Threatened or Threatening? How Ideology Shapes Asylum Seekers' Immigration Policy Attitudes in Israel and Australia.

Can different political ideologies explain policy preferences regarding asylum seekers? We focus on attitudes regarding governmental policy towards out-group members and suggest that perceptions of threat help to shape these policy attitudes. Study 1 compared public opinion regarding asylum policy in Israel (N = 137) and Australia (N = 138), two countries with restrictive asylum policies and who host a large number of asylum seekers; Study 2, a longitudinal study, was conducted during two different time periods in Israel-before and during the Gaza conflict. Results of both studies showed that threat perceptions of out-group members drive the relationship between conservative political ideologies and support for exclusionary asylum policies among citizens. Perceptions of threat held by members of the host country (the in-group) towards asylum seekers (the out-group) may influence policy formation. The effect of these out-groups threats needs to be critically weighed when considering Israeli and Australian policies towards asylum seekers.

Introduction

Over the past two decades, the world has witnessed a radical shift in public perceptions and political reactions to asylum seekers (Suhnan ). As the number of asylum seekers has risen, governments of all political leanings have implemented policies designed to deter asylum seekers from entering their countries for the purpose of seeking legal protection as refugees. The result has been an ongoing struggle to balance internal pressures for border control against international law, which aims to create a compassionate and humanitarian environment for individuals who find themselves unable to live safe, secure lives in their home countries (Gammeltoft-Hansen 2011).

An asylum seeker is ‘an individual who has sought international protection and whose claim for refugee status has not yet been determined’ (United Nations High Commissioner for Refugees (UNHCR) 2011: 3; UNHCR Convention and Protocol Relating to the Status of Refugees 2010). According to the UN International Convention on the Rights of Refugees (1951, 1967), an asylum seeker must be afforded protection from harm by a host country while their application is being considered (UNHCR). Until the early 2000s, a number of nations, including Denmark and the Netherlands, maintained flexible immigration laws that gave many asylum seekers access to a generous welfare system. Yet today, across Europe and other Western nations, asylum seekers are often viewed through the lens of the threats they might pose.

In the present study, we examine the role of political ideology (political position or political preference) and threat perception as predictors of different levels of support for exclusionary asylum policies. Building on past research that dealt with threat perceptions (Hartley and Pedersen 2015) and political ideology, the novelty of this research lies in its comparative aspect and the longitudinal study in Israel that was done before and during a period of war. We hypothesize that individuals who subscribe to a more right-wing political outlook will perceive asylum seekers as posing a greater threat than do their compatriots with a more left-wing outlook and, as a consequence, will have a greater tendency to support exclusionary asylum policies.

We examine this hypothesis in two studies. Study 1 employs a comparative framework, examining public opinion towards national asylum policy in two different contexts: among citizens of Tel Aviv, Israel, and Perth, Australia. We followed John Stuart Mill’s most different design (Mill 1970; Seawright and Gerring 2008). This design involves the selection of two cases that are similar in outcome, namely the support for exclusion of asylum seekers, but differ in many of other characteristics. We chose Israel and Australia as two countries that have many differences, but have experienced communal ‘angst’ about asylum seekers.

Study 2 was conducted during two different time periods in Israel: before and during the war between Israel and the Palestinian Authority in the summer of 2014. The relationship between political ideology and threat perception on exclusionary attitudes has been investigated before (Canetti-Nisim and Pedahzur 2003; Pedersen ). However, to the best of our knowledge, only a few studies (e.g., Hartley and Pedersen 2015) have specifically assessed attitudes towards asylum policy in Australia and we are certainly the first to do so in Israel. Both studies complement each other as we are investigating whether our suggested mechanism works similarly across countries (Study 1) and during a period where attitudes towards out-groups might change (i.e. a time of war). Findings from these two different tests might validate our results even more if we show that threat perceptions mediate the relationship between political ideology and support for exclusionary asylum policy regardless of where or when.

Asylum Seeking within the Context of Attitudes towards Immigrants and Immigration

Given the broad theoretical and empirical overlap between immigration and asylum seeking, we draw upon the literature about attitudes towards immigration in constructing our theoretical framework. There is a common tendency to refer to attitudes towards immigrants and towards immigration policy together, rather than as two distinct domains (Ceobanu and Escandell 2010). As Ceobanu and Escandell argue: Focusing on the individual’s attitude towards an asylum policy, we aim to contribute to the body of research that deals with the important connection between the individual and decision-making in intergroup relations.

two broad types of … public views can be identified: attitudes toward immigrants and attitudes toward immigration. The two constitute distinct domains of study, despite the tendency of some scholars to address them together …. In instances when the two types of attitudes are approached as distinct objects of study, the literature illustrates a heightened interest in public views toward immigrants and a comparatively diminished focus on reactions to immigration (Ceobanu and Escandell 2010: 313).

Attitudes towards immigrants or immigration and asylum seekers were found to be likely influenced by group dynamics (e.g. group conflict theory; Meuleman ), societal norms and individual traits that help to mould the effects of intergroup contact on personal attitudes (Pettigrew 1998). Socio-demographic factors were examined (Wilkes ), as well as individuals’ deeply held values and beliefs. These societal and individual factors may contradict or reinforce one another. Australian scholars found that government rhetoric has employed stereotypical representations of asylum seekers as dishonest and opportunistic (Every and Augoustinos 2007)—a characterization that is readily accepted by citizens who are predisposed to be prejudiced against individuals unlike themselves. A number of studies have addressed the role of personal prejudice in people’s willingness to accept false beliefs about asylum seekers (Pedersen ; Suhnan ).

One area in which group norms and individual characteristics meet and reinforce one another is the realm of political ideology. In this work, we examine the role of political ideology in support for an exclusionary asylum policy.

Political Ideology and Attitudes towards Immigrants and Asylum Seekers

Political ideology constitutes a coherent worldview, derived from one’s culture, which can provide a sense of meaning in the face of individual and collective threats (Milner and Tingley 2010). A substantial literature on intergroup relations and immigration attests that people’s attitudes towards immigrants and immigration tend to be influenced by their political orientations (Semyonov ). The pattern emerging from this literature suggests that people who identify with a right-wing ideology value tradition and conformity, are more internally focused and are more likely to adopt anti-immigrant attitudes and support restrictive immigration policies. Those identifying with a left-wing ideology have greater tolerance for out-groups, a dislike of hierarchy and authority, and high empathy, and therefore are more likely to welcome immigrants and favour lenient immigration policies (McLaren 2001; Semyonov ; Semyonov and Glikman, 2009; Mudde and Rovira Kaltwasser 2013).

Asylum seekers form a particular subgroup of immigrants. Research focused specifically on this category of migrants has found similar patterns with respect to citizens’ attitudes. Studies in Australia have found that age, gender, education, socio-economic status and political views all influence attitudes towards asylum seekers, with the most negative views being held by older males with less formal education, low socio-economic status and right-wing political views (McKay ). Yet, the unique characteristics of contemporary asylum seeking calls for a deeper understanding of this political phenomenon and its implications for political behaviour. These complex factors include the large number of asylum seekers, their lack, in most cases, of official authorization, the indefinite nature of their claims, the fact that most will require long-term assistance from the state, their visible differences from native-born citizens, and the difficulty, in many cases, of accurately distinguishing between those truly needing asylum and individuals seeking to immigrate for personal or economic reasons.

Mediation Hypothesis: Threat Perceptions

Most scholars regard threat perceptions as the single best predictor of hostile out-group attitudes (Canetti-Nisim ; Ben-Nun Bloom ). Contemporary social psychology theories, such as integrated threat theory (Stephan and Stephan 2001), distinguish between two sets of threats that promote negative attitudes towards out-groups (Corenblum and Stephan 2001). Realistic threats relate to tangible objects, such as money, jobs, land or human life (McLaren 2001). These are resources the in-group members perceive as being threatened as a result of new competition with out-groups. Theoretical models (Semyonov , 2008) suggest that individuals with low socio-economic status tend to adopt more negative attitudes towards out-group members, since they feel most directly threatened by competition for resources. Symbolic threats relate to relatively abstract features of the in-group, such as its sense of identity and its system of values or beliefs, as expressed in its language, religion or moral code (Semyonov ). Asylum seekers are perceived as a threat to the culture of a given society (Raijman and Semyonov 2004). Symbolic threats can often be more emotive than realistic threats. Researchers in Australia found that the most common reason given for negative attitudes towards asylum seekers was their perceived violation of the country’s values (Pedersen ).

The accumulation of individual opinions may lead to a collective perception that one’s nation or group is threatened, which predicts support for exclusionist policies (Canetti-Nisim ). Perceived threats from immigrants predict popular support for punitive or restrictive immigration policies (Coenders and Scheepers 2003). In Israel, threat perceptions have yet to be examined in regard to attitudes towards asylum seekers, but they have been found to affect the Jewish public’s attitudes towards other minority groups (Canetti-Nisim and Pedahzur 2003). Given Israel’s situation of prolonged conflict, studies have also examined the role of threat perceptions among individuals exposed to political violence. Those studies showed that exposure to political violence is an important antecedent of threat perceptions, and that these perceptions mediate the relationship between exposure to violence and hard-line or exclusionist attitudes (Canetti-Nisim ). Canetti-Nisim, Ariely and Halperin (2008) found that many Israeli Jews perceive Israel or its values as under threat even from minority groups that have traditionally lived peacefully within the state. Those studies offer only partial support for the present hypothesis since, in the case of exposure to violence, perceived threat is, at least partially, a reflection of actual threatening occurrences (Canetti-Nisim ).

Relying on the large body of migration literature as well as previous research regarding asylum seekers, we hypothesized that threat perceptions would mediate the relationship between political ideology and attitudes towards asylum policy, such that people who are right-wing will (i) perceive asylum seekers as threatening and (ii) support exclusionary asylum policies to a greater extent than people who are left-wing. We further predicted that this mechanism would function identically in Israel and Australia (Study 1) and at times of both relative peace and during violent conflict in Israel (Study 2).

Setting

Both Israel and Australia are developed Western states and signatories to the Refugee Convention, they have a high proportion of immigrants, allow immigration and were founded by immigrants. While Australia is an immigration society potentially offering integration to all nationalities and ethnicities, Israel is an immigration society based on ethnicity, which offers integration only to Jews (Markus and Semyonov 2010). Both are social welfare countries, yet the benefits of the Israeli social system are granted only to Jews. The list of differences between the countries is long and varied. However, with regard to the asylum phenomenon, both experience asylum seekers who differ in visible ways from the country’s majority population, and both have responded to the asylum phenomenon with relatively harsh and restrictive policies.

Asylum Seekers in Australia

Thousands of people have sought asylum in Australia over the past decade (Siegel 2013). Australian society has seen three major waves of asylum seekers since 1976: the first from 1976 to 1981, the second from 1989 to 1998, and finally the current wave, which began in 1999 many of who are from a Middle Eastern origin. Over the years, successive Australian governments formulated strict asylum policies, including, from 1994 through 2008, mandatory detention for all individuals entering the country by boat without official authorization. After the Tampa Affair in 2001, Australian asylum policy became even harsher, with greater deterrence measures and fewer legal rights for those arriving in Australia. These changes fell within the framework of a policy called the Pacific Solution (Mares 2002), whereby Christmas Island and other islands were excluded from Australia’s migration zone, meaning that asylum seekers arriving in these territories could not automatically apply to the Australian Government for refugee status and would not have access to Australian courts. This process subjected thousands of asylum seekers to indefinite detention.

Polls and studies conducted over the years have found that many Australians hold overtly negative attitudes towards asylum seekers. Findings testify to strong prejudice against the asylum seekers, accompanied by threat perceptions and the acceptance of false beliefs concerning the refugees’ goals, attitudes and behaviour (Pedersen ; Croston and Pedersen 2013).

Asylum Seekers in Israel

Between 2005 and 2013, approximately 64,000 migrants entered Israel (Population and Immigration Authority, Israeli foreign data, 2012). Israel refrained from deporting both Sudanese and Eritrean nationals back across the border because of the dire situation in their home countries, and because of the UNHCR’s designation of Eritreans as a group in need of temporary humanitarian protection. At first, Israel automatically granted all Sudanese and Eritrean migrants temporary permission to stay in the country, but generally did not give them work permits or access to social benefits. By 2012, tens of thousands of asylum seekers were living in the poorer neighbourhoods of south Tel Aviv and in southern Israel.

Initially, the migrant phenomenon was accepted with indifference by the Israeli public and the government but, as their presence became more conspicuous in particular areas, local tensions began to surface and the issue entered the public agenda.

From 2010 through 2013, the government’s response to the crisis was aimed chiefly at reducing the number of new asylum seekers coming into Israel. The main components of the policy were the construction of a 240-kilometre barrier along the Israel–Egypt border and the internment of many asylum seekers in detention centres (fines were imposed on businesses that employed migrants). In addition, about 4,500 migrants from South Sudan and other African nations were deported in 2012. The new policy led to a substantial reduction in the number of asylum seekers entering Israel (from about 10,000 in 2012 to fewer than 40 in the first half of 2013). By 2013, Israeli asylum policy was able to focus on finding ways to deal with the social, political and legal status of those who were already in the country. Currently, Israel permits the detention of asylum seekers in semi-open detention centres for up to one year, though large numbers of migrants are still living in urban centres, and particularly in impoverished neighbourhoods of Tel Aviv, which offers them access to education and welfare services.

The Israeli–Palestinian Conflict as Context

The arrival of tens of thousands of foreigners, different in race, language and culture, and from predominantly Muslim countries, added tension to an already complicated political landscape. Study 2 was conducted during a period of war (which lasted from early July to the end of August 2014). During the fighting, militants in Gaza launched a relentless barrage of rocket attacks that potentially threatened all Israeli communities within geographical range, up to and including Tel Aviv (see Figure 2). To convey a sense of the stressors facing Israelis during that time, 4,600 air raid sirens sounded somewhere in Israel over that period, for an average of 29 per day in Tel Aviv (Rotter Net 2014).

Figure 2

Range of missiles launched from Gaza Strip.

Hypothesized model

Empirical Strategy

We study the effects of political ideology on policy attitudes through the mediation of threat perceptions in the context of asylum seekers. We employ a unique combination of two distinct yet complementary designs. Study 1 consists of a survey conducted in Perth, Australia, and Tel Aviv, Israel, in mid-2010 and early 2014, respectively. Study 2 is based on the baseline Israeli data included in Study 1, along with data from a second survey conducted in Israel several months later, during Operation Protective Edge. Study 2 uses a longitudinal panel design (Gerhart ) which tests for changes in attitudes over a period in which the country went from relative quiet to violent conflict, to evaluate whether the overall sense of threat varies within participants and between time points (though the number and status of asylum seekers remained fairly stable during that time).

The two studies provide a unique fusion of external validity (through the comparative framework of Study 1) and internal, ecological validity (through the panel design of Study 2). The characteristics of the two studies are summarized in Table 1.

Table 1

Characteristics of Study 1 versus Study2

Study 2	Study 1
Two samples: Before and during Operation Protective Edge (both Tel Aviv, Israel)	Two samples: Perth, Australia, and Tel Aviv, Israel
Panel design	Cross-sectional design
Data collection: March 2014 (Wave 1), August 2014 (Wave 2)	Data collection: 2010 (Australia), 2014 (Israel)
N = 94 (final sample for both waves)	N = 274 (138 in Australia and 137 in Israel)
High internal validity	High external validity

Study 1

Method: Sample and Procedure

For Study 1, a structured internet survey measuring political ideology, threat perceptions and attitudes towards the government’s policy regarding asylum seekers was administered in two different contexts: Perth, Western Australia, and Tel Aviv, Israel. The Australian survey was conducted between June and August 2010, using a questionnaire developed based on previous research in that country (Suhnan ). Participants were drawn from SCORED (the Social and Community Online Research Database), which is maintained by researchers at Murdoch University in Perth. In Israel, the survey was translated into Hebrew using a standard forward- and back-translation procedure. The Israeli survey was administered in March 2014 to participants randomly drawn from the Midgam Panel, a large online Israeli survey pool. Both studies were described to participants as surveys about attitudes regarding asylum seekers.

Australian sample

One hundred and thirty-eight Australians who resided in Perth took part in this study (59.4 per cent women). The participants ranged in age from 20 to 72 (mean age = 40.56, SD = 14.03). Eighty-two of the participants (59.4 per cent) reported being left-wing and 27 (19.6 per cent) reported being right-wing (see under ‘Measures’ below).

Israeli sample

One hundred and thirty-seven Jewish Israelis who resided in Tel Aviv took part in this study (51.8 per cent women). Participants ranged in age from 23 to 64 (mean age = 34.57, SD = 8.47). Forty-nine participants (35.8 per cent) identified as left-wing and 35 (25.5 per cent) as right-wing.

Measures

Three main variables were tested: a predictor variable (political ideology), a mediator (threat perceptions) and the dependent variable (attitudes towards the government’s asylum policy).

Predictor variable: political ideology

Participants reported their ideology using a single item that assessed their subjective political position (‘How would you describe your political stance? Extreme right, right, centre, left, extreme left’). Answers were coded into three groups: 1 = right, 2 = centre, 3 = left. We did not explain the terms ‘right’ and ‘left’, but assumed respondents’ interpretations would reflect the general conventional understanding of those terms.

Predictor variable: threat perceptions

Perceptions of threat were measured using a set of questions adapted from previous research (Suhnan ) and tailored to relate specifically to asylum seekers. This scale originally included two subscales containing four realistic threat items and four symbolic threat items (two of the items were deemed irrelevant to circumstances in Israel and therefore excluded from the Hebrew version). Answers were given on a seven-point scale ranging from 1 = not threatening to 7 = very threatening. Sample items include ‘The values and beliefs of asylum seekers regarding family issues and socializing children are compatible with the values and beliefs of most Australians’ (reverse scored); ‘The quality of social services available to Australians has remained the same, despite asylum seekers coming to Australia’ (reverse scored). Following a preliminary analysis which revealed a substantial correlation between the symbolic and realistic threat scales (r = 0.762; p < 0.001), we created a single threat perception scale by averaging participants’ scores for all eight (Australia) or six (Israel) items. Following the initial analysis, we factor-analysed all the threat items together. The obtained scree plot clearly pointed to the presence of one underlying factor. An inspection of the corrected item-total correlations (CITC) revealed no items with a CITC under the target value of 0.30. We therefore treated all the threat items as a single scale. Cronbach’s alpha was 0.63 for the Israeli sample and 0.87 for the Australian sample.

Dependent variable: attitudes towards asylum policy

Attitudes towards the government’s policy regarding asylum seekers were measured using one question: ‘Overall, do you think the Federal Government’s, or in the Israeli context, Netanyahu government’s stance toward asylum seekers is too lenient, too harsh, or fairly balanced?’ (Nickerson and Louis 2008). Higher scores indicated that the respondents think the policy is too harsh.

Covariates

These included gender (male = 0, female = 1) and age (years). Past research showed that men and older people had more conservative views (Semyonov ).

Results

Descriptive statistics

The means, standard deviations and intercorrelations among the study variables are displayed in Table 2. Participants in both samples reported relatively high levels of exclusionary asylum policy attitudes. Sample t-tests showed that such attitudes were higher among the Australian participants compared to the Israelis (M = 3.90, SD = 2.12; M = 3.84, SD = 1.85, respectively). Sample t-tests also showed significant differences in levels of perceived threat between the two samples, with the Israeli respondents reporting higher levels of threat than the Australians (M = 4.10, SD = 1.05; M = 3.52, SD = 1.46, respectively).

Table 2

Means, Standard Deviations and Intercorrelations among Study Variables (Study 1)

		M	SD	1	2	3	4	5
1.	Political ideology 1 = right to 3 = left	2.10	0.79	1.00
		2.40	0.79	1.00
2.	Threat perception	4.10	1.05	–	1.00
		3.52	1.46	0.48***	1.00
				–
				0.46***
3.	Policy attitude	3.84	1.85	–	0.60***	1.00
		3.90	2.12	0.57***	0.74***	1.00
				–
				0.50***
4.	Age	34.57	8.47	0.09	–0.08	–0.16	1.00
		40.56	14.03	0.12	0.08	–0.05	1.00
5.	Gender (female)	1.49	0.50	–0.19	0.14	0.02	0.09	1.00
		1.40	0.49	–0.08	0.00	0.07	0.01	1.00

p < 0.05;

p < 0.01;

p < 0.0001. First row = Israel, second row = Australia.

Table 2 shows the bivariate correlations between all the variables assessed in this study. For both samples, negative correlations were found between threat perceptions and liberal political ideology (r = –0.48, p < 0.001; r = –0.46, p < 0.001, for the Israeli and Australian samples, respectively) and between exclusionary policy attitudes and liberal political ideology (r = –0.57, p < 0.01; r = –0.50 p < 0.001, respectively). A strong correlation was also found between exclusionary policy attitudes and threat perceptions (r = 0.60, p < 0.01; r = 0.74, p < 0.001). There was no relationship between political ideology, threat perceptions or policy attitudes and age and gender.

Mediation analysis

We hypothesized that threat perceptions would mediate the relationship between political ideology and attitudes towards government policy on asylum seekers. We conducted a path analysis to test our mediation hypothesis. The results of these analyses are shown in Table 3 and in Figures 3 and 4.

Table 3

Mediation Effect of Threat Perceptions on the Relationship between Political Ideology and Policy Attitudes (Study 1)

		SE	B
				Threat perception
		0.10	–0.63***	Political ideology
		0.14	–0.87***
				Policy attitude
		0.12	0.74***	Threat perception
		0.09	0.94**
		0.17	–0.91***	Political ideology
		0.17	–0.50*
Bootstrap results for mediation effects; 95% confidence interval (CI)
Upper	Lower	SE	B	Mediation effect: threat perception
–0.28	–0.70	0.10	–0.47	Indirect effect of X on Y
–0.51	–0.1.17	0.17	–0.82

p < 0.05;

p < 0.01;

p < 0.0001. First row = Israel, second row = Australia.

Figure 3

Structural model for Israeli sample (Study 1).

Figure 4

Structural model for Australian sample (Study 1).

As the figures show, after accounting for the mediating effect of threat perceptions, the total effect of political ideology on policy attitudes fell in the Israeli sample from B = –1.35, SE = 0.17, p < 0.001 to B = –0.90, SE = 0.17, p < 0.015, and in the Australian sample from B = –1.32, SE = 0.20, p < 0.001 to B = –0.50, SE = 0.17, p < 0.001. The total indirect effect in both cases is significant. The analysis in both samples shows that a person’s political ideology can explain support for an exclusionary policy, but that this relationship is partially accounted for by threat perceptions. This finding supports our hypothesis that perceptions of threat mediate the relationship between political ideology and policy attitudes.

To determine whether factors specific to one or the other country influence this mechanism, we tested these results by conducting a moderated mediation analysis (utilizing PROCESS model no. 15, available online at http://www.processmacro.org) with the country as moderator (1 = Israel, 2 = Australia). We ran this model on the relationship between threat perceptions on policy attitudes and again on the relationship between political ideology and policy attitudes. The interactions were not significant, supporting our hypothesis that ideology influences attitudes through the mediation of threat perceptions regardless of which country the individual comes from (see Appendix).

We used Preacher and Hayes’s (2008) method of calculating standard errors and 95 per cent confidence interval (CI) values to examine the effect of political ideology on policy attitudes through threat perceptions on the Israeli and Australian samples. We used 1,000 bootstrapped samples to estimate the bias-corrected and accelerated confidence intervals. The results of these analyses confirm that threat perceptions mediate the relationship between political ideology and support for exclusionary policies (mediated effect: Israel, B = –0.47, SE = 0.10, 95 per cent CI = –0.70, –0.28; Australia, B = –0.82, SE = 0.17, 95 per cent CI = –1.17, –0.51) (see Table 3). Specifically, the effect size was –0.98 in the Israeli sample and –1.68 in the Australia sample, and the 95 per cent CI values did not include zero. These results support the hypothesis that threat perceptions partially explain the relationship between political ideology and policy attitudes towards asylum seekers.

Study 2

Study 1 was conducted during a period of relative calm in the Israeli–Palestinian conflict. However, intergroup conflicts can swiftly shift from relatively uneventful times to periods of escalation and open violence. The second wave of Study 2 was conducted during the conflict in August 2014 known in Israel as Operation Protective Edge. Study 2 thus follows a panel design in which we examine our model within the same subject group during two different time periods (referred to below as the March and August samples). Specifically, we sought to evaluate whether our sample would express similar attitudes during both of these periods, while also evaluating the possibility that the conflict might increase the strength of the association between threat and exclusionary policy attitudes in the second wave, during a period of heightened threat from other sources.

Through the Midgam Panel, we were able to re-establish contact in August with all original participants from the March 2014 survey and 94 respondents agreed to complete the second survey (69 per cent of the original respondents) and were sent a similar structured internet survey.

As described above, the respondents were all Jewish Israelis residing in Tel Aviv. Half the participants in the second sample (47 respondents) were women. Participants ranged in age from 23 to 64 (mean age = 34.57, SD = 9.56). In terms of their political outlook, 34 participants (36.2 per cent) identified as left-wing and 25 (26.6 per cent) as right-wing. An examination of the demographic and political indicators found no differences between those participants in Wave 1 who took part in Wave 2 (i.e. the final sample) and those who did not.

Measures

The tools used to assess political ideology and threat perceptions (the predictor and mediator variables) were identical to those used in Study 1.

Dependent variable: Attitudes towards asylum policy

In Study 2, this variable was measured using a scale of 12 questions. (The full set of 12 items was included in the Israeli survey in Study 1, but was not analysed for that study so that the scales for the Israeli and Australian samples would be consistent.) Seven of the 12 items related to specific Israeli government asylum policies and legislative rulings, and the remaining five to asylum policy in general. Two of these questions were drawn from the Eurobarometer—a set of public opinion surveys conducted by the European Commission, which periodically surveys European Union residents about their attitudes towards immigration and asylum seekers (among many other topics). Cronbach’s alpha for this scale was 0.76. Example items include ‘The right of asylum is a fundamental human right’ and ‘Border controls should be abolished throughout the world’.

Covariates

These included gender (male = 0, female = 1) and age (years).

The means, standard deviations and intercorrelations among the study variables are displayed in Table 4. Importantly, rather than showing more hard-line attitudes in the second wave (during a period of heightened threat), participants reported relatively high levels of exclusionary attitudes towards asylum policy in both waves (M = 4.20, SD = 0.99 and M = 4.06, SD = 0.98 in the first and second waves, respectively). A paired samples t-test revealed no significant differences between the two samples regarding the study variables.

Table 4

Means, Standard Deviations, and Intercorrelations of Political Ideology, Threat Perceptions, Policy Attitudes, and Control Variables (Study 2)

			M	SD	1	2	3	4	5
1.	Political ideology 1 = right to 3 = left		2.10	0.80	1.00
			2.11	0.80	1.00
2.	Threat perception		4.20	1.06	–	1.00
			4.20	1.46	0.48***	1.00
					–
					0.50***
3.	Policy attitude		4.20	1.00	–	0.62**	1.00
			4.06	1.00	0.54***	*	1.00
					–	0.65**
					0.54***	*
4.	Age		34.46	8.46	0.08	–0.05	–0.02	1.00
			34.46	8.45	0.02	–0.04	0.05	1.00
5.	Gender (female)		1.51	0.50	0.20*	–0.14	–0.10	0.09	1.00
			1.51	0.50	–0.03	–0.17	–0.10	–0.10	1.00

p < 0.05;

p < 0.01;

p < 0.0001. First row = March 2014, N = 137; second row = August 2014, N = 93.

Table 4 shows the bivariate correlations between all the assessed variables. A negative correlation was found in both samples between threat perceptions and political ideology (r = –0.48, p < 0.001; r = –0.50, p < 0.001 for the first and second waves, respectively) and between policy attitudes and political ideology (r = –0.54, p < 0.01; r = –0.54, p < 0.001, respectively). A strong correlation was also found between policy attitudes and threat perceptions (r = 0.62, p < 0.01; r = 0.65, p < 0.001 for the first and second waves, respectively). A significant correlation was found between gender (female) and political ideology in the first sample, but not in the second.

As in Study 1, we conducted path analysis to test our hypothesis that threat perceptions would mediate the relationship between political ideology and policy attitudes. As shown in Figures 5 and 6, accounting for the mediating effect of threat perceptions reduced the total effect of political ideology on policy attitudes from B = –0.67, SE = 0.09, p < 0.001 to B = –0.39, SE = 0.09, p < 0.001 in the March sample, and from B = –0.67, SE = 0.11, p < 0.001 to B = –0.36, SE = 0.11, p < 0.001 in the August sample (see Table 5). The total indirect effect is significant, meaning that threat perceptions mediate the effect of political ideology on policy attitudes.

Figure 5

Structural model for March 2014 sample (Study 2).

Figure 6

Structural model for August 2014 sample (Study 2).

Table 5

Mediation Effect of Threat Perceptions on the Relationship between Political Ideology and Policy Attitudes (Study 2)

		SE	B
				Threat perception
		0.11	–0.64***	Political ideology
		0.12	–0.68***
				Policy attitude
		0.08	0.44***	Threat perception
		0.08	0.45***
		0.09	–0.39***	Political ideology
		0.11	–0.36***
Bootstrap results for mediation effects; 95% confidence interval (CI)
Upper	Lower	SE	B	Mediation effect: threat perception
–0.18	–0.43	0.06	–0.28	Indirect effect of X on Y
–0.15	–0.48	0.08	–0.31

p < 0.05

p < 0.01

p < 0.0001. First row = March 2014, second row = August 2014.

As before, we used Preacher and Hayes’s (2008) method of calculating standard errors and 95 per cent CI values to corroborate the effect of political ideology on attitudes to asylum policy through threat perceptions across the two waves of data. We used 1,000 bootstrapped samples to estimate the bias-corrected and accelerated confidence intervals. The results of these analyses are shown in Table 5. The effect size was –0.61 in the March sample and –0.63 the August sample, with a 95 per cent CI value which did not include zero. For this reason, we can conclude that threat perceptions help explain the relationship between political ideology and attitudes towards asylum policy.

Discussion

The current research addressed the question of how threat perceptions affect the construction of preferences towards asylum policy. The findings uncovered, across two studies, a mediating effect of threat perceptions on the relationship between political ideology and exclusionary attitudes towards asylum seekers. The robustness of the effect is especially convincing as the two studies differed in design: Study 1 compared policy attitudes and threat perceptions in two contexts, namely between Israel and Australia, while Study 2 used a panel design to examine attitudes and perceptions among Israelis over two time periods and against two different contextual backdrops. Our findings suggest that, in all three settings examined (Tel Aviv, Israel, at a time of relative peace and at a time of open conflict; and Perth, Australia), a right-wing political ideology predicts support for an exclusionary policy towards asylum seekers, with threat perceptions being the driving mechanism behind such attitudes.

In terms of the comparative framework (Study 1), one of the sparks for the current research was the puzzling phenomenon of widespread hostile attitudes towards asylum seekers, and a concomitantly harsh government policy, in both Israel and Australia, despite numerous differences between the two countries. Our findings from both samples confirmed our hypothesis that people who identify as right-wing are likely to perceive higher levels of threat from asylum seekers and to support a more exclusionary policy towards them, and that threat perceptions help explain the relationship between political ideology and these policy attitudes. However, we found that, although levels of perceived threat were high in both samples, they were higher among the Israeli respondents compared to the Australians, while the latter reported significantly greater support for exclusionary policies.

Several factors might help explain these intriguing findings. First, looking specifically at the two samples (residents of Tel Aviv and Perth), it is clear that the former are personally exposed to asylum seekers far more than the latter. In Tel Aviv, where tens of thousands of asylum seekers live in all areas of the city, their large numbers combined with their appearance make them highly visible. In Perth, by contrast, the majority of asylum seekers are restricted to a detention centre or dedicated housing. Most residents of Perth only encounter asylum seekers through the media and political discourse (Johnson ; Pedersen ). It thus stands to reason that Israelis living in Tel Aviv would perceive asylum seekers as a greater threat than Australians living in Perth. This could also be due to the omnipresent threat of hostile border states. However, these circumstances cannot account for the greater support for exclusionary policies among our Australian sample.

Historical and societal factors may take us further in resolving this puzzle. With regard to the Israeli findings, a combination of Jewish history (particularly the Holocaust), the prolonged Israeli–Palestinian conflict and the continuing rejection of Israel’s legitimacy by other states in the region mean that many Israeli Jews dwell in a chronic state of anxiety, and even a sense of living under existential threat (Bar-Tal 2000; Gordon and Arian 2001). These feelings can at times give rise to a ‘siege mentality’—a core societal belief that other groups have negative intentions towards the focal group, which stands alone in a hostile world (Bar-Tal 2000). The asylum seekers, a new minority group in Israel, may be perceived by Israeli Jews as presenting threats on several levels: economic (in that they can be expected to compete for unskilled jobs), social (in that they are blamed for rising crime rates) and national security (in that many of them are Muslim and, as such, they are feared to constitute an immediate realistic threat connected to the Israeli–Palestinian conflict). The combination of existential fear among Israeli Jews and the unique characteristics of the asylum seekers thus offer a convincing explanation for the higher levels of threat reported by the Israeli sample. The same history that helps make Jewish Israelis feel under threat—particularly the Holocaust and other experiences of persecution, which made many Jews into refugees themselves—may also serve to counter the natural tendency to respond to feelings of threat by shutting the gates on others.

What about the exclusionary attitudes among the Australian sample? The asylum phenomenon is not new to Australian society, and negative attitudes towards asylum seekers (along with harsh asylum policies) have been present for years. Since the late 1990s, a number of Australian researchers have investigated hostility to refugees and support for exclusionary policies in relation to models of prejudice and racism (Johnson ; Pedersen ). According to this literature, in a country where about 90 per cent of the population traces their ancestry to white Europeans, asylum seekers—who are largely non-white—are positioned as threatening via a discourse in which foreigners are said to undermine the national culture and Australian way of life (Hage 2003). According to Hage (2003), hostility to immigrants in Australia represents a ‘paranoid nationalism’ among white Australians who feel ideologically and economically marginalized by globalization (Crock 2004; Louis ). These characteristics of Australian society may help explain the high support for exclusionary asylum policies among our Australian sample due to the lack of personal encounters with asylum seekers.

With respect to Study 2, the unstable security situation in Israel allowed us an opportunity to test our hypothesis amidst a military conflict. In the post-9/11 world, the relationship between exposure to political violence and political attitudes is a frequent topic of investigation, and several studies have pointed to a strong link between exposure to violence, threat perceptions and hard-line or uncompromising political attitudes (Hobfoll ; Canetti-Nisim ). At the time of our second survey, Tel Aviv was within range of rockets being launched from the Gaza Strip, and residents of the city were continually alert for air raid sirens. Israel is a small country with a near-universal draft, therefore it could be expected that many of our participants felt under threat because of concern for friends or family living closer to the Gaza Strip or serving in the military. It might therefore have been expected that the second survey would reveal even more uncompromising and exclusionary attitudes towards asylum seekers than the first, reflecting an inability to cope with this additional source of threat. This possibility was not borne out by the data, suggesting that threat perception towards out-groups is a stable phenomenon not driven by exposure to increased realistic threats by a different out-group member.

However, the findings may also reflect limitations to our study. First, it may be that the difference in external threat between March and August 2014 was not sufficient to create a change in attitudes. That is, it may be that, because Israeli citizens are constantly exposed to political violence as part of the ongoing Israeli–Palestinian conflict, the additional threat posed by the summer’s fighting did not harden their attitudes towards asylum policy, which were relatively exclusionary to begin with (though, as we have discussed, less so than in our Australian sample). Alternatively, the findings may simply reflect our small sample size and the fact that only about a quarter of the respondents identified as right-wing.

Yet, with respect to our study’s limitations, its findings are important in light of the fact that the Western world is becoming more and more similar to Israel, as terrorism casts its shadow over ordinary people in cafes and shopping centres and public transportation networks everywhere. The bomb attacks in Sydney, Australia, that took place on December 2014 and were treated as a terrorist attack, during its early stages (Ralston and Partridge 2014), surprised Australians. Future studies might revisit the current research model in Australia to see what influence the fear of terrorism, real or perceived, might have on Australians’ attitudes towards asylum policy.

We also extend previous research on attitudes towards asylum seekers—a subgroup of immigrants who may be most at risk for prejudice and social exclusion. We propose a mediation model linking political ideologies to policy attitudes and consider threat perceptions as the mechanism linking the two. More importantly, we test and replicate this model in a comparative framework and during two different periods of time.

At a policy level, our work provides useful guidance for social activists and policy-makers seeking to understand intergroup relations on the one hand, and the connection between public opinion and policy formation on the other. Specifically, the current findings highlight the role played by threat perceptions in increasing exclusionary attitudes towards vulnerable out-group members. This study highlights the need to better understand the sources of threat perception and how these can be modified on a population level. We also encourage potential host countries that have not already developed a policy on asylum seekers to do so as soon as possible, before they too are forced to deal with the political and social consequences of this phenomenon.

Table 6

Mediation Effect of Threat Perceptions on the Relationship between Political Ideology and Policy Attitudes (Study 2)

		SE	B
				Threat perceptions
		0.09	–0.80***	Political ideology
				Policy attitudes
		0.08	0.83***	Threat perceptions
		0.12	–0.71***	Political ideology
		0.17	0.80***	Country
		0.16	0.20	Int _1
		0.24	0.36	Int_2
Bootstrap results for mediation effects; 95% confidence interval (CI)
Upper	Lower	SE	B	Mediation effect: threat perceptions
–0.36	–0.80	0.17	–0.88	Indirect effect of X on Y
–0.53	–0.98	0.17	–0.53

p < 0.05

p < 0.01

p < 0.0001. First row = Israel, second row = Australia.