienced antisocial behavior in general. I have lived opposite an Indigenous family for 4.5 years, and they have been a constant source of noise, rubbish, loud and abusive language, and requests for money, cigarettes, car lifts, use of phone, etc.

Following closely were acknowledgments and descriptions of both positive and negative experiences by 24% of the qualitative sample; for example, “When I was younger, I had mainly negative experiences, due to the city train station. When I got older, I had more contact and different experiences in different situations that changed my view positively.”

The prevalence of negative experiences coincides with the documented level of social disadvantage suffered by Indigenous Australians and associated problems, such as high rates of incarceration, substance abuse, and violence, which contribute to experiences of antisocial behavior (Pedersen et al., 2000; Sutton, 2009). However, only a handful of respondents made this connection between Indigenous disadvantage and antisocial behavior. Previous discursive analyses have likewise revealed that in relating negative experiences with Indigenous Australians, some participants viewed these behaviors as independent of other factors, and felt that Indigenous Australians only had themselves to blame for such behavior (Augoustinos, Tuffin, & Every, 2005).

Finally, just over 10% of respondents related only positive experiences. As discussed previously, these experiences were predominantly described in work contexts or through intimate levels of contact, such as family or friends. For example,

One of my children married an Aboriginal woman, and I now have an extended Aboriginal family. My daughter has worked extensively with Indigenous children and their families, and I have also been involved. This contact has given me greater knowledge and insight into the lives of Indigenous people and has resulted in greater respect and compassion for the depth of their disadvantage.

Quality of experience was clearly important to respondents, regardless of whether that quality was good or bad. The large effect size found between contact quality and prejudice in the quantitative data and other quantitative contact studies regarding Indigenous Australians (e.g., Pedersen, 2009) lends support to this.

Indirect experience

Another theme voiced by 18% of respondents was indirect experiences; that is, things they had not actually experienced themselves, but had heard about through other means. Almost all of these comments were made in relation to the media and how predominantly negative stories influenced the public; for example, “The portrayal of Aboriginal people in the media has a huge impact on perceptions. The portrayal is so often very negative—usually depicting violence or abuse. It is rare to see a positive story about Aboriginal people.”

That the media predominantly reports with a negative focus on Indigenous issues has been argued and documented by various professionals and journalists (Sweet, 2009). Yet, positive television campaigns targeting negative beliefs about Indigenous employment have significantly changed beliefs about the proportion of Indigenous Australians in paid employment and the proportion of Indigenous people that remained employed for an extended period of time (Donovan & Leivers, 1993). The way the media report (or don't report) on Indigenous issues has a marked influence on both attitudes toward Indigenous Australians and on the self-esteem and well-being of Indigenous people (Mellor, 2003).

Extraneous issues (special treatment)

Of the respondents, 12% felt that Indigenous Australians were given too much preferential treatment by the government in areas such as welfare benefits and housing. A sample comment is “All Australians should be treated the same, yet Aboriginals get money, land, housing, communities all for free. … We should all have the same rights and not have special ones for Aboriginals!”

The issue of special treatment has been previously documented in Western Australia (Pedersen, Dudgeon, Watt, & Griffiths, 2006). Participants who reported that they felt Indigenous Australians received special treatment were more prejudiced than who those who did not report special treatment. It seems that believing Indigenous Australians are given better treatment may provide people with reasons or justification for negative attitudes, as shown in this example:

I constantly hear about special benefits given to Aboriginal people that White people are not entitled to. … Not being able to walk past an Aboriginal without being asked for money is incredibly annoying. Especially since they piss away all the money the government gives them on alcohol/paint/glue/petrol/whatever else. And they want a f***ing apology from us every year?

Augoustinos, Tuffin, and Rapley (1999) similarly found that negative or racist comments about Indigenous Australians were rationalized in terms of “justifiable” anger over the government support and social advantages they were perceived to have.

Refugees

Extraneous issues

Interestingly, the most common theme for refugees was that of extraneous issues; areas that did not actually relate to experiences with refugees. Close to half of respondents (44%) made comments in relation to the need for integration, confusion, and false beliefs, all three subthemes being mentioned in almost equal numbers. This may be because respondents had less actual contact with refugees and so, instead, felt the need to make comments about other issues in relation to refugees. Regardless of the reason behind this observation, it is interesting that so many respondents talked about other issues, even when specifically asked about their experiences.

One fifth of respondents commented on the need for refugees to integrate into Australian society or noted that they did not integrate well. A particularly illustrative example gave a list of the type of “integration” required:

My main concern with refugees is that they are bringing all their hatred with them, they say they come for a better life, but when they get here they are trying to push their way of life upon us, e.g., religion, dress code. If they want to live here they should

  1. speak English (Australian),
  2. keep religion out of working hours,
  3. take no notice of what other people wear or eat,
  4. keep beliefs and language to their own.

However, integration involves bidirectional acculturation and requires ongoing acceptance of differences and adjustments from both sides (Hollands, 2001). Although respondents use the word integrate, their comments and requirements are instead more reflective of assimilationist ideas; that newcomers must change to fit in with Australian norms. These attitudes not only create a sense of inclusion and exclusion for those who do not “fit in,” but reflect a need to perpetuate the dominance of the stereotypical Anglo-Saxon Australian identity (Fisher & Sonn, 2007). The comments also indicate people feel anxious and threatened by what they believe refugees represent. One respondent reflected on this with the comment “Most Australians are apprehensive about refugees due to the perception of a lack of integration/acceptance of our ‘culture.’ ” People may also feel physically and economically threatened. Previous research found that participants viewed asylum seekers/refugees as potential terrorists (Klocker, 2004) and that attitudes toward multiculturalism are influenced by perceived competition for resources (Esses, Dovidio, Jackson, & Armstrong, 2001).

The second subtheme became apparent through the way some comments were written. It appears that many respondents were unsure about who/what a refugee actually is, and confused them with immigrants, asylum seekers, Muslims, and “ethnic people.” A sample comment is “Refugees must speak English when they migrate here”; “The ethnic groups and ladies I have met”; “In favor of migration at the appropriate time”; “My partner had contact with a Moslem lady”; “Refugees … let them apply through proper channels”; “Don't like the preference being given to boat people” (emphases added).

A particularly illustrative example came from a respondent who in the demographic section stated that his parents were born in Scotland (i.e., he is a second-generation immigrant): “If a family member of a refugee or immigration family commits a crime, the whole family should be deported and never allowed back into this country” (emphasis added). This respondent appears to be unaware that he, too, is a member of an “immigration family,” highlighting the confusion in these terms and how people apply them. This is even though refugee was specifically defined at the start of the qualitative section, identifying a refugee as “any person who seeks refuge in Australia and is granted full humanitarian refugee status.” This confusion is perhaps not surprising, as it seems the terms have been used interchangeably by both politicians and the media (see Every & Augoustinos, 2008). It exhibits a lack of understanding and education about refugee issues (and immigration issues in general) in the community, which may contribute to negative attitudes toward refugees.

The third subtheme found in the data was false beliefs about refugees. Research on asylum seekers has documented that there are a number of false beliefs about asylum seekers pervasive in the community, including “they are ‘illegals’,” “they must be cashed up” (i.e., having a great deal of money), “they get lots of government handouts,” “they are queue jumpers,” and “people who come unauthorized are not proper refugees” (see Pedersen, Attwell et al., 2005). Although our study asked about refugees and specifically stated that a refugee is anyone who is granted full humanitarian status, one of these false beliefs was mentioned in almost 20% of the comments. A sample comment is “When a refugee comes here, [he/she gets] more benefits than an age pensioner who has paid tax for his/her working life in Australia.” High levels of false beliefs have been shown to be strongly correlated with prejudiced attitudes toward asylum seekers (Pedersen, Attwell et al., 2005), and it is likely that attributing these beliefs to refugees would likewise be associated with higher levels of prejudice toward refugees.

Nature of experience

The second most common theme was the nature of experience. Similar to that found for Indigenous Australians, one third of respondents (34%) who made comments in relation to refugees spoke about the nature and quality of their experiences; that is, positive, negative, or a combination of positive and negative experiences. Unlike the results for Indigenous Australians, descriptions of positive experiences (17%) were more common than were negative experiences: “I have dealt with African refugees in business and socially, and associate regularly with Sudanese families. They are a warmhearted, happy, and friendly race of people who integrate into our communities well.” This was followed by statements indicating both positive and negative experiences (13%): “It would be inappropriate to typecast ‘refugees’ based on a range of experiences. Good and bad experiences.”

The comments indicate that those who have had actual experiences with refugees (or who they perceived to be refugees) generally found the contact positive, or acknowledged a variety of experiences. This is supported by the higher level quality of contact found for refugees in the quantitative data and may explain the initial significant bivariate correlation between increased generic contact and decreased prejudice. This may also provide some explanation of the finding by Pedersen (2009) that increased generic contact with asylum seekers in Albany significantly decreased prejudice, especially as Albany had recently had a number of refugees (previously asylum seekers) relocated to the area, which would have increased the possibility of positive contact.

Finally, there were a handful of respondents who related only negative experiences (or what appear like experiences): “Violent, aggressive a***holes from some war-torn country don't deserve to be in Australia when they can't assimilate and carry on here like they are still in their homeland.” Whether this comment is actually reflective of the respondent's own experiences or whether it is based on stories heard from others or the media is unknown. Overall, the focus of nature of experiences again suggests that when people do have contact with refugees, it is the quality of that contact that is deemed most important, which is supported by other Australian data on quality of contact and prejudice (Pedersen, 2009).

Context of experience

The third theme was context of experience with 28% of respondents mentioning places they had experienced contact with refugees. Over half of the contact described was through work (e.g., “I have worked with many refugees and have not had a negative one [experience]”), followed by friends/family (e.g., “My husband's family were refugees from Vietnam”). As argued earlier in relation to Indigenous Australians, both of these types of contact imply a certain quality of contact and more intimate relationships, which helps explain the increased prevalence of positive experiences described previously.

Unlike most Indigenous Australians, refugees are also not a defined ethnic/racial group that can be easily identified without having more detailed information about them (Correa-Velaz, 2006). People are less able to generalize negative superficial contact (i.e., contact in public places) because it is impossible to know whether someone is a refugee without speaking to him or her.

Discussion: quantitative and qualitative findings

Indigenous Australians

The predicted relationships between the variables were supported, for the most part, in relation to Indigenous Australians. Both quantity and quality of contact had a significant direct effect on prejudice. This supports Pettigrew and Tropp's (2006) meta-analytic findings that any contact generally reduces prejudice. Additionally, quality of contact was a stronger predictor of prejudice, compared to quantity of contact. This is consistent with other Australian data. Pedersen's (2009) review of 19 analyses of contact quality and prejudice found that 95% of those analyses showed significant reductions in prejudice when there was better quality of contact. Pettigrew and Tropp also found that contact under optimal conditions yielded stronger reductions in prejudice.

In contrast to previous findings (e.g., Islam & Hewstone, 1993), the role of intergroup anxiety in the contact–prejudice relationship was only confirmed in relation to contact quality. Increased quality of contact indirectly decreased prejudice through a decrease of intergroup anxiety. There was no significant pathway between contact quantity and intergroup anxiety and, therefore, no significant indirect effect on prejudice. The dominant themes of context and nature of experiences for Indigenous Australians in the qualitative data may provide some explanation; for example, “In a work situation, I deal with [Aboriginals]. I am not afraid and have respect for them all. Out of work, where there are a group of antisocial Aboriginals, I am afraid and avoid them.”

It appears that when non-Indigenous Australians encounter Indigenous Australians in public and non-intimate settings, Indigenous Australians unfortunately are often associated with antisocial behavior; for example, “I have met very few Indigenous Australians, so my experience is skewed by those begging or aggressively and loudly arguing in public areas.” This type of interaction will not decrease intergroup anxiety. However, when they have positive experiences (i.e., increased quality of contact) with Indigenous Australians, this anxiety is lessened. Further, in line with our predictions, decreased levels of prejudice significantly increased support for legislation. These results support the findings of Sidanius and Pratto (1999) and Leach et al. (2007).

It is encouraging that reducing prejudice seems to have further impact on support for policies designed to improve the situation of minority groups. In the qualitative data, there was little comment in relation to legislation support. Those who did comment seemed to have differing opinions (e.g., “Aboriginal people as the original inhabitants of this land must be given much more access to employment opportunities” vs. “Affirmative action undermines the ability of these groups to reach equality, as less skilled people may be given preference and jobs are no longer filled on merit”). As shown through the theme of extraneous issues, respondents seemed more concerned about special treatment toward Indigenous Australians in the form of welfare and monetary benefits. Without commenting on the complex issue of whether increased welfare or government benefits actually would improve the situation for Indigenous Australians, the qualitative data suggest that the results may have been different if the policy in question concerned an increase of benefits.

Finally, it was predicted that increased support for legislation would increase willingness to act on that support, which was confirmed. This supports Leach et al.'s (2006) findings that increased support for affirmative action in education and employment for Indigenous Australians increased the willingness to take political action. Since willingness to act on beliefs was asked in relation to both support of and opposition to equal-opportunity legislation, this indicates that those who are in favor of the legislation are more likely to pursue political action in support of the legislation, while those who oppose it are not as willing to act in opposition.

Refugees

There was a significant correlation between contact quantity and prejudice that was lost in the path analysis, this correlational result being supportive of Pettigrew and Tropp's (2006) overall findings. The loss of association in the path analysis is suggestive of the power of quality of contact over quantity of contact on prejudice; that is, once quality is taken into account, this overrides the effects of quantity. Like the results for Indigenous Australians, contact quantity did not influence intergroup anxiety. Considering the past and present portrayal of asylum seekers and refugees as a threat to national security and sovereignty by some Australian politicians and media outlets, this finding for intergroup anxiety is understandable (Klocker & Dunn, 2003). That some in the community feel threatened by what they believe refugees represent is likewise reflected in some examples from the qualitative data; for example, “The anti-European–USA stance by Muslin terrorists has made me aware of being an enemy.” Generic contact—without quality—will not necessarily change this belief. Optimistically, this study indicates through both types of data that quality contact with refugees challenges these views and reduces anticipated anxiety and, consequently, prejudice; for example, “The ethnic groups and ladies I have met have been so polite and friendly, and you don't feel at all nervous.”

The remaining relationships found in the path analysis for refugees supported predictions. Quality of contact once again reduced prejudice, both directly and indirectly through intergroup anxiety. The direct effect size on prejudice was higher than that for Indigenous Australians, but the indirect effect through intergroup anxiety was lower. This suggests that reduction of intergroup anxiety may play a comparatively larger role in relation to attitudes toward Indigenous Australians.

In line with predictions, reductions in prejudice increased support for legislation, which, in turn, increased willingness to act. This supports previous findings by Leach et al. (2006) and confirms the far-reaching impact contact can have on effecting social change. As stated in relation to Indigenous Australians, it also indicates that those who are supportive of social changes are more likely to act on that positive support, while those who do not support such an equal-opportunity policy are perhaps more apathetic about taking action against it.

Interestingly, there were very few participants who made specific comments about affirmative action or equal-opportunity legislation. Of those who did comment, 75% were negative in nature. A sample comment is

Coming from a deprived background, I don't believe legislation succeeds unless there is a willingness on the part of any disadvantaged person to take advantage of any education or vocational training on their own initiatives. Being forced to hire someone because of legislation is detrimental to the employer and employee. Legislation causes social misunderstanding and discrimination.

It may be that the results found here are an artifact of the measure used; that is, using affirmative action for policy and the types of action that could be taken. As is seen in the remainder of the paper, there were numerous comments made about special treatment/receipt of benefits in relation to both Indigenous Australians and refugees (e.g., “Indigenous [Australians] and refugees receive too much assistance from the government”). The results found for legislation support and willingness to act may be very different if changes were proposed for economic benefits for the target groups.

Comparison of target groups

The core findings of the quantitative data, while not identical, were similar for both target groups. Contact quality influenced prejudice, both directly and indirectly (via intergroup anxiety) for both Indigenous Australians and refugees. It is clear that quality of contact plays a significant role in the reduction of prejudice toward out-groups. The data suggest that prejudice is more strongly associated with people's quality of interactions, rather than the actual amount of contact with out-groups, which is in line with Allport (1954). In the qualitative data, the theme of nature of experience found for both target groups corroborates this finding. When asked about how their experiences affect their attitudes, respondents spoke frequently about how their contact was positive, negative, or both; for example, “My contact with refugees has been mostly in my adult life and generally positive.”

This does not negate quantity of contact, as it is obviously impossible to have any quality contact without actual contact. In relation to the Indigenous path model, increased quantity of contact decreased prejudice, but not to the extent that increased quality of contact decreased prejudice. However, it does indicate that we cannot simply expect that having different groups living together in the same neighborhoods and cities will eventually lead to improved relationships. More intimate and positive contact in encouraging contexts is required.

The results also show that reduction of intergroup anxiety is an important factor in the relationship between contact quality and prejudice. Intergroup anxiety may play a larger mediating role in influencing attitudes toward Indigenous Australians than attitudes toward refugees. As mentioned previously, this may be because of the nature of interactions with each target group. The qualitative data also suggest that the anxiety relating to each target group may stem from different issues: one the witnessing of antisocial behavior (e.g., “Some Aboriginal people are loud, aggressive, violent, pushy for money, so I tend to view them all with caution”), and the other seemingly from media reporting (e.g., “They [refugees] protest and demonstrate and attack in groups against the elderly and young Australians in cities [as shown] by TV coverage”). Knowledge of how anxiety is “caused” and has different effects is important for any effort to change levels of prejudice.

It is interesting that intergroup anxiety did not play a significant mediating role in the relationship between contact quantity and prejudice for either target group. Although there was initially a significant correlation between contact quantity and intergroup anxiety for both target groups, this was lost in the path analysis. Having knowledge of the qualitative data described previously, the finding is perhaps not so surprising. If contact only reinforces negative interactions, anxiety will not be decreased. The confusion found in relation to what/who refugees are also shows that people may be making assumptions in their contact with various other groups (e.g., Muslims, ethnic groups), which has a negative impact for refugees.

The major point of difference between the target groups in terms of the quantitative data was found in the relationship between contact quantity and prejudice. Contact quantity directly reduced prejudice for Indigenous Australians, but not for refugees. However, there was initially a significant correlation between contact quantity and prejudice with respect to refugees. This indicates the power of quality of contact in relation to refugees; once contact quality was taken into account, it negated the effect of contact quantity. When considering the qualitative data, this may be explained by factors relating to the nature of the target group itself. First, it is difficult for anyone to determine whether someone is a refugee, based on generic contact. Some quality of contact is required to even be aware that a person is a refugee. Second, perhaps as a consequence of the first factor, there is obvious confusion about who refugees are.

In relation to legislation support and willingness to act, the qualitative data reveal that Australians appear to be more aware of disadvantage and need for policy change with respect to Indigenous Australians when compared to refugees. Some affirmative-action-type policies are already in place for Indigenous Australians and are typically discussed more in relation to Indigenous Australians than refugees. Because of the history of Australia and colonization, social disadvantage in relation to Indigenous Australians may also be more dominant in the public mind. This is confirmed by comments from the qualitative data, such as “I am sympathetic to the social and health problems of Indigenous Australians and wish there was a solution to [the] current situation.”

Implications, limitations, and future research

Through both the quantitative and qualitative data, quality of contact clearly appears to be a dominant factor in improving attitudes and encouraging social action. Over 50 years ago, Allport (1954) gave us conditions that attempt to increase positive contact. It seems that these, along with other methods, must be revisited for future anti-prejudice programs or interventions to be successful.

The fact that the qualitative data revealed that intergroup contact is frequently negative in relation to Indigenous Australians is important. When viewing the quantitative data alone, respondents reported moderately positive interactions with Indigenous Australians. Given the qualitative comments about negative contact, this moderate level is likely reflective of both positive and negative contact. It would be an idea for future contact research to review how contact is measured quantitatively. For example, future measures could integrate positive contact and negative contact in separate items so they can be analyzed separately.

The qualitative data also reveal significant wider social and political factors that impact on attitudes. How the media report clearly has ramifications for prejudice and contact. However, the role of the media in the context of intergroup relations has not been studied extensively (for a notable exception, see Paluck, 2009). Further research on the relationships between contact, media reporting, prejudice, and their implications could provide important information for changing attitudes.

The documented beliefs of special treatment in relation to Indigenous Australians may also be a factor influencing the quantitative findings, particularly in relation to legislation support and willingness to act. Future research measuring other social policies or action in relation to Indigenous Australians would be necessary to determine if the relationships found in this study are unique to equal-opportunity legislation/affirmative action or whether they replicate across other policies and social changes.

The findings of the present study should also be interpreted with its limitations in mind. First, the sample size was restricted in number. A community sample was chosen to increase the relevancy of the study's findings; the disadvantage being that the amount of responses could not be estimated or controlled. Having said that, other community-based studies using survey and questionnaire materials have found similar response rates (e.g., Watt & Larkin, 2010). Second, only Perth residents were surveyed, limiting the generalizability of these findings. Considering the importance found for context of contact, it would be interesting to look at how contact with these target groups operates in different locations.

The present study indicates that under the right circumstances, increased contact with Indigenous Australians and refugees can reduce prejudice, although contact alone is not as influential as contact quality. It also indicates that there are differences in the prejudice–contact relationship, depending on the target group and the social circumstances surrounding those groups. Interestingly, there are other factors and issues that were important enough for participants to bring up, even when the question asked of them did not relate to these issues (e.g., special treatment in relation to Indigenous Australians, false beliefs in relation to refugees). Finally, context is important: We do not hold our attitudes in a social vacuum; our surroundings are important—especially when the political climate gives rise to prejudiced attitudes. Clearly, these issues also need attention. Contact is only one piece of the puzzle, albeit at times an important one.

Given the harmful and pervasive impact prejudice and discrimination have on their targets, as well as society as a whole, it is important to know why these community attitudes persist and how they can be changed. However, in addition to dealing with such attitudes, anti-prejudice sentiments must be publicly endorsed, which has not always been the case. In short, structural, interpersonal, and personal factors must all be addressed. Then, Australia may be able to claim that it is an accepting, multicultural society. At the moment, it cannot do so.

Footnotes

  1. 1

    Once again, we describe the themes in order of prevalence.

Ancillary