Violence,displacement,contact,and attitudes toward hosting refugees

Abstract

How do individuals’ personal experiences with various aspects of political violence affect their attitudes toward hosting conflict refugees? More specifically, how do their personal exposure to violence, their own personal experience of being displaced, and their recent contact with refugees influence these attitudes? To explore answers to these questions, we draw upon a recent survey of 2,400 Lebanese residents where we identify individuals who experienced violence during the Lebanese civil war (1975–90), those forced to flee their homes during that conflict, and those who enjoy recent contact with Syrian immigrant and/or displaced populations. We examine whether these distinct experiences affect respondents’ regard for members of the Syrian refugee population. Results demonstrate that historical exposure to violence and experience of displacement have no discernible impact on individual attitudes toward hosting refugees. We find much stronger evidence that attitudes are associated with whether individual respondents have had contact with Syrians in Lebanon; those with such interactions are significantly more likely to support hosting refugees, to consider hiring a refugee, or to allow one of their children to marry a refugee. Our findings suggest exposure to violence by itself does not correlate to positive sentiments toward refugees, especially over time. Further, finding ways to create positive contact between refugees and native populations may be associated with improving attitudes and relations between the two populations.

Keywords

contact theory exposure to violence Lebanon public attitudes refugees Syrian civil war

Introduction

The ongoing Syrian civil war displaced more than half of the country’s populace. This includes more than five million refugees who have been forced to flee to neighboring countries and further afield (UNHCR: United Nations High Commissioner for Refugees, 2017). While a great deal of heart has been demonstrated by populations hosting Syrian refugees during this crisis, attitudes toward and support for hosting appear to have frayed over time in many countries (Dempster & Hargrave, 2017). These changing attitudes accompany rising numbers of attacks against refugee populations (see e.g. Benček & Strasheim, 2016; Savun & Gineste, 2018) and overblown accusations of problems caused by refugees (Kermoliotis, 2016). Conventional wisdom suggests economic conditions on the ground, intercultural interactions, and fear regarding security implications affect local attitudes toward hosting refugees.

We contribute to research on this topic in four important ways. First, we focus on Lebanon, which hosts more than one million refugees from Syria – the highest per capita hosting rate globally. Given the extraordinary effort of the Lebanese to host Syrian refugees, this case represents an important litmus test for the limits of native population support for hosting refugees. While much of the current discussion concerning the costs of hosting refugees centers upon the European continent, we suggest more evidence is needed on the dynamics in Syria’s overburdened and fragile neighboring countries.

Second, we speak to the growing literature on the effects of exposure to violence on attitudes and empathy toward refugees. We do so by following standard practice and looking at a direct indicator of whether or not respondents were exposed to acts of violence historically – that is, during the Lebanese civil war (1975–90). We then offer an innovative twist on this concept by assessing whether respondents were themselves forced to flee their own homes at some stage during that conflict. It seems plausible to expect these experiences might affect respondents’ attitudes toward hosting refugees (Bauer et al., 2016). On the one hand, individuals that have experienced the horrors and costs of conflict themselves may be more likely to feel empathy for refugees and consider them deserving of support (Hartman & Morse, 2018). On the other hand, these experiences can have the effect of hardening prejudice and undermining support for hosting (Bauer et al., 2014; Beber, Roessler & Scacco, 2014).

Third, in addition to focusing upon aggregate levels of support for hosting refugees (Adida, Lo & Platas, 2017), we assess attitudes regarding other ways in which locals themselves are willing to aid or associate with refugees. Specifically, we assess Lebanese residents’ willingness to hire refugees or to allow them to marry their children. This combination of attitudes helps, we argue, to build a more robust sense of what Collier (2013) refers to as ‘mutual regard’, which could be thought of as necessary for successful integration of refugee populations.

Fourth, we test the validity of the traditional contact hypothesis under very difficult conditions. We argue that respondents’ attitudes regarding refugees are conditioned by the extent of their contact with Syrians in Lebanon. Building upon the ‘contact hypothesis’ (Allport, 1954), we suggest respondents’ contact with Syrians helps reduce their prejudices and enhance their levels of support (c.f. Pettigrew & Tropp, 2006; Savelkoul et al., 2011; Homola & Tavits, 2018). As noted above, however, Lebanon is under considerable stress as a host of such a large refugee population. Accordingly, levels of support for hosting refugees are in decline (Ghosn & Braithwaite, 2018). Thus, ours represents an especially taxing examination of the contact hypothesis.

To test these various expectations, we analyze data on individual attitudes towards hosting refugees drawn from a new survey of 2,400 Lebanese residents carried out during the summer of 2017. We compare attitudes among individuals, differentiating between their relative experiences during the Lebanese civil war (1975–90), including whether or not they were exposed to violence and/or forced to flee their home, as well as the extent of their more recent contact with Syrians (including refugees) in Lebanon. Our tests demonstrate that individuals’ exposure to violence and experience of displacement are not associated with their relative levels of support for hosting refugees. We find much stronger evidence, however, that individuals’ attitudes are associated with whether or not they have had contact with Syrians in Lebanon. Those with contact are shown to be significantly more likely to support hosting refugees.

The Lebanese context

Before reviewing existing research and proceeding to derive testable hypotheses, we first provide more context regarding the case of interest in this study. Lebanon has its own long history with violence and population displacement. The Lebanese civil war, which lasted from 1975 to 1990, left more than 144,000 dead, 184,000 wounded, and at least 17,000 still missing (Ghosn & Khoury, 2013). Moreover, nearly 175 towns were partially or completely destroyed and more than 30% of the population (i.e. 1.2 million) were uprooted from their homes (Charif, 1994).

Between 1990 and 2005, most of Lebanon’s domestic and foreign policy decisionmaking was in the hands of the Syrian civilian and military apparatus in Lebanon. Despite the fact civil war did not recur, there were several assassinations, assassination attempts, and disappearances during this period (Knudsen, 2010). The presence of the Syrian military for 25 years in the country left many Lebanese with negative feelings towards their Arab neighbor. As a result, Lebanon’s relationship with Syria in the context of the current Syrian civil war is a hotly contested issue in Lebanese foreign policy (Ghosn, 2016).

Since the escalation to civil war in Syria in 2012, Lebanon received and hosts more than one million Syrian refugees. In addition, Lebanon hosts 40,000 Iraqi refugees and 400,000 Palestinian refugees.¹ This means Lebanon has the highest per-capita concentration of refugees worldwide, with one in six residents a refugee (UNHCR, 2017). Many of the Syrian refugees in Lebanon are housed in districts along the border regions that are among the poorest in the country with higher than average levels of unemployment.² The sheer scale of the influx relative to Lebanon’s population severely affects the socio-economic situation in the country. Due to the Syrian refugee crisis, according to the International Labour Organization, unemployment doubled to around 20% by 2015 and economic losses incurred are estimated around US $7.5 billion.³ In 2016, the World Bank stated that the public debt widened to 157% of the GDP.⁴

As the years of the current refugee crisis have passed and the number of arrivals continues to increase, Lebanon suffered an intensification in rhetoric and escalation in violence against refugees. When the first wave of refugees arrived in Lebanon between April 2011 and 2012, most of the Lebanese, especially within the Sunni community, felt sympathy for those fleeing.⁵ Nonetheless, public attitudes toward hosting refugees in Lebanon appear to have soured over recent years. According to a survey conducted in 2013, 25% of Lebanese surveyed felt unsafe because of the presence of Syrian refugees (International Alert, 2015). By 2015, the number of individuals who felt unsafe rose to 46% and that number continued to rise to 47% in 2016, and 51% in 2017 (Ghosn & Braithwaite, 2018). Despite these intensifying sentiments, nearly all studies show there is individual and regional variation in public attitudes.

Individual attitudes towards immigrants and refugees

A vibrant literature addresses and evaluates patterns of public attitudes toward immigrant populations generally and refugee populations more specifically (Bansak, Hainmueller & Hangartner, 2016; Dempster & Hargrave, 2017). A recent survey of respondents across 140 countries between 2012 and 2014 highlighted that citizens across all regions globally, with the exception of Europe, wish to see immigration rates maintained or increased. In Europe, however, a majority wish to see immigration levels reduced significantly (International Organization for Migration; IOM, 2015). A majority of respondents across the European continent are uncomfortable with current levels of immigration, with only one-fifth in agreement with the evidence that immigration has a positive effect on receiving countries and more than one-third supporting the full closure of their country’s borders to refugees (Ipsos MORI, 2016).

While citizens across Europe tend, in the aggregate, to hold negative attitudes toward immigrants, there is quite significant variation across the continent. On the upper end, 85% in Turkey and 65% in Italy believe there are too many immigrants in their home countries, with majorities in both countries, as well as in Hungary, also supporting calls to close their borders to refugees (Ipsos MORI, 2016). At the lower end of the negative attitudes scales are a series of countries with smaller populations of immigrants, including Sweden, Denmark, and Finland (Heath & Richards, 2016).

Aggregate survey numbers also mask meaningful subnational variation with respect to geographic location. There are greater levels of tolerance and favorable attitudes towards immigrants and refugees in urban areas in which more diverse populations reside (Crawley, Drinkwater & Kauser, 2013; Blinder & Allen, 2016). There are, however, lower levels of tolerance in ethnically homogenous areas adjacent to locations with more diverse populations (Kaufman, 2015). This reflects a more general trend in which people are less supportive of the arrival of members of ethnic and racial out-groups, who they worry will struggle or refuse to integrate into the fabric of society (Hainmueller & Hiscox, 2007; Pearson, 2010; Heath & Richards, 2016).

Peoples’ evaluations and attitudes also vary depending upon which subgroup of the immigrant population is identified (Ford, Morrell & Heath, 2012). Attitudes are generally less sympathetic if the group is referred to as illegal immigrants rather than asylum seekers (Augoustinos & Quinn, 2003) and are most sympathetic if the group is referred to as refugees (Mayda, 2006; O’Rourke & Sinnott, 2006; Rowe & O’Brien, 2014; Hatton, 2016). This difference appears to hold even if there are costs and risks associated with supporting refugee populations (Bansak, Hainmueller & Hangartner, 2016).

Underlying public concerns regarding hosting migrant populations – even among populations otherwise supportive of compassionate policies toward refugees – are concerns about undue burdens placed on local welfare provision and economic performance (Scheve & Slaughter, 2001; Wike, Stokes & Simmons, 2016), as well as perceived threats to security (Salehyan & Gleditsch, 2006; Salehyan, 2008; Böhmelt, Bove & Gleditsch, 2018; Fisk, 2018; Rüegger, 2019). These concerns fit within the perspective that immigration is a ‘state of the nation’ issue with attitudes shaped by concerns about the impact on broader society, rather than personal circumstances (Hainmueller & Hopkins, 2014; Hatton, 2016).

Collier (2013) offers an extensive discussion of the notion of ‘mutual regard’, which he contends is fundamental to two behaviors that are important to building successful social models: the willingness of the successful to finance transfers to the less successful and the facilitation of cooperation to generate public goods. In other words, ‘[m]utual regard is valuable for the trust that supports cooperation and the empathy that supports redistribution’ (Collier, 2013: 72), and, accordingly, ought to play an especially important role in influencing public attitudes towards hosting refugees. Public levels of support for social welfare programs or state generosity towards refugees may be driven by government policies and/or media representations that dehumanize arriving populations via the use of terms such as ‘swarms’, ‘invasions’, and ‘floods’ (Esses, Hamilton & Gaucher, 2017; Crawley & McMahon, 2016).

These attitudes also depend upon judgments regarding the ‘deservingness’ of recipients (van Oorschot, 2006; Peterson, 2012; Feldman et al., 2017). Deservingness rests upon perceptions of recipients’ need for assistance, responsibility for their own circumstances, and apparent willingness to contribute to a societal good (van Oorschot, 2006; Petersen, 2012). For instance, Doherty (2015) demonstrates that opponents to immigration in Australia look especially unfavorably upon those arriving by boat, because they view them as having ‘jumped the queue’.

Feldman et al. (2017) contend the long-running Syrian refugee crisis and powerful images from 2015 of the young Syrian child, Alan Kurdi, who drowned while trying to flee from Turkey to mainland Europe, swung attitudes somewhat more in favor of refugees. They suggest that, in this context, ‘hot’ reactions grounded in empathy and related emotions of sympathy and compassion dominated ‘cold’ judgments of refugee deservingness (Feldman et al., 2017: 4). In other words, while the literature speaks to a generally negative set of attitudes toward hosting refugees (Getmansky, Sınmazdemir & Zeitzoff, 2018), there are conditions under which to expect more positive attitudes. We next explore two such possible conditions: experience with violence and displacement and contact with refugee populations.

Experience with violence and displacement

Recent studies in Turkey and Lebanon reveal that as time passes, attitudes towards refugees in host communities shift depending on economic factors like employment rates, housing costs, and market competition. In fact, as the length of stay and number of Syrian refugees increase, public opinion becomes more negative, reaching hostile levels in both countries (İçduygu, 2015; Alsharabati & Nammour, 2017; Ghosn & Braithwaite, 2018). One of the prime motivations for the present study is to determine whether there is meaningful variation in attitudes within host countries. In deriving expectations regarding the roots of variation in support for hosting refugees, we begin with the observation that the vast majority of refugees fleeing conflict zones, including those exiting Syria, settle in countries that have themselves a meaningful history of violence (Hartman & Morse, 2018). It stands to reason, therefore, that host population’s attitudes towards hosted populations reflect their own personal experiences with conflict.

The literature on the social and political legacies of violence upon the fate of post-conflict societies offers competing claims. On the more negative side of the coin, this includes increasing support for militants and right-wing parties, loss of trust in institutions, opposition to disarmament and out-groups, and hardening attitudes in support of exclusionist policies (Canetti-Nisim et al., 2009; Bauer et al., 2014, 2016; Canetti et al., 2017). Evidence of these effects is demonstrated in Israel (Berrebi & Klor, 2008; Getmansky & Zeitzoff, 2014), Northern Ireland (Hayes & McAllister, 2001), Tajikistan (Cassar, Grosjean & Whitt, 2014), and Turkey (Kibris, 2011).

One plausible underlying theme in explaining these negative effects is the suggestion that the psychological and sociocultural infrastructures that develop during a conflict lead to ‘threat-buffering’ beliefs. That is, in order to ‘cope with the negative consequences of ongoing perceived threat, societies adopt conflict-supporting beliefs, giving them a meaningful picture of the conflict, reducing their sense of uncertainty and stress’ (Canetti et al., 2017: 86). While many of these applications of parochial legacies of violence relate to outcomes within conflict countries, Hartman & Morse (2018) note they could easily be expected to apply to intergroup relationships during refugee crises. Refugees may flee into border regions that are rife with security concerns or ethnic divisions. They may also trigger group-level perceptions of threat by increasing the risk associated with the spread of conflict (Whitaker, 2003; Salehyan & Gleditsch, 2006).

By contrast, there are a growing number of studies detailing potentially positive effects and outcomes from individual exposure to violence during civil conflicts. This includes lower than expected out-group biases in Bosnia (Whitt & Wilson, 2007), higher than expected within-group trust in Burundi (Voors et al., 2012) and Nepal (Gilligan, Pasquale & Samii, 2014), and increased political mobilization and collective action in Sierra Leone (Bellows & Miguel, 2009) and Uganda (Blattman, 2009). At their core, these studies reflect a diffuse set of logics which center upon the potential for growth in prosocial behaviors in the aftermath of trauma. In the context of refugee crises, these prosocial behaviors and attitudes reflect a mechanism of empathy-driven altruism, which transcend identity boundaries (Hartman & Morse, 2018).

Given these effects, civilians exposed to violence are expected to harbor relatively positive attitudes and perceptions regarding hosting Syrian refugees, at least as compared to the average member of society. This is because those individuals, having experienced the horrors and costs of conflict for themselves, are more likely to feel empathy for refugees and consider them deserving of support. Moreover, those that experienced displacement due to conflict are considered especially likely to display empathy toward those that more recently shared this fate.

Given Lebanon’s history with political violence, as well as its ongoing experience with Syrian refugees, it is important to adjudicate as to which of these two competing logics is more likely to influence public attitudes. From the perspective of developing mutual regard, for instance, one may conclude personal exposure to violence and experience with displacement during the Lebanese civil war hardens attitudes against out-group cooperation and undermines development of mutual regard for fleeing Syrians in the ongoing crisis. On the other hand, traumas associated with experiences of violence and displacement can result in Lebanese individuals viewing displaced Syrians as kindred spirits, which ensures a measure of empathy and helps to normalize out-group differences. These are precisely the conditions that would help generate mutual regard. Accordingly, we present competing hypotheses:

Hypothesis 1a: Individuals with exposure to violence and/or experience with displacement are less likely to support hosting refugees, as compared to those without such exposure or experience.

Hypothesis 1b: Individuals with exposure to violence and/or experience with displacement are more likely to support hosting refugees, as compared to those without such exposure or experience.

Contact with refugee populations

Given the declining state of support for hosting refugees in most recipient countries, a great deal of energy is focused upon identifying any means of remedying these worsening trends. Given the considerable resource burden faced in host countries, it is important to establish under what conditions positive out-group attitudes and mutual regard are developed rather than undermined. With these goals in mind, Dempster & Hargrave’s (2017) comprehensive review concludes with the recommendation governments and agencies must design policies around the principle of encouraging contact between local and immigrant populations. They follow a rich tradition in suggesting that such contact helps reduce prejudice and unfavorable attitudes, as well as improve prospects for integration and cooperation (Allport, 1954; McLaren, 2003; Savelkoul et al., 2011).⁶

Drawing upon this more optimistic perspective, we suggest Lebanese respondents’ contacts with Syrian immigrants attenuate prejudices and enhance their mutual interests and regard (c.f. Pettigrew & Tropp, 2006). According to the contact hypothesis, interactions expose individuals to new information regarding one another. This new information helps generate affective ties, which reduce fears and help develop positive views of the out-group (Homola & Tavits, 2018). Such contact has the effect of reducing prejudices and enhancing mutual interests because interaction leads to opportunities for self-disclosure (Pettigrew, 2008). Pettigrew & Tropp’s (2006) meta-analysis of 515 studies of intergroup contact theory finds greater intergroup contact is consistently associated with lower levels of prejudice. As an example, Rodon & Franco-Guillén (2014) find that Catalonians who had contact with immigrants – notably, in the form of close contact and family contact – were less likely to have negative attitudes toward immigration and immigrants. In the context of rising anti-immigrant views across Europe, especially with respect to anti-Muslim attitudes, Savelkoul et al. (2011) demonstrate that individuals who have had contact with colleagues belonging to ethnic minority groups were less likely to harbor negative attitudes toward Muslims.

In line with this logic, we anticipate:

Hypothesis 2: Contact with Syrian immigrants is associated with individuals’ support for hosting refugees, as compared to those without such contact.

Empirical strategy

In order to gather the variety of variables of interest required to test our hypotheses, we draw upon responses from a survey of 2,400 members of the Lebanese population between 31 May and 31 August 2017. By design, we imposed two sets of restrictions on the sample of individual respondents surveyed. First, we were interested in hearing from individuals who could have been exposed to violence or displaced during the civil war and their perception of hosting Syrian refugees. As a result, we restricted participation in the survey to individuals currently 40 years of age or older (i.e. at least two years old when the civil war began). Accordingly, we have a more mature sample than would be fully representative of the population of Lebanon. Research shows there is typically less hostility towards refugees and other migrant populations among younger individuals (Crawley, 2009; IOM, 2015; Heath & Richards, 2016). Consequently, we feel confident to the extent this would bias our results, it would do so by producing results in line with Hypothesis 1a rather than 1b. Second, given the majority of the violence and displacement occurred in three governorates (62% from Mount Lebanon, 24% from the South, and 8% from Beirut), we slightly oversampled from these areas of the country. These data are depicted in Table I and in Figure 1.⁷

Dependent variables

With regard to capturing attitudes toward support for hosting refugees, we use three questions from our survey to generate dependent variables. The first is an ordinal variable gauging how much a respondent supports the Lebanese government’s decision to host refugees, ranging from 1 if they completely disagree with hosting through to 5 if they completely agree. This first measure captures a generic sense of how willing individuals are to

Table I.

Population distribution

	Total population	Displaced population	Sampling population
Beirut	10%	8%	11%
Mount Lebanon	35%	62%	49%
South & Nabiteh	20%	24%	24%
North & Akkar	23%	4%	10%
Bekaa & Baalbek-Hermel	12%	2%	6%
Total (Lebanon)	100%	100%	100%

While there has been no official census in Lebanon since 1932, the distribution of the total population is based on the distribution of residency.

have the state provide goods and resources to support refugees. In order to understand how willing individuals would be to provide some of these resources themselves, we then operationalize two binary variables assessing whether the respondent would be willing to (a) hire a refugee, or (b) allow their child to marry a refugee. Full details of the survey questions used to operationalize the dependent and explanatory variables are provided in the Online appendix.

Explanatory variables

As detailed earlier, we are primarily concerned with three sets of experiences among Lebanese respondents. In order to examine our pair of competing Hypotheses, 1a and 1b, we operationalize two indicators of experience of conflict. First, we are interested in whether or not a respondent was previously exposed to violence during the Lebanese civil war. Our survey includes questions asking whether an individual experienced violence by means of physical assault, torture, abduction, sexual violence, shelling, being shot at, forced labor, and/or wage theft. If they fell victim to any of these acts, they are coded as exposed to violence. Second, we are interested in whether respondents became displaced because of the Lebanese civil war. For this, we asked individuals if they left their homes during the course of the civil war.

Third, as suggested in Hypothesis 2, contact with Syrian immigrants might have a positive influence on individuals’ levels of support for hosting Syrian refugees. Naturally, skeptical readers may be concerned that contact may not be exogenous to the formation of attitudes toward refugees – that those that are predisposed to want to support refugees may seek out contact with them as volunteers, employers, etc. We attempt to gauge the extent of individual contacts to try to address this concern and because brief contacts are shown

Figure 1.

Percentage of Lebanese population, displaced population, and survey sample, by governorate

Table II.

Summary statistics of dependent and independent variables

Dependent variable	N	Mean	Std. dev.	Minimum	Maximum
Support hosting refugees	2,356	2.709	1.437	1	5
Hire refugee	2,336	0.270	0.444	0	1
Allow child to marry refugee	2,288	0.137	0.344	0	1
Experienced violence	2,356	0.751	0.432	0	1
Displaced	2,356	0.338	0.473	0	1
Syrian friend(s) in Lebanon	2,356	0.272	0.445	0	1
Know displaced Syrian(s)	2,356	0.424	0.494	0	1

elsewhere to be potentially counterproductive in the formation of attitudes toward out-groups (see e.g. Enos, 2014).

In our survey, we ask individuals a series of questions about their contact with and knowledge of Syrians in Lebanon. These questions enable us to create a series of variables measuring the extent of contact between respondents and Syrian immigrants in Lebanon. Capturing the extent of contact allows us to judge whether more contact improved attitudes. First, we ask whether the respondent has Syrian friends in Lebanon, without prompting whether the ‘friend’ is displaced or not. Second, we ask if the respondent has had any contact with a displaced Syrian. Contact can include interactions and relationships such as a neighbor, family, acquaintance, employee, or through humanitarian assistance. Table II provides summary statistics for each of these key explanatory variables, as well as for the aforementioned dependent variables.

Before proceeding to discuss the remainder of our model specification and our estimation strategy, we briefly highlight the distribution of responses on each of our dependent variables across the various experiences captured in our explanatory variables. In Figure 2, we summarize the levels of support individuals have for hosting refugees (on a five-point scale from strongly disagreeing to strongly agreeing) along the x-axes. Each quadrant of the figure then highlights the number of respondents and their respective levels of support according to the four experiences of interest to our test hypotheses. In the top left quadrant, we observe many more respondents experienced violence during the Lebanese civil war than did not. We also note that many more respondents disagree with supporting refugees than agree with providing support. Both of these trends are consistent with prior data. Finally, we see similar trends between the two conditions, that is, between those that did and did not experience violence.

Indeed, as we compare this to the other quadrants in Figure 2, it becomes clear that across a majority of conditions and experiences, there is a fairly stable distribution of levels of support, with relative peaks towards the disagree end of the scale. This holds whether respondents were or were not displaced during the civil war, as well as if respondents do not have Syrian friends or do not know displaced Syrians. However, when we look at these two contact variables, there is a subtle and marginally discernible increase in the relative likelihood of agreement when individuals have had contact with Syrians.

Figure 3 provides a similar overview of responses for our second dependent variable: whether or not individuals would consider hiring a Syrian refugee. Once again, a similar pattern emerges. Whether or not individuals were exposed to violence or displaced during the civil war, they tend to be less likely to be willing to consider

Figure 2.

Levels of support for hosting refugees by conflict experience and contact

Figure 3.

Willingness to hire a refugee by conflict experience and contact

Figure 4.

Willingness to allow child to marry a refugee by conflict experience and contact

hiring a Syrian refugee. However, those with contact even out the responses between positive and negative.

A similar, though less pronounced, pattern exists with respect to the respondent’s willingness to have their child marry a Syrian refugee. Figure 4 demonstrates, in this instance, that having a Syrian friend has a stronger effect than simply knowing a displaced Syrian with respect to determining support for marriage.

Control variables

Similar to other studies using survey evaluations of individual attitudes, we include a series of common demographic control variables in our model specifications (e.g. Branton, 2007; Savelkoul et al., 2011). We control for the respondent’s age, gender, marital status, religion, education, and income. Male is coded as 1 if the respondent identified as male and 0 if female. Married is a binary indicator coded 1 if the individual is married and 0 otherwise. Individuals identified themselves as Sunni (baseline), Shia, Maronite, Eastern Orthodox, Catholic, Druze, or other. Education is a multicategorical variable, with the categories representing education up to primary school (baseline), some secondary school, high school, and college and above. Respondents are grouped into three categories for Monthly income, based on whether they earn less than minimum wage (baseline), $501–1500, or exceed $1501.

Methodology

Given our dependent variables are categorical and limited in nature, ordinary least squares (OLS) regression is not the most appropriate model for estimation (King, 1998). For our binary dependent variables, we run logistic regressions. For our ordered dependent variables, we employ ordered logistic regressions (Long, 1997).⁸ In order to capture any unobserved heterogeneity based on geography, we include district fixed effects in all models (Clark & Linzer, 2015). District fixed effects offer the virtue of capturing cross-unit variation, for instance, in numbers of refugees hosted and levels of violence experienced at various times in the past, as well as proximity to Syria and experience of violent events carried out by Syrians, each of which might be thought of as influencing Lebanese attitudes towards hosting

Figure 5.

Coefficient plot of support for hosting refugees

refugees. Finally, we also include a simple set of time fixed effects. Our survey was in the field between 31 May and 31 August 2017. Treating observations from June (including 31 May) as our excluded baseline category, we include fixed effects variables, respectively, for responses provided in July and August. These time fixed effects help us control for any unobserved variation that may be associated with the passing of three months while responses were collected.

Discussion of results

We present the results of our models in a series of coefficient plots. Full tabulated regression results are reported in Tables III–V in the Online appendix. Each coefficient plot presents results for a different dependent variable. The plot depicts point estimates of coefficients with associated 95% confidence intervals. The vertical line is 0, such that point estimates to the right demonstrate a positive relationship between the covariate and the dependent variable and estimates to the left reflect a negative relationship. Figure 5 depicts results for Lebanese support for hosting Syrian refugees. Figure 6 shows results for willingness to hire a Syrian refugee. Figure 7 displays results for willingness to allow a child to marry a refugee.

Across our dependent variables, a respondent’s experience during the Lebanese civil war does not have a systematic bearing on their attitudes toward refugees – in each plot the confidence intervals around the point estimates for experience of violence and displacement always overlap with zero. Therefore, we do not find support for either Hypothesis 1a or 1b. This null result could either reflect that prior experiences now have no effect on attitudes or that they have competing effects that cancel one another out. As things stand, we cannot easily adjudicate between these two explanations; however, this is a fruitful topic of future research.

With respect to Hypothesis 2, that contact will improve Lebanese attitudes towards Syrian refugees, we find much clearer and consistent evidence. Contact with Syrians is positively associated with attitudes and statistically significant for both forms of contact – with Syrian friends and with members of the displaced population. Individuals that have contact with Syrians in either way are significantly more likely to support hosting, hire, or

Figure 6.

Coefficient plot of hiring refugees

allow their children to marry refugees. From a substantive perspective, there is only a 16% chance a respondent with no contact with Syrians or refugees would strongly agree with hosting refugees. This probability rises to 24% for those that have had contact. Respondents who have had contact with Syrians are also more than twice as likely to hire or to allow their child to marry a refugee, as compared to those that have not had such contact. This means that in Lebanon, even at a time where favorable attitudes towards refugees continue to decline year-on-year, contact with Syrians has the effect of reducing prejudice and hostility.

Moving next to a discussion of our control variables, our model specifications return a small number of interesting conclusions. We find men are significantly more likely than women to be supportive of hosting refugees, hiring refugees, and allowing their child to marry a refugee. In addition, we find Shia Lebanese are consistently less likely than Sunni Lebanese to relay positive attitudes regarding Syrian refugees. This effect holds across each of our models. With respect to general support for hosting, all religious groups display lower levels than do Sunnis. Shias were also less likely to hire a refugee than a Sunni. With respect to allowing a child to marry, Maronite Lebanese join Shia in reflecting lower levels of willingness. Collectively, these findings with respect to religion reflect that the majority of refugees are Sunni.

Additionally, respondents with lower levels of education are more likely to support hosting refugees and those with medium and high levels of income are more likely to be willing to hire refugees. We also find some variation in responses and effects over time. In general, those interviewed later in our survey are more likely to hold positive attitudes towards refugees. The one exception is among respondents in July when asked whether they support hosting refugees. These individuals are less likely to support the government hosting refugees than those surveyed in June, though this relationship is not quite statistically significant. We may be seeing significant time effects because of violent events involving refugees during the time of our survey (see e.g. Braithwaite et al., forthcoming).

In sum, we find strong evidence for our second hypothesis, suggesting an individual’s contact with (displaced) Syrians bolsters their support for hosting refugees. This provides support for contact theory, conveying interactions with an out-group are associated with harboring more positive attitudes toward them. At

Figure 7.

Coefficient plot of allowing child to marry refugee

the same time, we do not find evidence of any relationship between respondents’ own experiences in the Lebanese civil war –in terms of exposure to either violence or displacement – and their attitudes toward Syrian refugees. We believe this may be because of the protracted nature of the Syrian refugee crisis in Lebanon.⁹ While other studies find that individuals who also go through civil war experiences are more likely to empathize with refugee populations (see e.g. Hartman & Morse, 2018), our findings suggest that over time, these positive sentiments begin to wane. Lebanon has the highest density of Syrian refugees, even if they do not host the most. Since we interviewed Lebanese citizens in year five of this ongoing process, their attitudes have hardened, reducing the likelihood that they are able to empathize.

Conclusion

In this article, we explore how an individual’s experiences during the Lebanese civil war (1975–90), as well as their recent contact with Syrian immigrants, affect their attitudes toward hosting conflict refugees from the ongoing civil war in Syria. This is a meaningful context in which to investigate this puzzle because Lebanon serves as host to more than one million refugees from the Syrian civil war, which represents almost one-third of their own total population.

The conflict processes literature reveals a debate between those arguing exposure to violence generates empathy for subsequent victims and studies claiming the affected judge others in the same boat more harshly than those never affected. Drawing on a new survey of 2,400 Lebanese residents, we examine how experiencing violence during the civil war, being displaced because of the conflict, having Syrian friends, and having contact with displaced Syrians influence attitudes toward hosting Syrian refugees.

Our tests demonstrate that exposure to violence and experience of displacement do not have an impact on individuals’ attitudes regarding hosting refugees. However, we do find compelling evidence that individuals’ attitudes are associated with whether they have contact with Syrians in Lebanon. In fact, those with contact, whether via friends or otherwise, are shown to be more likely to support hosting refugees, which we ask directly. In addition to being associated with attitudes regarding state provision of support, we also show that respondents’ contact with Syrians correlates with their willingness to hire a refugee or allow their children to marry refugees. Our findings are important, as understanding public attitudes towards refugees within their host communities has become a major interest, and debate, in almost every European and Middle Eastern country, as well as the United States. Knowledge of such attitudes help us in building better policy responses.

While our findings do not confirm a causal relationship between these factors, they do suggest contact and positive sentiments go hand-in-hand. Accordingly, encouraging more contact between local communities and refugees might have the benefit of facilitating greater levels of support within communities towards the refugees that they host. This most likely requires a radical rethinking of the conditions under which most refugees are housed. In other words, it may be necessary to find ways of moving refugees out of isolated (and overcrowded) camps and into local communities.

Footnotes

Replication data

The Online appendix, dataset, and replication code for the empirical analysis in this article can be found at .

Acknowledgments

The authors are grateful to the editors of Journal of Peace Research, the special issue editors, and the three anonymous reviewers for their helpful suggestions. The views expressed are those of the authors and should not be attributed to the US Department of Defense or the US Navy Research Office/Army Research Laboratory.

Funding

The research reported here was supported by award W911-NF-17-1-0030 from the Department of Defense and US Army Research Office/Army Research Laboratory under the Minerva Research Initiative.

Notes

References

Adida

Claire L

Adeline

Platas

Melina

(2017) Engendering empathy, begetting backlash: American attitudes toward Syrian refugees. Working paper.

Allport

Gordon W

(1954) The Nature of Prejudice. Cambridge, MA: Perseus.

Alsharabati

Carole

Nammour

Jihad

(2017) Survey on perceptions of Syrian refugees in Lebanon: Between resilience and vulnerability (http://www.sciences-po.usj.edu.lb/pdf/Presentation%20EN.pdf).

Augoustinos

Martha

Quinn

Cheryl

(2003) Social categorization and attitudinal evaluations: Illegal immigrants, refugees or asylum seekers? New Review of Social Psychology 2(1): 29–37.

Bansak

Kirk

Hainmueller

Jens

Hangartner

Dominik

(2016) How economic, humanitarian, and religious concerns shape European attitudes toward asylum seekers. Science 354(6309): 217–222.

Bauer

Michal

Blattman

Christopher

Chytilová

Julie

Henrich

Joseph

Miguel

Edward

Mitts

Tamar

(2016) Can war foster cooperation? Journal of Economic Perspectives 30(3): 249–274.

Bauer

Michal; Alessandra Cassar, Julie Chytilová & Joseph Henrich

(2014) War’s enduring effects on the development of egalitarian motivations and in-group biases. Psychological Science 25(1): 47–57.

Beber

Bernd

Roessler

Philip

Scacco

Alexandra

(2014) Intergroup violence and political attitudes: Evidence from a dividing Sudan. Journal of Politics 76(3): 649–665.

Bellows

John

Miguel

Edward

(2009) War and local collective action in Sierra Leone. Journal of Public Economics 93: 1144–1157.

10.

Benček

David

Strasheim

Julia

(2016) Refugees welcome? A dataset on anti-refugee violence in Germany. Research & Politics 3(4). https://doi.org/10.1177/2053168016679590.

11.

Berrebi

Claude

Klor

Esteban F

(2008) Are voters sensitive to terrorism? Direct evidence from the Israeli electorate. American Political Science Review 102(3): 279–300.

12.

Blattman

Christopher

(2009) From violence to voting: War and political participation in Uganda. American Political Science Review 103(2): 231–247.

13.

Blinder

Scott

Allen

William L

(2016) UK Public Opinion toward Immigration: Overall Attitudes and Level of Concern. Oxford: Migration Observatory.

14.

Böhmelt

Tobias

Bove

Vincenzo

Gleditsch

Kristian Skrede

(2019) Blame the victims? Refugees, state capacity, and non-state actor violence. Journal of Peace Research 56(1): 73–87.

15.

Braithwaite

Alex

Chu

Tiffany S

Curtis

Justin

Ghosn

Faten

(forthcoming) Violence and the perception of risk associated with hosting refugees. Public Choice. https://doi.org/10.1007/s11127-018-0599-0.

16.

Branton

Regina

(2007) Latino attitudes toward various areas of public policy: The importance of acculturation. Political Research Quarterly 60(2): 293–303.

17.

Canetti

Daphna

Elad-Strenger

Julia

Lavi

Iris

Guy

Dana

Bar-Tal

Daniel

(2017) Exposure to violence, ethos of conflict, and support for compromise: Surveys in Israel, East Jerusalem, West Bank and Gaza. Journal of Conflict Resolution 61(1): 84–113.

18.

Canetti-Nisim

Daphna

Halperin

Eran

Sharvit

Keren

Hobfoll

Stevan E

(2009) A new stress-based model of political extremism: Personal exposure to terrorism, psychological distress and exclusionist political attitudes. Journal of Conflict Resolution 53(3): 363–389.

19.

Cassar

Alessandra

Grosjean

Pauline

Whitt

Sam

(2014) Social preferences of ex-combatants: Survey and experimental evidence from postwar Tajikistan. In: Wärneryd

Karl

(ed.) The Economics of Conflict: Theory and Empirical Evidence. Cambridge, MA: MIT Press, 231–262.

20.

Charif

Hassan

(1994) Regional development and integration. In: Collings

Deirdre

(ed.) Peace for Lebanon. Boulder, CO: Lynne Rienner, 151–161.

21.

Clark

Tom S

Linzer

Drew A

(2015) Should I use fixed or random effects? Political Science Research and Methods 3(2): 399–408.

22.

Collier

Paul

(2013) Exodus: How Migration is Changing Our World. Oxford: Oxford University Press.

23.

Crawley

Heaven

(2009) Understanding and Changing Public Attitudes: A Review of Existing Evidence from Public Information and Communication Campaigns. Swansea: Centre for Migration Policy Research, Swansea University.

24.

Crawley

Heaven

McMahon

Suzanne

(2016) Beyond Fear and Hate: Mobilising People Power to Create A New Narrative on Migration and Diversity. Coventry: Centre for Trust, Peace and Social Relations, Coventry University.

25.

Crawley

Heaven

Drinkwater

Stephen

Kauser

Rukhsana

(2013) Regional Variations in Attitudes Towards Refugees: Evidence from Great Britain. London: Centre for Research and Analysis of Migration (CREAM).

26.

Dempster

Helen

Hargrave

Karen

(2017) Understanding public attitudes towards refugees and migrants. Working paper, Overseas Development Institute.

27.

Doherty

Ben

(2015) Call Me Illegal: The Semantic Struggle Over Seeking Asylum in Australia. Oxford: Reuters Institute for the Study of Journalism.

28.

Enos

Ryan D

(2014) Causal effect of intergroup contact on exclusionary attitudes. Proceedings of the National Academy of Sciences 111: 3699–3704.

29.

Esses

Victoria M

Hamilton

Leah K

Gaucher

Danielle

(2017) The global refugee crisis: Empirical evidence and policy implications for improving public attitudes and facilitating refugee resettlement. Social Issues and Policy Review 11(1): 78–123.

30.

Feldman

Stanley

Huddy

Leonie

Wronsky

Julie

Lown

Patrick

(2017) Compassionate policy support: The interplay of empathy and ideology. Working paper, Stony Brook University.

31.

Fisk

Kerstin

(2019) Camp settlement and communal conflict in sub-Saharan Africa. Journal of Peace Research 56(1): 58–72.

32.

Ford

Robert

Morrell

Gareth

Heath

Anthony

(2012) ‘Fewer but better?’ Public views about immigration. In: Park

Alison

Clery

Elizabeth

Curtice

John

Phillips

Miranda

Utting

David

(eds) British Social Attitudes: The 29th Report. London: NatCen Social Research, 26–44.

33.

Ford

W Scott

(1986) Favorable intergroup contact may not reduce prejudice: Inconclusive journal evidence, 1960–1984. Sociology and Social Research 70: 256–258.

34.

Getmansky

Anna

Zeitzoff

Thomas

(2014) Effect of rocket threat on voting in Israeli elections. American Political Science Review 108(3): 588–604.

35.

Getmansky

Anna

Sınmazdemir

Tolga

Zeitzoff

Thomas

(2018) Refugees, xenophobia, and domestic conflict: Evidence from a survey experiment in Turkey. Journal of Peace Research 55(4): 491–507.

36.

Ghosn

Faten

(2016) Lebanon: The lack of a unified and independent foreign policy. In: Halabi

Yakub

(ed.) Democratic Peace Across the Middle East: Islam and Political Modernisation. London: IB Tarus, 74–94.

37.

Ghosn

Faten

Braithwaite

Alex

(2018) Could contact stem the rising tide of negative attitudes towards hosting Syrian refugees in Lebanon? Oxford Monitor of Forced Migration: Field Monitor 7(2): 69–74.

38.

Ghosn

Faten

Khoury

Amal

(2013) The case of the 2006 war in Lebanon: Reparation? Reconstruction? Or both? International Journal of Human Rights 17(1): 1–17.

39.

Gilligan

Michael J

Pasquale

Benjamin J

Samii

Cyrus

(2014) Civil war and social cohesion: Lab-in-the-field evidence from Nepal. American Journal of Political Science 58(3): 604–619.

40.

Hainmueller

Jens

Hiscox

Michael

(2007) Educated preferences: Explaining attitudes toward immigration in Europe. International Organization 61(2): 399–442.

41.

Hainmueller

Jens

Hopkins

Daniel J

(2014) Public attitudes toward immigration. Annual Review of Political Science 17: 225–249.

42.

Hartman

Alexandra

Morse

Benjamin S

(2018) Violence, empathy and altruism: Evidence from the Ivorian refugee crisis in Liberia. British Journal of Political Science. https://doi.org/10.1017/S0007123417000655.

43.

Hatton

Timothy J

(2016) Immigration, public opinion and the recession in Europe. Economic Policy 31(86): 205–246.

44.

Hayes

Bernadette C

McAllister

Ian

(2001) Sowing dragon’s teeth: Public support for political violence and paramilitarism in Northern Ireland. Political Studies 49(5): 901–922.

45.

Heath

Anthony

Richards

Lindsay

(2016) Attitudes towards Immigration and Their Antecedents: Topline Results from Round 7 of the European Social Survey. London: European Social Survey ERIC.

46.

Homola

Jonathan

Tavits

Margit

(2018) Contact reduces immigration-related fears for leftist but not for rightist voters. Comparative Political Studies 51(13): 1789–1820.

47.

Hopkins

Daniel J

Tran

Van C

Williamson

Abigail Fisher

(2014) See no Spanish: Language, local context, and attitudes toward immigration. Politics, Groups, and Identities 2: 35–51.

48.

İçduygu

Ahmet

(2015) Syrian Refugees in Turkey: The Long Road Ahead. Washington, DC: Migration Policy Institute.

49.

International Alert (2015) Citizens’ Perceptions of Security Threats Stemming from the Syrian Refugee Presence in Lebanon. Beirut: International Alert.

50.

IOM (2015) How the World Views Migration. Paris: International Organization for Migration (IOM).

51.

Ipsos MORI (2016) Global views on Immigration and the Refugee Crisis. London: Ipsos MORI.

52.

Kaufman

Stuart J

(2015) Nationalist Passions. Ithaca, NY: Cornell University Press.

53.

Kermoliotis

Teo

(2016) Hoaxmap: Debunking false rumours about refugee ‘crimes’ (http://www.aljazeera.com/news/2016/02/debunks-false-rumours-refugee-crimes-160216153329110.html, accessed 19 September 2017).

54.

Kibris

Arzu

(2011) Funerals and elections: The effects of terrorism on voting behavior in Turkey. Journal of Conflict Resolution 55(2): 220–247.

55.

King

Gary

(1998) Unifying Political Methodology: The Likelihood Theory of Statistical Inference. Ann Arbor, MI: University of Michigan Press.

56.

Knudsen

Are

(2010) Acquiescence to assassinations in post-civil war Lebanon? Mediterranean Politics 15(1): 1–23.

57.

Long, J Scott (1997) Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA: Sage.

58.

Mayda

Anna Maria

(2006) Who is against immigration? A cross-country investigation of individual attitudes towards immigrants. Review of Economics and Statistics 88(3): 510–530.

59.

McLaren

Lauren M

(2003) Anti-immigrant prejudice in Europe: Contact, threat perception, and preferences for the exclusion of migrants. Social Forces 81(3): 909–936.

60.

O’Rourke

Kevin H

Sinnott

Richard

(2006) The determinants of individual attitudes towards immigration. European Journal of Political Economy 22(4): 838–861.

61.

Pearson

Matthew R

(2010) How ‘undocumented workers’ and ‘illegal aliens’ affect prejudice toward Mexican immigrants. Social Influence 5(2): 118–132.

62.

Petersen

Michael Bang

(2012) Social welfare as small-scale help: Evolutionary psychology and the deservingness heuristic. American Journal of Political Science 56(1): 1–16.

63.

Pettigrew

Thomas F

(2008) Future directions of intergroup contact theory and research. International Journal of Intercultural Relations 32(3): 187–199.

64.

Pettigrew

Thomas F

Tropp

Linda R

(2006) A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology 90(5): 751–783.

65.

Rodon

Toni

Franco-Guillén

Núria

(2014) Contact with immigrants in times of crisis: An exploration of the Catalan case. Ethnicities 14(5): 650–675.

66.

Rowe

Elizabeth

O’Brien

Erin

(2014) Genuine refugees or illegitimate ‘boat people’: Political constructions of asylum seekers and refugees in the Malaysia Deal debate. Australian Journal of Social Issues 49(2): 171–193.

67.

Rüegger

Seraina

(2019) Refugees, ethnic power relations, and civil conflict in the country of asylum. Journal of Peace Research 56(1): 42–57.

68.

Salehyan

Idean

(2008) The externalities of civil strife: Refugees as a source of international conflict. American Journal of Political Science 52(4): 787–801.

69.

Salehyan

Idean

Gleditsch

Kristian S

(2006) Refugees and the spread of civil war. International Organization 60(2): 335–366.

70.

Savelkoul

Michael

Scheepers

Peer

Tolsma

Jochem

Hagendoorn

Louk

(2011) Anti-muslim attitudes in the Netherlands: Tests of contradictory hypotheses derived from ethnic competition theory and intergroup contact theory. European Sociological Review 27(6): 741–758.

71.

Savun

Burcu

Gineste

Christian

(2019) From protection to persecution: Threat environment and refugee scapegoating. Journal of Peace Research 56(1): 88–102.

72.

Scheve

Kenneth F

Slaughter

Matthew J

(2001) Labor market competition and individual preferences over immigration policy. Review of Economics and Statistics 83(1): 133–145.

73.

UNHCR (2017) Syria regional refugee response (http://data.unhcr.org/syrianrefugees/regional.php, accessed 19 September 2017).

74.

van Oorschot

Wim

(2006) Making the difference in social Europe: Deservingness perceptions among citizens of European welfare states. Journal of European Social Policy 16(1): 23–42.

75.

Voors

Maarten J

Nillesen

Eleonora M

Verwimp

Philip

Bulte

Erwin H

Lensink

Robert

Van Soest

Dan P

(2012) Violent conflict and behavior: A field experiment in Burundi. American Economic Review 102(2): 941–964.

76.

Whitaker

Beth Elise

(2003) Refugees and the spread of conflict: Contrasting cases in Central Africa. Journal of Asian and African Studies 38(2–3): 211–231.

77.

Whitt

Sam

Wilson

Rick K

(2007) The dictator game: Fairness and equity in postwar Bosnia. American Journal of Political Science 51(3): 655–668.

78.

Wike

Richard

Stokes

Bruce

Simmons

Katie

(2016) Europeans Fear Wave of Refugees Will Mean More Terrorism, Fewer Jobs. Washington, DC: Pew Research Center for the People & the Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.44 MB