This is an open access article distributed under the terms of the Creative Commons
Attribution License (Most of the world's refugees currently reside in developing countries (United Nations Higher Commissioner for Refugees [UNHCR], 2013). Despite this, refugees’ arrival to western countries has become highly politicised. Asylum seekers and refugees are often seen as a threat to national sovereignty, the economy and social cohesion, giving rise to questions about the type of entitlements, support and rights that should, or should not, be received or be granted in the host society (Verkuyten, 2004). While there is large body of research that has explored community attitudes towards immigrants (e.g., Esses, Deaux, Lalonde, & Brown, 2010), relatively few studies differentiate between subgroups of new arrivals, such as refugees and asylum seekers (Murray & Marx, 2013). This is, in part, due to the assumption that all members of an outgroup will be evaluated and perceived in similar ways by ingroup members (for a review see Fiske, 1998). Yet, differences in perceptions towards such subgroups can affect attitudes and behaviours (Kaiser & Wilkins, 2010). Verkuyten (2004), for example, found feeling empathy towards political refugees predicted more generous policy attitude, whereas feeling angry towards those perceived to have fled due to economic reasons predicted support for harsher policy. Thus, a subgroup analysis can provide a more nuanced understanding of how and why people may support particular social policies for one subgroup but not another.
In this study, we examine Australian community members’ support for social policies aimed at providing assistance to resettled refugees (who are flown from overseas to Australia via its official refugee resettlement program) and asylum seekers (who arrive in Australia by boat and then seek refugee status). While there are also asylum seekers who arrive on a valid visa, such as a tourist or student visa, and then seek refugee status, only those who arrive by boat are explored here, as they are the group who have faced the most punitive treatment by successive Australian governments. We turn briefly to an overview of the socio-political context in which the study was conducted.
Constructions of Resettled Refugees and Asylum Seekers in Australia
Australia has a Refugee and Humanitarian Program (RHP) which consists of an onshore and offshore component. As a signatory to the 1951 United Nations Convention relating to the Status of Refugees as amended by the 1967 Protocol (Refugee Convention), the onshore component attends to people who have entered Australia and lodged a refugee protection claim (Kenny & Fiske, 2014). People who are ‘onshore’ when they apply for refugee status are known as asylum seekers until their cases are determined and either enter Australia without a valid visa, with the majority in this category arriving by boat, or with a valid visa, such as a student or tourist visa (Lusher, Balvin, Nethery, & Tropea, 2007). While international law does not distinguish between refugees and asylum seekers and their mode of arrival, those who arrive in Australia without a valid visa are subject to mandatory detention while their claims are being processed.
The second component of the RHP is the offshore component which offers resettlement visas to refugees outside of Australia. This may include refugees who have been recognised by a UNHCR camp or by an Australian embassy. Refugees who arrive in Australia in this manner are often referred to as “resettled refugees”. Although the Refugee Convention does not oblige Australia to accept offshore refugees, Australia sets an annual quota for the numbers of visas granted in its offshore component of the RHP (Kenny & Fiske, 2014). In 2013-2014, 13,768 visas comprising 11,016 offshore refugees and 2752 to onshore refugees (Department of Immigration and Border Protection [DIBP], 2013). As a matter of policy, Australia links onshore and the offshore components; if there is an increase visas given to those found to be refugees who arrived in Australia onshore, there is a decrease in visas offered to refugees offshore. The linking the two RHP components has contributed to the politicisation of asylum seekers, in part due to the idea that such arrivals are taking the places of refugees waiting elsewhere (Kenny & Fiske, 2014).
In this context, resettled refugees and asylum seekers who arrive in Australia by boat are often represented in political discourse in binary terms such as “genuine refugees” and “legal” (resettled refugees) versus “non-genuine refugees” and “illegal” (asylum seekers) (Klocker & Dunn, 2003; McKay, Thomas, & Kneebone, 2012; Pedersen, Watt, & Griffiths, 2008; Rowe & O’Brien, 2014). During the 2013 Australian Federal Election campaign, the former opposition party, now the Australian government, consistently referred to asylum seekers as “illegal arrivals” who “turn up illegally” on “illegal boats” (Creek, 2014). The Australian media has also been found to depict asylum seekers as “illegal immigrants” and “queue jumpers” (e.g., Every & Augoustinos, 2008). Constructions that pit one group of refugees over another can be found in other countries such as Canada (e.g., Bauder, 2008), the United Kingdom (e.g., Lynn & Lea, 2003) and the Netherlands (e.g., Verkuyten, 2004). It is important to note that being a signatory to the Refugee Convention, Australian law permits unauthorised entry for the purposes of seeking asylum, and thus asylum seekers who arrive by boat are not illegal (Hartley, Pedersen, Fleay, & Hoffman, 2013). In addition, over 90% of most asylum seekers who arrived by boat over the last decade have eventually been found to be refugees (DIBP, 2013).
Social policies that restrict the rights and supports offered to asylum seekers who arrive by boat over the past 20 or so years also reinforce these oppositional constructions. For example, as noted previously, asylum seekers who arrive in Australia without a valid visa are subject to mandatory detention. In November 2010, some asylum seekers have been released from detention into the community while waiting for the outcome of their refugee claims but are given limited welfare support (Hartley & Fleay, 2014). On the other hand, the supports and entitlements offered to resettled refugees have been argued to be some of the best in the world (Fozdar & Hartley, 2013). To a certain degree, community attitudes towards refugee groups in democratic countries may affect the policies implemented regarding refugees’ access to resources, entitlements and support in the host society (Marr & Wilkinson, 2003). Thus, the question of what factors underpin social policy attitudes is pertinent.
Social Policy Attitude, Prejudice and Emotions
There are a number of social psychological factors that may be useful in predicting social policy attitude for resettled refugees and asylum seekers, such as prejudice and emotions. While these factors have been explored relating to asylum seekers in various ways (for a review see Pedersen & Hartley, 2015), no research has focused their relationship with attitudes towards resettled refugees. Traditionally, prejudice has been conceptualized as a general attitude and has been the focus of much research designed to explain the origins, processes, and implications of feelings and behaviors towards outgroups (for a review see Brewer & Brown, 1998). In Australia, community attitudes towards asylum seekers have been found to be largely negative (e.g., McKay et al., 2012; Schweitzer, Perkoulidis, Krome, Ludlow, & Ryan, 2005). By contrast, little research has focused on prejudice towards resettled refugees (Fozdar & Hartley, 2013). One exception is a community survey of over 2000 Australian community members which found that 75% were more positive towards resettled refugees compared to asylum seekers who arrive by boat (Markus, 2012).
Although prejudice may offer some insights to this issue, value also lies in considering how emotions towards different groups can help to predict behaviors (Cottrell & Neuberg, 2005; Cottrell, Richards, & Nichols, 2010; Iyer & Leach, 2009). For example, refugees have been associated proximally with negative emotions such as fear and anger (e.g., Lynn & Lea, 2003; Schweitzer et al., 2005). Fear and anger have also been found to promote negative action tendencies toward outgroup members (e.g., Mackie, Devos, & Smith, 2000). Similarly, Cottrell et al. (2010) argue that the specific feelings towards an outgroup should be related to perception of threat posed by an outgroup, and that threat should predict action tendencies relating to negative treatment.
Fear and perceptions of threat have been suggested to play an important role in shaping attitudes towards outgroups in general, and immigrants in particular, including policy attitudes relating to these groups. Fear towards an outgroup, for example, has been conceptualised as fostering self-protective behaviour which in turn promotes preferences for harsher and protective policy initiatives (Cottrell et al., 2010). Similarly, Montada and Schneider (1989) have argued that fear of losing one’s group advantages can be related to support for policies that seek to stabilise the status quo. While no research has explored fear towards resettled refugees, Pedersen, Watt, and Griffiths (2007) examined prejudice towards asylum seekers and found that community participants who held more negative views about asylum seekers were also significantly more likely to report a fear of terrorism.
Previous research has also found a relationship between the perception of threat and the emotion of fear (Cottrell & Neuberg, 2005). These authors argue that perceiving the obstruction of valuable goals, or the taking of valuable resources, produces anger and a tendency to aggress, whereas perceiving a threat to physical safety produces fear and a tendency to flee. One group of threats that has been proposed are realistic threats, which relate to threats to the political and economic power of an ingroup (e.g., Stephan, Renfro, Esses, Stephan, & Martin, 2005). On the other hand, symbolic threats concerns group differences in values, beliefs, morals, and attitudes. Research has found that attitudes towards immigrants are more likely to be negative when immigrants are perceived to pose realistic and symbolic threats to one’s ingroup (e.g., Stephan et al., 2005). In this vein, Schweitzer et al. (2005) found that negative attitudes towards asylum seekers were related to perceptions that they pose a symbolic threat to Australia’s culture, beliefs and moral and religious values. Previous Australian research has found that realistic and symbolic threat with regard to asylum seekers factors into one factor: they are statistically inseparable (Suhnan, Pedersen, & Hartley, 2012).
Anger has also been linked with attitudes towards outgroups and policy attitudes in particular. Traditionally, research has explored the role of anger in predicting intergroup behaviour and actions when focusing on disadvantaged group members’ anger towards an advantaged group (e.g., van Zomeren, Spears, Fischer, & Leach, 2004). On the other hand, advantaged members who deny their structural advantage have been found to feel anger towards a disadvantaged group if such a group are perceived to receive “unfair” advantages (such as welfare benefits or perceived preferential treatment). Similarly, Leach, Iyer, and Pedersen (2006) found that non-Indigenous Australians who perceived their ingroup as relatively deprived compared to Indigenous Australians were more likely to oppose government action to redress. This type of anger may also be relevant for understanding social policy attitudes for resettled refugees and asylum seekers. For example, Verkuyten (2004) found that when refugees were constructed as “economic refugees”, anger predicted support for more harsh policy regarding refugees. The key issue here was the perception that “economic refugees” are responsible for their situation and thus undeserving of help. While previous research has found that anger towards asylum seekers has been linked with prejudice (e.g., for a review see Pedersen & Hartley, 2015), the relationship with policy attitude has not been explored. However, as asylum seekers who arrive by boat are often constructed as ‘illegal’ compared to resettled refugees, anger may lead to more restrictive social policy attitude.
Overview of the Study
Using a mixed methods approach, community attitudes towards social policies regarding resettled refugees and asylum seekers were quantitatively examined in relation to feelings of anger, fear, and perceived threat, as well as levels of prejudice. Following Greene’s (2006) mixed-method framework, quantitative findings were compared with participants’ responses to an open-ended question to extend our understanding of social policy attitudes.
The research involved three primary aims. The first was to compare participants’ support for social policies relating to resettled refugees and asylum seekers and the levels of prejudice, threat, fear and anger towards the two groups. In view of the evidence that suggests resettled refugees and asylum seekers who arrive by boat are often constructed in oppositional ways by the media, politicians, and the general public, our first hypothesis is that levels of threat, anger, fear and prejudice will be significantly higher for asylum seekers than for resettled refugees, and participants will favour more restrictive policy for asylum seekers than for resettled refugees.
The second aim was to examine predictors of restrictive social policy attitude. As discussed, researchers have moved beyond general prejudice to examine the specific emotions that people experience towards other groups (Cottrell et al., 2010). Given that previous research finds that many Australians see asylum seekers who arrive by boat as illegal and queue-jumpers and thus violating moral codes, we hypothesised that anger would be more predictive of a more restrictive policy attitude. We also hypothesised that there would be little difference in the predictive value of threat and fear for resettled refugees versus asylum seekers as, given that each one is an outgroup, they both could be perceived as threats to the Australian economy and culture. In line with arguments made in previous research (e.g., Iyer & Leach, 2009), emotional reactions that are relevant to the outgroup should be more predictive of policy attitude towards outgroups over and above prejudice.
Our final aim involved an analysis of the qualitative data. Given previous research that suggests that there is a degree of negativity among the general population to immigration generally but high degrees of negativity towards asylum seekers specifically (Fozdar & Hartley, 2013), we were interested to see what participants spontaneously mentioned in response to a broad, open-ended question. In particular, we were interested to see whether participants mentioned issues relating to asylum seekers, resettled refugees or both and whether these groups were referred to differently in some aspects (e.g., asylum seekers as violating some form of moral code) and similarly in others (e.g., new arrivals perceived as threats). Following Greene’s (2006) mixed-method framework, to provide breadth to the quantitative data, an open-ended question was asked so that participants were able guide what issues they wanted to discuss in relation to resettled refugees and asylum seekers which would highlight what was salient and important to them. It was anticipated that both the quantitative and qualitative data would provide complementary, and possibly unique, contributions to the research questions.
Method
Participants
A total of 185 adults from the Perth community in Western Australia participated in our study using a convenience sample with six people acting as research assistants (see acknowledgements). The mean age of the sample was 38 years with a range from 18 to 80 years, and there were more male (54%) than female (46%) participants. Overall, the participants were well educated, with 38% holding or currently completing bachelor’s degrees, while a further 9% had achieved or were completing higher university degrees. Compared with the population figure provided by the Australian Bureau of Statistics (2014) in which 24% of those aged 15-74 held a Bachelor degree or higher, the study’s sample was much more educated than the general population. The other socio-demographics were relatively comparable to the general population. Most participants were from a White European background (76%). The remainder were Asian, Indigenous, Indian, Middle Eastern, Pacific Islander, and African. Of those who stated their political orientation, 41% were left leaning, 26% were right leaning and 32% were centre. Just under half of the participants (44%) reported being Christian, 49% reported having no religion, and 7% reported having other religions.
Measures
Information concerning the participants’ gender, age in years, education level (1 = did not complete secondary school to 6 = higher degree) and their political orientation (1 = strongly left through to 5 = strongly right, or 6 = don’t care) was collected. Those who did not care about politics were coded as “missing data”. We also asked participants for their religion and ethnicity.
Two attitude thermometers were used to measure prejudice towards resettled refugees and asylum seekers. The prefacing question read, “In general how positive or favourable do you feel about resettled refugees/asylum seekers?”. Participants could respond from 0º (extremely unfavourable) to 100º (extremely favourable). These were reversed in order for them to measure negativity or prejudice rather than positivity.
Participants were then asked to use a check list (following Leach et al., 2006) to indicate the degree to which they experienced ten emotions when thinking about resettled refugees and asylum seekers. The adjectives were rated on a six-point Likert scale (0, not at all; 5, extremely). The following six adjectives were included to assess participants’ anger level: indignant, angry, annoyed, hostile, outraged, and furious. The first five anger items were taken from Leach et al. while furious was new to the present study. The following five adjectives were used to assess participants’ fear level: scared, frightened, fearful, afraid, frightened, and nervous. Two of the fear items (fearful, afraid) were taken from van Zomeren, Leach, and Spears (2010) and the other four were new to the present study. Higher scores indicated increased feelings of anger and fear.
There were 10 items measuring perceptions of threat regarding resettled refugees and asylum seekers. Four sub-scales of two items each were adapted from Cottrell et al. (2010): “threat to values” (e.g., promote values that directly oppose the values Australians hold); “threat to reciprocity due to unwillingness” (e.g., choose to take more from Australians than they give back); “threat to reciprocity due to inability” (e.g., need to take more from Australia than they are able to give back); and “threat to physical safety” (e.g., endanger the physical safety of Australians). While Cottrell et al. measured the perceived threat of different towards groups towards “people like me” in this study we changed this to “Australians”. Cottrell et al.’s fifth threat subscale related to threat to personal property; thus, a final subscale of two items each was substituted that related to “threat to Australian resources” (e.g., increase the tax burden on Australians, items adapted from Schweitzer et al., 2005). The 10 items were responded to on a Likert scale ranging from 1 to 7. After appropriate recoding, higher scores indicated increased threat perceptions.
Social policy attitude was measured by participants rating their level of support on a 1–7 Likert scale for four statements amended from Verkuyten (2004): “Resettled refugees/asylum seekers should have immediate access to all social services such as education and healthcare”; “Resettled refugees/asylum seekers should be given the chance to work as soon as possible”; “There is too much effort put into the care and support of resettled refugees/asylum seekers in the community” (reverse scored) and “The government needs to do more to prevent discrimination against resettled refugees/asylum seekers in the labour market”. After recoding, higher scores indicated a more restrictive policy attitude. An overall mean score for the four policy items relating to resettled refugees, and asylum seekers respectively was created. Last, participants were asked an open-ended qualitative question, “Finally, are there any other comments you would like to make to help us understand your views in more detail?” This question was purposefully broad so as to tease out whether participants spontaneously mentioned issues relating to resettled refugees, asylum seekers, or both.
Procedure
Participants were drawn from the Perth metropolitan area in July-August 2013 through convenience sampling through a team of research assistants. Participants were recruited in large group settings; for example, at social barbeques. Participants were told in the cover letter that “There is ongoing debate and a range of views about refugees and asylum seekers in Australia. For the purpose of our study, resettled refugees are those people flown from overseas as part of Australia’s refugee and humanitarian program, and asylum seekers are people who arrive in Australia by boat and then seek refugee status”. Participants were informed about the study and could either complete the questionnaire or not. There was a collection box out of sight of the team of research assistants so there was no pressure on people to participate (the research assistants were not aware of who took part). All participants were given the same questionnaire with the measures in the same order. Participants completed the socio-demographic questions followed by the prejudice thermometers, and then the questions relating to their perceptions of threat, emotion and policy attitude with reference to resettled refugees. The same questions were then repeated with reference to asylum seekers. The last section involved the open-ended question.
Results
Quantitative Data
Aim One: Compare Social Policy Attitude, Prejudice and Emotions for Resettled Refugees and Asylum Seekers
We investigated whether there were differences between the dependent variable (social policy) and the four predictor variables with respect to resettled refugees and asylum seekers. First, it was noted that there was high correlations between the five threat sub-scales (r’s between .63 and .80 with the refugee items; r’s between .63 and .82 with the asylum seeker items). As a result, we conducted a factor analysis on both sets of threat items. Both scree plots indicated the existence of one underlying factor. We then conducted an inter-item reliability analysis on both threat scales. After an inspection of the corrected item-total correlations (CITC), no items produced a CITC under the target value of .30 (indeed; all were over .55). As such, two modified scales were constructed and labelled as “Threat – Refugees” and “Threat – Asylum Seekers” which included all threat items; no items needed to be deleted from this scale to increase reliability.
Table 1 set out the descriptive statistics for each scale including means, standard deviations, alpha coefficients, the number of items, the range of scores and the t-tests statistics. Reliability was satisfactory for all scales (α = .79 to α = .97). All items were included in the scales except for both policy items where one item reduced reliability from α = .79 to α = .71 with the refugee policy scale and from α = .81 to α = .77 with the asylum seeker policy scale. In both cases, this item was removed. Scores were just below the midpoint on the two policy scales and threat scales with relatively high levels of prejudice towards asylum seekers (this being above the midpoint unlike prejudice towards refugees).
Table 1
Descriptive Characteristics of Scales
| Scale | Resettled refugee
|
Asylum seeker
|
t | df | p | d | No of items | Score range | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M | SD | α | M | SD | α | |||||||
| Policy | 3.02 | 1.71 | .79 | 3.88 | 1.94 | .81 | -7.31 | 184 | < .001 | 0.472 | 3 | 1-7 |
| Prejudice | 37.46 | 24.17 | - | 64.67 | 28.00 | - | -12.27 | 137 | < .001 | 1.040 | 1 | 1-100 |
| Threat | 3.18 | 1.49 | .93 | 3.81 | 1.66 | .94 | -7.47 | 181 | < .001 | 0.400 | 10 | 1-7 |
| Anger | 0.69 | 1.20 | .96 | 1.29 | 1.53 | .95 | -7.73 | 182 | < .001 | 0.432 | 6 | 0-6 |
| Fear | 0.62 | 1.07 | .96 | 0.78 | 1.19 | .97 | -3.10 | 182 | .002 | 1.430 | 6 | 0-6 |
To investigate the first aim, five t-tests were conducted for each target group. As detailed in Table 1, prejudice, threat, fear and anger levels were significantly higher for asylum seekers compared with refugees. Support for harsher policy was also significantly higher for asylum seekers compared with refugees.
We then conducted a correlation analysis of all the scales, political orientation and the socio-demographics (see Table 2). With respect to the asylum seeker and refugee scales outlined in Table 1, significant correlations ranged from r = .20 for refugee prejudice and asylum seeker fear, to r = .81 for refugee fear and asylum seeker fear (see the Appendix for the full correlation matrix). Support for more restrictive policies for refugees in the community was significantly correlated with prejudice against refugees, the perception that refugees were a threat, and high levels of fear and anger. Furthermore, older participants, those who held right-wing views, and those with less formal education were also more supportive of a harsher refugee policy. The same variables were found to be significant with the asylum-seeker correlations with the exception that education was not significant.
Table 2
Correlation Matrix
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |
|---|---|---|---|---|---|---|---|---|---|
| 1. Policy Scale | — | .50** | .63** | .61** | .58** | .19* | .04 | .24** | -.22** |
| 2. Prejudice | .53** | — | .43** | .41** | .22** | .14 | -.13 | .18 | -.25** |
| 3. Threat | .58** | .70** | — | .70** | .64** | .20** | .10 | .30** | -.28** |
| 4. Anger | .55** | .56** | .68** | — | .82** | .14 | .04 | .18* | -.27** |
| 5. Fear | .37** | .30** | .48** | .71** | — | .19* | .17* | .19* | -.27** |
| 6. Age | .20** | .25** | .28** | .26** | .29** | — | .03 | .22** | -.21** |
| 7. Sex | -.02 | -.01 | .06 | .03 | .17* | .03 | — | -.04 | -.04 |
| 8. Politics | .40** | .35** | .31** | .29** | .19* | .22** | -.04 | — | -.04 |
| 9. Education | -.11 | -.22** | -.26* | -.23** | -.21** | -.21** | -.04 | -.04 | — |
Note. Above the diagonal: Resettled refugees. Below the diagonal: Asylum seekers.
*p < .05. **p < .01.
Aim Two: Explore the Predictors of Restrictive Social Policy Attitude
We examined what variables predicted social policy attitudes with respect resettled refugees and asylum seekers. To examine the additional predictive power of emotions for policy attitude, over and above prejudice, two hierarchical multiple regressions were performed, one for resettled refugees and one for asylum seekers. Political position and significant socio-demographics were entered on the first step, prejudice and threat were entered on the second step, and anger and fear were entered on the third step (see Table 3). Participants who answered “don’t care” to the political view question were not included in the regression analyses. This involved a total of 17.8% of participants (33 out of 185) leaving a sample of 152 for the regression analyses.
Table 3
Hierarchical Regression Predicting Social Policy Attitude Towards Resettled Refugees and Asylum Seekers
| Variables entered | r | β
|
R2 Change | Total R2 | ||
|---|---|---|---|---|---|---|
| Step 1 | Step 2 | Step 3 | ||||
| Social policy attitude: Resettled refugees | ||||||
| Step 1 | .04 | .04 | ||||
| Age | .19* | .03 | -.01 | .01 | ||
| Political Position | .24** | .14 | -.02 | -.21 | ||
| Education | -.22** | -.10 | .03 | .09 | ||
| Step 2 | .34* | .38** | ||||
| Prejudice | .50** | .28** | .26** | |||
| Threat | .63** | .44** | .23* | |||
| Step 3 | .07* | .45** | ||||
| Fear | .58** | .26 | ||||
| Anger | .61** | .11 | ||||
| Social policy attitude: Asylum seekers | ||||||
| Step 1 | .16** | .16** | ||||
| Age | .20** | .07 | -.09 | -.09 | ||
| Political Position | .40** | .38** | .23** | .22** | ||
| Step 2 | .22** | .38** | ||||
| Prejudice | .53** | .32** | .27* | |||
| Threat | .58** | .24* | .13 | |||
| Step 3 | .03* | .41** | ||||
| Fear | .37** | .07 | ||||
| Anger | .55** | .23* | ||||
With respect to policy attitude regarding resettled refugees, age, political position or education did not have significant beta weights at step one and this was the case with steps two and three also. At step two, both prejudice and threat were significant; at step three, prejudice and threat remained significant. At step three, fear also had a significant beta weight. In other words, the most important variables for predicting more restrictive policy attitude towards refugee policy were high levels of prejudice, the perception of threat together with an increased fear of refugees. A significant amount of variance in more restrictive policy attitude towards refugees was explained by prejudice and threat at step two (R2 = .34) as well as by fear at step three (R2 = .07), with 45% of the overall variance explained.
With respect to policy attitude relating to asylum seekers, right-wing political position had a significant beta weights at all three steps. At step two, prejudice and threat both were significant; at step three, prejudice remained significant. Furthermore, at step three, anger had a significant beta weight. In other words, the most important variables for predicting more restrictive policy attitude regarding resettled refugees were high levels of prejudice together with high anger against refugees and a right-wing political attitude. A significant amount of variance in more restrictive policy attitude towards asylum seekers was explained by political attitude at step one (R2 = .16), prejudice and threat at step two (R2 = .22) as well as by anger at step three (R2 = .19), with 41% of the overall variance explained.
Qualitative Data
Aim Three: Explore What Issues and Subgroups Participants Mentioned in the Qualitative Data to Seek Convergence With the Quantitative Data.
To analyse the qualitative dataset, the first author conducted a thematic analysis in an inductive way (see Braun & Clarke, 2006). That is, the themes that were identified were strongly linked to the data, without trying to fit the data into a pre-existing coding frame. The data were analysed in five phases. The first four stages of the analysis were completed by the first author, using a dataset of all responses that had been transferred to a single word document by a research assistant, which had the corresponding participant code attached to each transcription. The first phase involved generating initial codes from the data to organise them into meaningful groups. In the second phase, the initial codes were collated into broader themes. The third phase involved reviewing the themes and the removal of any themes that overlapped or did not have enough data to support them. The final phase involved defining and naming the themes where the second author crossed checked the themes. To do this, attention was given to whether the excerpts in each theme fitted with the description of the theme. For excerpts that were seen to be unclear or counter-intuitive, both authors discussed the appropriate place for the excerpt.
The percentage of participants who left qualitative responses was 46% (N = 86). Of these responses, a total of 72 excerpts were coded into themes. The breakdown percentage of excerpts falling into each theme is detailed in Table 4. Despite the lower response rate than the quantitative questions, qualitative data from open-ended questions has been found to provide rich information from relatively few respondents (Geer, 1991). Indeed, as will be outlined below, the data provided complementary explanations for understanding the quantitative responses but also helped to enhance our understanding of quantitative findings. Three major themes emerged and are described in order of prevalence (see Table 4). Following the method adopted by Pedersen and Fozdar (2010), we analysed each theme separately.
Table 4
Key Themes From Qualitative Data (N Excerpts Coded Into Themes = 77)
| Theme 1: Asylum Seekers as ‘Illegals’ and ‘Queue Jumpers’ (42% of total themes) |
|---|
| Participants identified the method of arrival of asylum seekers as illegal and as ‘queue-jumpers’. Some also identified the method of arrival of asylum seekers as an example of how they ‘exploit’ Australia’s systems and processes. |
| Theme 2: Threat Posed by New Arrivals (30% of total themes) |
| Symbolic Threat. Participants expressed concerns that new arrivals were not willing to integrate into the Australian way of life and threatened the Australian culture. |
| Realistic Threat. Participants expressed concerns that new arrivals were a threat to the Australian economy, or may take Australian jobs. |
| Theme 3: Positive Comments Towards Resettled Refugees and Asylum Seekers (28% of total themes) |
| Empathy. Participants expressed their care and concern for resettled refugees and asylum seekers. |
| Human Rights. Participants positioned the act of seeking asylum as a human right. |
| Common Humanity. Participants highlighted the bond shared between Australians and resettled refugees and asylum seekers by virtue of being human |
The most prevalent theme was “Asylum seekers as ‘illegal’ and ‘queue jumpers’” which comprised of 42% of total themes coded. Within this theme, participants wrote about asylum seekers with regards to two broad issues. The first related to perceptions that asylum seekers who arrive by boat are ‘illegal’. For example, Participant No 4 (a 30 year old man) stated: “Those that come here illegally, expect everything to be handed to them on a silver plate, then cause problems & riots, jump up & down about our religion & political views should then be put on a plane & sent back to their own country”. Similarly, Participant No. 28 explains “People who use criminal acts to gain access to Australia, should not be welcome into the country. All illegal emigrants should immediately be sent back to their country of origin. If they have been able to pay $10,000 US dollars to get on an illegal boat then they are able to apply for genuine refugee status offshore”. The second issue related to descriptions of asylum seekers as ‘queue-jumpers’. For example, as Participant No 128 (a 23 year old woman) stated: “I disagree passionately with asylum seekers coming in by boat, as each time they ‘cue jump’[sic] those waiting in the refugee camps, often for 5+ years, get pushed to the back”. Similarly, Participant No. 83 states “I believe the animosity towards “boat people” does not stem from a vehement abhorrence towards people seeking asylum, rather in the fairness of people “jumping the queue” and entering Australia illegally.” Importantly, for most of the responses within this theme, some level of anger and indignation was expressed. For example, Participant No. 101 explains “I do feel some indignation at people who appear to be able to spend a lot of money to fly and then take a risky boat trip to reach Australian shores. My perception of these asylum seekers is that they are somehow “jumping the queue” and gaining an unfair advantage over other people who are waiting in refugee camps”.
The second most prevalent theme was “Threat posed by new arrivals” which comprised of 30% of the total themes. Only two respondents referred to asylum seekers in this theme with the remaining responses in made reference to refugees or migrants more generally. As such this theme was labelled threat posed by ‘new arrivals’ and is interpreted as indicating a more generalised concern about the threat posed by new arrivals to Australian society. Within this theme there were two sub-themes; the first related to symbolic orientated threats that involved concerns about a potential lack of integration of new arrivals. As Participant 22 (a 37 year old male) stated: My biggest issue with refugees is when they make no attempt to learn English or integrate into our society then take the freedoms this great country provides and use it to promote their own extremist views about how they think we should live. Some comments within this theme involved animosity towards Muslim immigrants specifically. For example, Participant No 98 (a 55 year old male) stated: “I was born in NZ but have spent most of my life in Australia. I am totally opposed to migration to this country by Muslims. I am not persay[sic] against migrants in general.” The second sub-theme involved realistic orientated threats that involved concerns that new arrivals are burden on the Australia’s economy. For example, as Participant 59 (a 50 year woman) stated: I believe that it is time the Australian Government STOP spending so much on these people and concentrate on Australia’s own trouble Indigenous people who struggle everyday to survive.”
The third most theme involved positive comments and comprised of 28% of the total themes. There were three subthemes; the first being empathy towards new arrivals. For example, as Participant No 39 (a 22 year old woman) said, “I think that (in a sweeping, generalised statement) Australians have an unjustified sense of entitlement. It appears that they have difficulty, or just don’t try, to imagine being in the situations of asylum seekers. I think that there is a great lack of understanding regarding these issues, partly because people don’t inform themselves, but also because the government tends to perpetuate a negative rhetoric concerning refugees.” The second subtheme involved evoking human rights in support for new arrivals. As stated by Participant No 57 (a 19 year old man) noted, “I believe every human deserves freedom & safety and if a privileged nation can provide this, it has an obligation to share it’. The final subtheme involved appealing to common humanity in support of new arrivals. As Participant 17 (a 60 year old man) stated: “I am a proud Australian, but we are one world, one humanity.”
Discussion
Through a mixed-methods community survey, this study examined the attitudes of members of the Australian community regarding social policies aimed at providing support and assistance to resettled refugees and asylum seekers who arrive by boat. The use of these methods allowed for the provision of complementary explanations for understanding the overall research question, with the qualitative data providing additional insights. At a global level, our findings underscore the differences in emotions and prejudice towards two groups of new arrivals to Australia. These findings reinforce previous research that points to the value in exploring ingroup attitudes, feelings and emotions beyond the broad label of ‘immigrant’ and focusing on specific subgroups (Kaiser & Wilkins, 2010; Murray & Marx, 2013). Our findings also reinforce previous research findings that resettled refugees and asylum seekers who arrive by boat to Australia are often constructed in oppositional ways by the media, politicians, and the general public (e.g., Klocker & Dunn, 2003; McKay et al., 2012), and adds to the small amount of research that has compared differences in prejudice towards these two groups (see Markus, 2012). In addition, our research builds on earlier work that examined the social psychological antecedents of community opinions about asylum seeker policy specifically (see Hartley & Pedersen, 2007).
The first of these antecedents is prejudice. Community members felt significantly more prejudice towards asylum seekers than resettled refugees. In addition, for both resettled refugees and asylum seekers, prejudice was an independent predictor of a more restrictive social policy attitude. Given the difference in levels of prejudice between the two groups, one could ask the question “is the animosity towards asylum seekers simply prejudice or racism?” However, our results indicate that the animosity goes beyond generalised prejudice. That is, prejudice alone cannot explain differences in attitudes towards the two groups (see also Markus, 2012 who makes similar claims).
The second of the antecedents is the influential, yet differential, role that emotions and prejudice play in predicting social policy attitudes. Specifically, fear and threat were independent predictors for more restrictive social policy relating to resettled refugees, while anger was an independent predictor for more restrictive social policy relating to asylum seekers. This is important as it suggests that emotions can play an important and different role in understanding outgroup reactions to specific subgroups, over and above measures of generalised prejudice. These findings contribute to a growing body of research that highlight the value in considering the differential roles emotions can play in predicting intergroup behaviour and attitudes (e.g., Cottrell et al., 2010; Iyer & Leach, 2009; Mackie et al., 2000).
That anger was an independent predictor of more restrictive social policy for asylum seekers but not for resettled refugees is particularly noteworthy. Anger is theorised to involve appraisals of wrongdoing (e.g., Iyer & Leach, 2009). It could be suggested that perceptions of wrongdoing are involved with shaping attitudes towards asylum seekers. While appraisals were not measured quantitatively in this study, the qualitative data offers insights as to what underpins higher levels of anger towards asylum seekers. The high proportion of responses that described asylum seekers as “illegal” or “queue jumpers” (42% of all coded themes) gives weight to quantitative findings that participants felt significantly higher levels anger towards asylum seekers. In particular, the qualitative responses in this theme carried tones of anger and disdain. The prevalence of these arguments regarding asylum seekers is in line with previous research. For example, in another Australian study that focused on resettled refugees only, one of the most prevalent themes emerging from the qualitative data related to false beliefs about asylum seekers – which was not the even focus of the study (Turoy-Smith, Kane, & Pedersen, 2013).
The results relating to fear and threat were also revealing. Our results found that while there were higher levels of fear and threat towards asylum seekers compared to resettled refugees, in the regression, fear and threat were independent predictors for a more restrictive social policy for resettled refugees only. Our qualitative data offer some additional insights in this regard. The second most prominent theme related to threat posed by new arrivals. While there were a few explicit mentions of asylum seekers, in general comments in this theme were concerned perceptions of threat (both symbolic and realistic) posed by refugees and migrants more generally. The qualitative data offers additional insights here, suggesting that fear and threat may be functionally more relevant when thinking about ‘new arrivals’ more generally who may be considered threats, as opposed to asylum seekers who arrive by boat and who may be perceived to have violated a moral code. The perceived threat highlighted in the qualitative data adds to previous quantitative research highlighting the importance of realistic and symbolic threat as important concepts underpinning behavioural orientations towards outgroups (e.g., Schweitzer et al., 2005; Stephan et al., 2005) and is in line with previous research that has found, Australian generally feel hostile towards immigration and new arrivals perceived as the ‘other’ (e.g., Muslims) (see Fozdar & Hartley, 2013).
Finally, that right-leaning political orientation predicted more restrictive social policy attitudes for asylum seekers but not for refugees is noteworthy. This points to the influence of the political environment and that asylum seekers who arrive to Australia by boat have been the target of much negative right-wing political and media attention (e.g., Klocker & Dunn, 2003).
In summary, the qualitative and quantitative data corroborate to provide an in-depth insight into the role of prejudice and emotions in social policy attitudes relating to resettled refugees and asylum seekers. First, when participants were explicitly asked about resettled refugees and asylum seekers in the quantitative section of the questionnaire, there were clear statistical differences in the level of felt emotions, prejudice, and support for social policy between these two groups. This suggests when the subgroup is explicitly defined, asylum seekers are viewed more negatively on a number of aspects, particularly with regards to feelings of anger. The qualitative data offers insights as to why anger may be so powerful. When participants were given the option to discuss any issues raised by the content of the questionnaire, the major theme that emerged related explicitly to asylum seekers and perceptions of illegality and ‘queue-jumping’. The results relating to resettled refugees, particularly regarding the predictive value of fear and threat for more restrictive social policy, and qualitative findings points a level of generalised outgroup hostility felt towards new arrivals perceived to threats Australians society, symbolically and realistically.
Practical Implications
The findings of this research present a number of implications. First, our findings reinforce the centrality of anger and of perceptions of illegality and wrongdoing of asylum seekers who arrive by boat to Australia in shaping policy attitudes. Since this research was conducted, the Australian Government has implemented a number of policies that perpetuate the idea that asylum seekers who attempt to arrive in Australia by boat are breaking the law. This includes the implementation of ‘Operation Sovereign Borders’, a military-led border security operation that includes turning asylum seeker boats on route to Australia back to Indonesia and Sri Lanka. It also includes the continued referral of asylum seekers as ‘illegal maritime arrivals’ in policy documents and media releases (Creek, 2014). This is concerning particularly in view of research that suggests political rhetoric is profoundly influential in shaping community attitudes (e.g., Pedersen, Watt, & Hansen, 2006).
Similarly, mainstream media reporting poses significant challenges for addressing negativity towards asylum seekers. As mentioned, the media often reinforces the false binary between asylum seekers and resettled refugees (e.g. Every & Augoustinos, 2008). This is despite the fact that the term ‘illegals’ with reference to asylum seekers has been deemed by the Australian Press Council (2012) as inaccurate and misleading. Consistent calls to account for journalists to use correct terms with reference to the Australian Press Council guideline may help to apply some form of accountability. On the other hand, innovative media campaigns have been used to challenge negative public discourse. This includes an award winning documentary television series, “Go back to where you came from”, in promoting six well-known Australians with vastly differing opinions on Australia’s asylum seeker debate were taken on a journey in reverse to that which refugees have taken. This series documents disturbing insights into life inside Australian immigration detention centres, immigration raids in countries en-route to Australia, and also in refugee camps. Such programs offer a potentially wider-reaching mechanism to foster positive changes in public debate by tapping into the emotions and attitudes of viewers, and challenging common negative misperceptions about asylum seekers and refugees.
Clearly, shifting negative attitudes towards asylum seekers requires a significant change in political rhetoric, social norms, and media reporting. How such change can be achieved structurally remains a significant challenge. Some research highlights the role that social identities, such as national identity or groups formed around a shared opinion, can play to promote pro-social collective action (for example, Druckman, 2006; Thomas, McGarty, & Mavor, 2009).
At the same time, individual change is also important and our findings do offer some practical implications in this regard. First, it is clear that prejudice and racism need to be addressed for both asylum seekers and resettled refugees. The fact that there is a difference between the perception of refugees and asylum seekers means that that anti-prejudice interventions have to take into account differences in prejudiced attitudes. Relatedly, anti-prejudice practitioners need to take into account the strong emotions involved in the attitude formation process. In an intervention, this cannot just involve giving people facts alone (e.g., that asylum seekers who arrive by boat are not illegal). Emotions, such as anger, fear and perceptions of threat need to be engaged as well. For example, with the fear and threat finding related to resettled refugees, interventions may seek to concentrate on the fact that integration into a host society is a two way process in that it requires a willingness for the host community to be welcoming and responsive to new arrivals (Hollands, 2001). With the finding relating to anger towards asylum seekers, some form of engagement with morality of ‘jumping the queue’ should be addressed and is likely that a person’s individual values are relevant here (see also Pedersen et al., 2008). Finally, while the way asylum seekers and resettled refugees are defined can influence more negative attitudes, emotional reactions, and support for restrictive social policy, as highlighted in our qualitative responses, the fact that these groups of people can be defined in positive ways also offers the possibility that more favourable attitudes can develop.
Limitations and Future Research
While this research highlights a number of important findings, there are some limitations that should be highlighted. First, our mixed-methods approach offers alternative insights that could not be obtained from other approaches; however, we acknowledge that the fact it is cross-sectional means that we cannot tease apart the social-psychological processes at play. Second, to minimise any priming issues, the more controversial questions relating to policy and the open-ended qualitative until the end of the questionnaire; however, because all items were placed in the same order we cannot rule out order effect. Future research may seek to address these limitations by exploring the studied variables in an experimental setting. Future research may also seek to explore the causal relationship between the variables; for example, the relationship between feelings of anger towards asylum seekers, perceptions of moral code violations, and social policy attitude. Also, because we were trying to keep the questionnaire as short as possible to maximise the response rate, we used a one-item measure of prejudice as per Pedersen and Barlow (2008). Future research should investigate the issues using more comprehensive measures. Finally, the sample had higher levels of education, compared to the general population and not all people left a qualitative response and thus responses may not be representative of the entire sample. Nevertheless, the responses provided rich information that gave additional insights over and above the quantitative data (Geer, 1991)
Conclusion
While attitudes toward resettled refugees versus asylum seekers are likely to affect the social policies implemented relating to their access to such support in the host society (Marr & Wilkinson, 2003), challenging negative attitudes requires a significant shift in political rhetoric, social norms and media reporting. Individual attitude change is also important. However, despite hard-line political discourse and media reporting relating specifically to asylum seekers, prejudice is not universally shared within the Australian community (Haslam & Holland, 2012). Previous research has found that negativity towards asylum seekers can be effectively tackled both in an educational setting (Pedersen, Paradies, Hartley, & Dunn, 2011) and in the community (e.g., Hartley, Pedersen, & Dandy, 2012). The challenge for researchers, refugee advocates and anti-prejudice practitioners is to continue to find ways to effectively engage with the emotions and attitudes of members of the community. An important avenue for this is to further understand the nuances in the perceptions of subgroups of new-arrivals to Australia.