Religion, securitization and anti-immigration attitudes: The case of Greece

Religion, securitization and anti-immigration attitudes: The case of GreeceAuthor(s): Georgios Karyotis and Stratos PatrikiosSource: Journal of Peace Research, Vol. 47, No. 1 (january 2010), pp. 43-57Published by: Sage Publications, Ltd.Stable URL: http://www.jstor.org/stable/25654527 .

Accessed: 31/07/2014 15:07

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Sage Publications, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Journal of PeaceResearch.

http://www.jstor.org

This content downloaded from 134.153.184.170 on Thu, 31 Jul 2014 15:07:02 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=sageltd

http://www.jstor.org/stable/25654527?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


journal of

r> 1 j.' 1 RESEARC H Research article

Religion, securitization and

anti-immigration attitudes: The case of Greece

Georgios Karyotis & Stratos Patrikios

Department of Government, University of Strathclyde

Journal of Peace Research 47(1) 43-57 ? The Author(s) 2010

Reprints and permission: http://www. sagepub. co. uk/j ournals Permission. nav DOI: 10.1177/0022343309350021

jpr.sagepub.com

?)SAGE

Abstract

This article revisits securitization theory of the Copenhagen School by addressing an empirical overemphasis on political actors

and offering a quantitative extension to typically qualitative assessments of the theory. Using Greece as a case study, it explores

the dynamics of competition and the relative discursive power of two actors, political and religious elites, regarding migration. After first documenting

a divergence in the two actors' rhetoric through discourse analysis, it proceeds to measure the relative

impact of their discourses on public immigration attitudes, employing structural equation modelling of European Social Survey data. Findings demonstrate that exposure to the securitizing religious discourse through church attendance immunizes citizens from the softening effect of the political message. This, in turn, explains the survival of the security frame on migration in Greece, even as political elites begin to move towards the desecuritized pole of the continuum. Crucially, the analysis of this case suggests that a methodological synthesis of qualitative and quantitative research methods to study securitization is possible despite limita tions. The authors call for greater efforts to combine the two methods which would allow for a better understanding of secur itization and desecuritization processes.

Keywords

Greece, migration, qualitative-quantitative synthesis, religion, securitization

Introduction

Since the 1980s and particularly after the terrorist attacks of 11

September 2001, migration in the European Union has been viewed through a security prism (Huysmans, 2000; Karyotis, 2007). Migration is perceived by many as a threat to the iden

tity, safety, economic development and, ultimately, the quality of life of citizens, who have been alarmed by the recent arrival of third-country nationals in the European area. In explaining

high levels of xenophobia, qualitative approaches have focused on the ways discourses of danger

are constructed, with most

prominent among them the security framework, developed by the Copenhagen School (Buzan, Waever & de Wilde, 1998).

The Copenhagen School argues that instead of examining security as something tangible, scholars should consider the

process by which actors construct issues as threats to security, a process they call 'securitization' (Waever, 1995). The social construction of security is typically associated with political elites presenting an issue as an existential threat, although in

issues such as migration, other societal actors may also be

involved in the process by supporting or opposing political discourse. Among those, religious elites are likely to be very influential, particularly in countries where national and reli

gious identities are closely interwoven.

Despite the prominence of the Copenhagen School

approach in security studies, only a few attempts have been

made to study the securitization of migration empirically by analysing political discourse (e.g. Buonfino, 2004; Ibrahim, 2005). Even scarcer are attempts to systematically analyse reli

gious discourse on migration, although religiosity is itself often linked to prejudice (Hunsberger & Jackson, 2005). This

emphasis of empirical work on political discourse, thus, fails to capture the complexity of a process where multiple actors

may develop complementary or competing discourses on

migration, the relative impact of which cannot be determined with the use of discourse analysis alone.

Our aim in this article is to fill a lacuna in securitization

theory by incorporating in a single analytical framework both

political and non-political actors and evaluating how their

potentially antagonistic discourses materialize into public acceptance or rejection of a (de)securitization move. We

hypothesize that in cultural contexts where religion holds a

prominent place in the public sphere, the influence of religious elites on public attitudes may even outweigh that of political elites. This is because while political actors usually construct rational - or at least apparently rational -

arguments in an

appeal to logic, religious actors require a leap of faith' on

Corresponding author:

Georgios Karyotis & Stratos Patrikios, University of Strathclyde, Department of Government, 16 Richmond Street, Glasgow Gl 1XQ, UK. Email:

[email protected] and [email protected]



44_journal of'Peace Research 47(1)

behalf of the individual, since faith discourses by definition make universalistic claims of primacy and infallibility (Laust sen & Waever, 2000).

To investigate this hypothesis and measure the relative

impact of each discourse, we propose a methodological

advancement of securitization theory with the inclusion of

public opinion evidence to complement discourse analysis.

Although measuring securitization is faced with inherent lim itations - not least the simplifying character of most quantita tive analyses

- we consider this a first step towards bridging the theoretical gap between rhetorical constructions and public threat perceptions (see Stritzel, 2007).

Greece has been chosen as a critical case study for this

attempt (Yin, 1994) because of its dual characteristic as the most religiously and ethnically homogenous country, with also the most negative attitudes towards immigrants in the EU

(European Social Survey, Round 1). The role of the Greek Orthodox Church as guardian of national identity increases the prospects for the involvement of religious elites in the securitization of migration, which is a prerequisite for our

analysis. Similar patterns may be observed in other countries

where ethnicity and religion overlap (e.g. Serbia, Israel, Cyprus and Poland).

Our argument is developed progressively over three inter

connected parts, with the use of quantitative and qualitative methods. The first part engages with the theoretical framework of the Copenhagen School and explains its application here. The second part examines the discourse of political and reli

gious elites on migration in Greece. Discourse analysis reveals

a softening of the political message in the new millennium, which however found opposition from the Orthodox Church. The final part attempts to quantify and measure the impact of

religious discourse on immigration attitudes in Greece. Using structural equation modelling of data from the first two rounds of the European Social Survey (ESS 2002, 2004), we find not

only that religiosity is a strong predictor of anti-immigration

attitudes, but also that exposure to the religious discourse

immunizes churchgoers from the softening effect of the polit ical message. Results unveil the importance of a

non-political actor in the securitization process, with implications for future

immigration policies in Greece, as well as for securitization

theory.

Constructing (in)security

In light of debates calling for a redefinition of the concept of

security, the Copenhagen School has offered one of the most

influential alternatives to the state-military approach pre

scribed by realism. The cornerstone of this approach is the

concept of securitization, a process that occurs when a political

actor pushes an area of'normal politics' into the security realm

by using the rhetoric of existential threat. In other words, securitization is the process through which an issue becomes

a security one, not necessarily because of the nature or the

objective importance of a threat, but because the issue is

presented as such. Therefore, according to Waever (1995:

55), we can regard security as a

speech act. 'the word security

is the act', pronounced as such by elites in order to produce hierarchical conditions in which security issues are dramatized

and presented as supreme priorities of the state or the actor in

question. To study securitization, the central question that needs to

be addressed is 'who can "do" or "speak" security successfully, on what issues, under what conditions, and with what effects'

(Buzan, Waever &; de Wilde, 1998: 27). The Copenhagen School refers to those actors that 'are placed in positions of

power by virtue of being generally accepted voices of security' as

securitizing actors, whereas actors such as the media that

popularize the security discourse are referred to as functional

actors. In the case of migration, the securitizing actors are typi

cally the 'holders of the collective identity' who value the

importance of safeguarding and reproducing a

language, a set

of behavioural customs, or a conception of ethnic purity

(Buzan, Waever & de Wilde, 1998: 23). The success of a secur itization move depends on their ability to specify a threat to a

collectivity and mobilize a 'we' against a

supposedly threaten

ing 'them'. In turn, such discourses of danger draw commu

nities together by emphasizing the differences between members of the community and those on the outside, which

implies that the construction of the 'Other' is inseparable from how the self is understood (cf. social identity theory and self

categorization in Tajfel & Turner, 1986).

Although the scholars of the Copenhagen School argue that in principle, nobody is excluded from becoming a securitizing actor, their view of security as a 'structured field' (Buzan,

Waever & de Wilde, 1998: 31) suggests that political and secu

rity elites are more likely to have the knowledge and social cap ital to write legitimate security discourses.1 However, religious elites can also be involved in the securitization of migration.

Religious teachings have been found to promote 'attitudinal

conformity' among congregants by stressing the importance of some values and denigrating the importance of others

(e.g. Wald, Owen & Hill, 1988). Additionally, the authority of religious elites is transcendental, since their messages are

considered to be direct interpretations of God's wisdom

(Laustsen & Waever, 2000). Consequently, religious actors can

themselves be powerful discourse entrepreneurs and therefore

influential in securitization processes.

The content of their message and the relative ability of reli

gious elites to influence immigration attitudes are context

dependent. Previous research has shown that religion can both reduce and exacerbate prejudice and insecurities towards the

'Other', although all religions include 'love thy neighbour' principles. A review of the literature demonstrates that 37 of

47 studies published between 1940 and 1990 showed a posi tive relationship between religiosity and prejudice, while just

1 For this reason, the Copenhagen School has been criticized for being state

centric (McSweeney, 1999).



Karyotis & Patrikios 45

two studies indicated a negative relationship (Hunsberger &

Jackson, 2005). Our point of departure in this article is to uncover the

under-researched role of religious actors in securitization the

ory, as already suggested by Laustsen & Waever (2000). The case we have selected for our analysis provides

a most-likely

setting for the observation of this phenomenon, that is, the involvement of the Greek Church in the securitization of

immigration. In the period under examination, a national

religion is already involved in political debates under the

leadership of a vocal prelate (Archbishop Christodoulos), while there is a noticeable influx of non-Christian Orthodox

immigrants into the country, which makes religious authority relevant in migration debates.

For these reasons, in the second part of this article, apart from the political discourse on

migration, we also analyse the

discourse of the Orthodox Church in Greece. However, the inclusion of two actors, political and religious elites, in a single analytical framework raises two further questions, one theore

tical and one methodological, relevant to our application of

securitization theory that need to be addressed.

The theoretical point has to do with the Copenhagen School's understanding of desecuritization as the optimal

long-term option for societal issues. Desecuritization is the

reversal process of securitization, referring to 'the shifting of

the issues out of the emergency model and into the normal

bargaining process of the political sphere' (Buzan, Waever & de Wilde, 1998: 4). However, the exact process through which this occurs is not clearly explained by the Copenhagen School.

A change in the political rhetoric, even if it aims to challenge the security discourse, may fail to reduce public threat percep tions towards an issue, since the debate continues to be con

ducted in security terms, thus reinforcing previously internalized fears (Huysmans, 2002). Furthermore, even if the

social power relations that facilitated the securitization of an

issue have changed, other actors not previously involved in the

construction of the security discourse may continue to

reproduce it. Bigo (2002) refers to the example of 'security professionals' such as police officers that are opposed to the

desecuritization of migration because this would challenge their professional legitimacy. Similarly, we expect that reli

gious elites may also sustain the security frame on migration,

despite moderation of the political message, particularly when

they perceive themselves as guardians of national identity. The

Greek case highlights the difficulties in desecuritizing a societal issue such as migration, where multiple actors may articulate

competing discourses, a process that remains under-theorized

(see Roe, 2004). The potential for competition in the discourses of the two

actors examined here raises a second question about the meth

odology used to study securitization. The 'obvious method' for the Copenhagen School would be discourse analysis, since securitization is 'a specific rhetorical structure that has to be

located in discourse' (Buzan, Waever & de Wilde, 1998: 176). The underlying assumption is that language is structured

into patterns that play an important role in shaping

our under

standing of reality. Foucault (1986) explains that the partici pants in the discourse attempt to present their view as

'objective' or 'true'. When such presentations of reality achieve

an authoritative status and become dominating, then discourse

becomes socially influential and appears as the only right and

possible perspective.

However, discourse analysis can only point towards the

influence of a securitizing actor without being able to gauge the impact of the security language on public threat percep tions. The only option available to the Copenhagen School is to assess acceptance of a

securitizing move based on the will

ingness of an audience to tolerate exceptional policy responses

and violations of rules that would otherwise have to be obeyed (Buzan, Waever & de Wilde, 1998: 26; on the problematic vagueness of this definition, see Jackson, 2006 and Stritzel, 2007). To overcome this limitation, we propose the inclusion of public opinion evidence to complement discourse analysis in studying securitization. The qualitative-quantitative synth esis will allow us to examine the securitization of migration from more than one

standpoint and determine the relative

impact of competing religious and political discourses on pub lic attitudes. This model is further developed and explained in the third part of this article. Before that, the next section dis cusses the securitization of migration in Greece, analysing

political and religious discourses.

Securitization of migration in Greece

For the most part of its modern history, Greece had been a

country of emigration. However, following the collapse of communist rule in Eastern Europe, Greece very rapidly and

unexpectedly became host to a great number of immigrants.

According to the 2001 census, 7.27% of the total population in Greece was

composed of foreigners, with unofficial figures

provided by NGOs and the scientific community raising the

immigrant population to 10%. Albanians constitute the

largest immigrant group (65%), followed by the Kurds, Pakistanis and Afghans. Although no official data exist on the

religion of immigrants, Mehmet Imam, the President of the Panhellenic Federation of Supporting the Muslims in Greece claimed that more than 200,000 Muslim immigrants lived in Athens alone (Antoniou, 2003: 166).

A survey by the European Parliament (1985: 43-44) on the rise of fascism and racism in Europe in 1985 showed that Greeks were 'tolerant and xenophilic and generally free of racial prejudice'. However, as

immigration to Greece grew, so did public anxieties and insecurities towards immigrants. For instance, a national survey in 1993 revealed that 90% of

respondents thought that immigrants take jobs from Greeks, while 84% considered them a public danger (Kiprianos, Balias & Passas, 2003: 154). The Copenhagen School framework can shed light on the techniques employed to construct the

security discourse on migration in Greece.



46_journal of 'Peace Research 47(1)

Political discourse Official discourse concerning migration in Greece became

highly securitized in the early 1990s. Discourse analysis reveals a strong offensive language towards immigrants, which served

as the main legitimizing factor for restrictive policy responses. The Greek political and security elites used symbolic language, metaphors, exaggerations, inaccuracies and a criminalization of

the 'Other' in order to actively promote the construction of

migration as a threat, as opposed to a multidimensional social

phenomenon. The foundations for the securitization of migration in

Greece were laid down with the introduction of the 'Law for Aliens' of 1991 (Law 1975/1991), which came to replace the

previous law dating back to 1929. Presenting the law to the

parliament, the government explained the security reasoning behind its introduction by emphasizing the threats that

migration posed to the economy, society and Greek national interests. Pointedly, its sponsor used the word 'problem'

twenty-eight times in total:

There are many problems that our country is faced with

because of the mass migration of these people to our country.

These are social problems; employment problems; health

problems; criminality problems, that we are all witnessing

every day.2

Such linguistic representation of migration was common in

the political debate during the 1990s, holding immigrants responsible for many social problems and assuming that it is common sense that the 'aliens' constitute a serious threat

('we are all witnessing every day'). Typically, political discourse

exaggerated the number of legal and irregular immigrants in

Greece, their participation in crime and their burden on the

economy. For instance, Minister of Public Order Stelios

Papathemelis noted in 1993 that 'the indicators of Albanian

criminality are increasing constantly' (Karydis, 1996: 130). Two years later, his successor, Sifis Valirakis, argued that

'aliens are responsible for an increase in the crime rates, which

for some types of criminality have risen up to 40%' (Karydis, 1996: 131). Criminology experts have repeatedly rejected these claims. Yet, such statements coming from the highest

authority in the Public Order Ministry should have had a

securitizing impact on the public mind. Ultimately, the secur

itization of migration in Greece was reflected in the policies of the state that were primarily restrictive, concerned with addres

sing short-term needs and inconsiderate towards the human

rights of immigrants (Baldwin-Edwards & Fakiolas, 1999).

Unsurprisingly, content analysis reveals that during the

1990s, a large part of the Greek media adopted a similarly

xenophobic standpoint (Triandafyllidou, 2002). The sensa

tionalist coverage presented migrants as people who come to

Greece only to exploit available opportunities, emphasizing

2 Parliamentary Proceedings, Greek Parliament, Session B, 10.10.1991, p. 35.

also their responsibility for the crime wave and thus contribut

ing to a particular reinforcing slant to anti-immigrant senti

ments (Mitropoulos, 1999). However, the role of the media in constructing the security discourse was subordinate to that

of the political elites. Triandafyllidou (2002: 156) noted that

normally in Greece, news related to migration is reported from

the perspective of dominant political actors, such as the gov ernment, state authorities or the police. In that way, the Greek

media was a transmitter of political discourse, objectifying claims that immigration is a threat with reference to official statements and the official version of events, even when these

were often inaccurate. Therefore, in line with the Copenhagen

School, the Greek media was not a securitizing actor per se but

a 'functional actor', who significantly influenced the securitiza

tion of migration by popularizing the security discourse and

reproducing negative stereotypes.

The security discourse on migration remained almost

unchallenged until the end of the 1990s, when various factors contributed to a gradual softening of the political message. These factors included, among others, the realization of economic benefits from immigration, the improvement of

bilateral relations between Greece and neighbouring, sending countries and the intense pressure from NGOs to improve the human rights conditions of immigrants. Initially, there was no

consistency in the new approach, with some political elites

continuing to reproduce the security rhetoric (Dimitras,

1999; Triandafyllidou, 2000). However, the moderation of

political discourse became increasingly crystallized when the

government prioritized boosting the international image of Greece as a modern, progressive democracy in order to

maximize the benefits from hosting the 2004 Olympic Games

(Lesser et al., 2001). As a result, political discourse on migra

tion, previously focusing almost exclusively on the threat that the 'aliens' posed to the state and society, started to decon

struct some of the myths that had rendered immigrants to be dangerous and inferior, a shift that was also reflected in media coverage (Triandafyllidou, 2002).

For instance, Public Order Minister Michalis Chrysochoi dis admitted that 'the criminality of foreigners has not risen as

much as it is believed' (Dimitras, 1999), while in 2003, Labour Minister Dimitris Reppas highlighted the benefits of

immigration to the economy. He argued that these can be

found 'in the restructuring and resurgence of the agricultural sector, the reinforcement of mainly small and medium-sized businesses ... as well as in the increase of contributions to the

social insurance funds from legal immigrants' (Tzilivakis, 2003b). Similarly, Anna Karamanou, member of the Eur

opean Parliament for socialist PASOK, depicted migration as a 'natural' phenomenon that could counterbalance the coun

try's ageing and declining population and fill in gaps in the labour market, since migrants generally do 'jobs most Greeks do not want to do any more' (Tzilivakis, 2003a). Prokopis Pavlopoulos, Interior Minister for the newly elected New

Democracy government confirmed in 2004 that 'forging a

substantial and effective immigration policy within a modern



Karyotis & Patrikios_47

framework and with respect to human rights is this govern

ment's top priority' (Tzilivakis, 2004). Tellingly, a central theme of New Democracy's electoral campaign in 2000 was

immigration, framed as a threat, whereas such messages were

omitted from the 2004 campaign (Mantouvalou, 2005:

134-135).

Religious discourse This section turns its attention to the discourse of religious elites,

focusing mosdy on the period after the turn of the century, when the political message became more moderate towards immi

grants. The weight of religious discourse in the public debate on immigration can be understood in the context of two factors.

First, although the Greek constitution attempts to guarantee the

de jure separation of church from state, the Orthodox Church continues to be very much identified with the state, both symbo lically and effectively (Kyriazopoulos, 2001; Anastassiadis,

2004). More specifically, the constitution identifies Eastern

Orthodoxy under the authority of the autocephalous Church of Greece as the 'prevailing religion' of the Greek population (Article 3) and grants the Church legal privileges (Art. 18)

compared to all other 'known religions, namely Judaism, Islam, Catholic and Protestant denominations' (Art. 13).

Second, Eastern Orthodoxy has consistently overlapped with Greek national identity. During the Ottoman occupation of Greece, the Orthodox faith was the distinguishing charac teristic emphasized by an emerging nationalist movement.

This was a useful platform for differentiation of the local pop ulation from Muslim authority, equating commitment to the

Orthodox dogma with Greek identity (Martin, 1978; Mavro

gordatos, 2003). Following 19th-century independence, the first constitutional texts of modern Greece did not distinguish between the notions of 'Greek citizen' and 'Greek Orthodox

Christian', thus highlighting 'the crucial role of Orthodoxy in identifying "Greekness" in a rather exclusionary manner'

(Chrysoloras, 2004: 40). The classification of members of the Greek minority in Albania as immigrants entitled to special privileges, based exclusively on religious grounds (rather than

linguistic or ethnic), serves as one of many examples of the

continuing treatment of religion as a substitute for nationality

(Mavrogordatos, 2003).

Religious discourse became openly politicized after the rise of a dynamic Archbishop, Christodoulos, who succeeded the moderate Seraphim as the head of the Greek Church in 1998 (Stavrakakis, 2004). In response to the government's

modernization and Europeanization agenda, Christodoulos

sought to redefine public debate in a purely religious/national ist context. His high popularity and his claim to speak 'both in the name of the people (and not in the name of Orthodoxy), as

well as in the name of God', gave his discourse 'an unusual sta tus of infallibility and influence' (Chrysoloras, 2004: 47).

More importantly, Christodoulos' discursive construction of

a social antagonism between 'the people' and a wide range

of perceived enemies that posed existential threats to them

echoes the key elements of securitization theory. In his inauguration speech and repeatedly after that, Chris

todoulos set the tone, saying that Orthodoxy 'is indeed a con

dition for the very survival of Hellenism. The historical truth is that a fundamental element of the unity and cohesion of our

people is Orthodoxy' (Gilson, 1998). Such rhetoric marked the revival, not of religion as such, 'but rather of nationalism identified with Orthodoxy' (Mavrogordatos, 2003: 130).

Although the Church did not usually target immigrants directly, its nationalist discourse, deriving from a self

declared role as custodian of national identity, was clearly hos

tile to those that did not belong to the in-group, that is, that were not born Greek Orthodox.

For example, in 2000, when the government decided to omit the declaration of faith from state identity cards, the inclusion of which had until then encouraged discrimination

against non-Orthodox individuals, the Church conducted an

unofficial referendum, collecting approximately three million

signatures without, however, managing to overturn the deci

sion (Molokotos-Liederman, 2003). Christodoulos forcefully attacked the government, stating that 'they want to make

Greece publicly pretend to be atheistic so that it will be recog nized as progressive, to appear as if it is denying its history and

traditions, which are Greek Orthodox' (Gilson, 2000). On another occasion, the decision of a school to allow an Albanian

A-grade student to carry the Greek flag at his school parade during a national holiday in 2000 and again in 2003 caused

widespread reaction from members of the local community and the clergy. In response, Christodoulos argued that immi

grants were threatening Greek national identity, stating that

'we are in danger of becoming refugees in our own country'

(Gatopoulos, 2000). The divergence of political and religious discourse on

migration becomes clearer when examining the debate con

cerning the construction of a mosque in Athens. Acknowled

ging that Athens is the only European capital without an official mosque, in June 2000, the government committed to constructing an Islamic cultural centre and mosque in time

for the 2004 Olympics. Foreign Minister George Papandreou argued that the mosque will be built 'in the spirit of the multi cultural democratic Europe of which Greece is a

part' and

explained that 'migration has made the necessity for a mosque even greater, because Athens' Muslim population has got that

much bigger' (Smith, 2003).3 Government planning identi fied an area for the mosque in Peania, close to the international

airport, yet construction never began, largely due to

objections from the Orthodox Church.

3 The inclusion of the law for the construction of the mosque in legislation

concerning the preparation for the Olympics (Antoniou, 2003) suggests that considerations about the country's international image contributed to

this more positive political message on immigration, as discussed earlier.




Religiosity x^^^

/ \ Media /_ \ discourse

Ideology r-^ : i / 1 bJ

1 Economic /

\v evaluations 1 / Socio-demographic

\^^^^^

^ I

Figure 1. Baseline model

Boxes indicate observed variables, while ovals correspond to latent variables with their own indicators. Single-headed arrows represent causal effects. Control

effects, disturbances, exogenous variable correlations, error correlations and surface indicators have been omitted. Economic evaluations are measured via

satisfaction with personal financial circumstances. Ideology is a standard left-right scale. Media discourse is hours spent watching political programmes on TV.

From the socio-demographic controls (age, male, education, size of residence area), education is expected to directly affect immigration attitudes (Adorno, Frenkel-Brunswik & Levinson, 1950). Effects from all control variables are hypothesized on all other endogenous variables (religiosity, ideology, economic

evaluations and exposure to media discourse). Therefore, with the exclusion of education, control effects on the latent variable are assumed to be indirect.

Until 2006, the construction of any religious building in Greece required the permission of the local Orthodox bishop, thus giving the Church a significant say on the issue. Christo doulos and Church representatives repeatedly stated that they did not oppose the construction of the mosque, since 'every one has the right to worship their god and to practise their reli

gion' (Tzilivakis, 2003c). However, the Church raised several

objections along the way, which revealed its uneasiness with

developments. Church spokesman Father Epiphanios noted that we are concerned that the mosque will be the first thing of Greece foreigners visiting the country will see from the aero

planes', which would send the wrong message about predomi

nantly Christian Greece (Tzilivakis, 2003c). A year earlier, in

2002, Father Epiphanios had rejected the idea of a mosque in

downtown Athens 'because the average Greek cannot yet

accept the idea of a minaret in the city centre' (Tzilivakis, 2003c). The Islamic cultural centre provoked further reaction

from Christodoulos, who argued, in an implied reference to

Islamic terrorism, that 'its existence contains dangers which are

known from similar centres in other European countries'

(Smith, 2003).4

4 In 2006, in a sudden change of heart, the Church proposed the construction

of the mosque in Elaionas, near the city centre, a proposal that was accepted by the government.

Religious discourse and securitization: A

quantitative view

The above qualitative evidence indicates an antagonism between religious and political rhetoric on migration during the early years of the millennium. The moderation of the polit ical message, although not supported by adequate policy change (see Skordas, 2002; Fakiolas, 2003), could be inter

preted as a first step towards deconstructing

some of the myths

associated with immigration, thus paving the way for its dese curitization. However, revival of nationalism in church dis

course, with its implied representation of immigrants as a

threat to Greekness, could override the political message. To

explore the relative impact of these competing discourses on

public attitudes, this section introduces a quantitative element

to the study of securitization.

We begin our investigation with a baseline model that will serve as the building bloc for subsequent elaborate formula

tions, more specific to our theoretical concerns. The postu

lated recursive model we employ first evaluates the central

place of religiosity in shaping reactions towards immigration. The expected relationships are illustrated in Figure 1. For the sake of parsimony, this model provides a simplified version of

competing theories regarding immigration attitudes, especially economic and cultural ones. We consider it adequate for

examining the theorized powerful influence of religiosity on




mass belief systems, especially within the framework of a

national religion. As posited in Figure 1, our central anticipation is for a

direct effect of Greek Orthodox commitment on boosting negative feelings towards immigrants, even after controlling for various socio-economic characteristics. The negative bias

of religiosity is also expected to work indirectly through factors identified below as the foundations of anti-immigrant senti

ments in European societies, especially ideological predisposi tion and exposure to media coverage of current political issues.

In short, we treat religiosity as a structuring principle for pub

lic reactions towards immigration threats to the nation, influ

encing, at the same time, other major predictors of the same

phenomenon.

Data and measures

Theorized relationships between political and religious dis courses on the one hand, and anti-immigration attitudes on

the other are evaluated with data from two ESS rounds col lected between September and December 2002 and 2004,

respectively. The ESS is a cross-national survey, with a mini

mum target response rate of 70% and rigorous translation pro

tocols (details in Jowell et al., 2005). The method of administration consists of hour-long face-to-face interviews

with particular emphasis on immigration, citizenship and

socio-political issues. Datasets include representative samples of the Greek population in both rounds.

Crucially, the timing of the surveys offers a rare opportunity to assess our

reading of securitization theory. The ESS team

has administered two measurements in the period under con

sideration: one lies closer to the time when the religious and

political messages were overlapping, while the second covers

a phase when the political message becomes consistently mod erate on immigration threats. We take advantage of this quasi

experimental element in the data and use the tension between

discourses as a testing ground for our hypothesis. Samples con

tain 2,566 cases for the first round and 2,406 for the second.

Throughout the analysis, these numbers are reduced slightly, since the small number of Catholic and Muslim respondents is filtered out. A similar treatment is applied to participants

without Greek citizen status.

Additionally, the ESS is useful because the framing of ques tions on

immigration attitudes direcdy touches upon securitized

discourses of existential threat and survival. This allows us to

construct a composite measure of the main variable in this

analysis, which we arbitrarily tide 'immigration threat'. The latent variable consists of three items, which correspond to the

most commonly observed sources of insecurity towards immi

grants (Ceyhan &: Tsoukala, 2002): identity, the economy and - in want of a better indicator - a generic feeling of well-being. For closer connection with the theoretical discussion, high scores in this new variable, purged from random measurement

error, denote a higher feeling of threat. The exact question wording of this and all variables of interest appears in Appendix

A. The reliability of the underlying construct is acceptable across both rounds, as well as across

subsamples used later in the analy sis (Cronbach's alpha varies between .81 and .89).

The first key independent variable in the analysis is measured

through a standard indicator of religiosity. The survey question asks respondents

to determine how religious they are on a 0 to

10 point scale ranging from 'not at all' to 'very religious'. Alter

native and sometimes more appropriate indicators of religiosity exist in other surveys. Such indicators tap religious salience

(importance of religion in life) or biblical inerrancy (literal inter

pretation of the meaning of holy texts) (see examples in Leege &

Kellstedt, 1993). The ESS, however, has not been designed to cover such religious characteristics. For the same reason, we can

not construct a more reliable, multi-item scale of religiosity. To operationalize the second key independent variable,

that is, political discourse, we use a measure of hours spent

watching political programmes on TV. As discussed earlier, the role of the media in securitizing an issue is viewed as sub ordinate to the political elites. This is because journalists, 'blinded by the myth of objectivity', are often reduced to a role of reproducing the discourse of the dominant ideology (Erja vec, 2003: 99). In the case of migration, journalists tend to

objectify claims about whether an issue is an existential threat or not with reference to the official version of events, as pre sented by state institutions. In that way, the media become

a transmitter of the political elites' discourse, which they legit imize with supporting evidence from the officials that created the discourse in the first place.

Following the relevant literature, we also consider other pos sible influences on

immigration attitudes. Quantitative research,

for example, has found educational attainment to be associated

with greater tolerance towards racial and cultural 'Others' (e.g. Citrin et al., 1997; McLaren, 2001; cf. Adorno, Frenkel Brunswik & Levinson, 1950). Lower income strata and those in lower status (more vulnerable) occupations will generally feel threatened by immigrants, seeing them as competitors in the

low-wage job market and in social welfare services (Fetzer,

2000; Scheve & Slaughter, 2001). Right-wing ideology is typi cally associated with a more

inward-looking perspective on the

nation, which predicts negative attitudes towards newcomers

(Chandler & Tsai, 2001). The social contact thesis also claims that exposure to out-group members (e.g. friendship with immi

grants) helps counter negative stereotyping of such out-groups

(Hayes & Dowds, 2006). Finally, men tend to see immigrants as a threat since they participate in the economy more than

women (Hemes & Knudsen, 1992).5 In order to explicate the vertical constraint exerted by religios

ity on a host of other political features and eventually on

5 The combined use of data from two rounds of the ESS does not permit a test

of the social contact hypothesis. Also, owing to the absence of a traditional class cleavage in Greek society, we have excluded occupational and income

indicators. However, we have included a question on satisfaction with

personal financial circumstances (also see note in Figure 1).




Table I. Religiosity and securitization effects

Endogenous variables (standardized solution)

Round 1 Round 2

Ideology Media exposure Immigration threat Ideology Media exposure Immigration threat

Religiosity .279** .037 .070** .256** -.107** .163** Education -.215** -.102**

Ideology .072** .113** Economic evaluations ?.073** ?.181** Media exposure ?.027 ?.053*

Sq. multiple correlation .094 .132

Summary statistics

N 2371 2229 X2 / df 88.341/20 (4.417) 70.246/20 (3.521)

IFI, CFI, RMSEA .987, .987, .038 .992, .992, .034

Source: ESS Rounds 1 & 2. Maximum likelihood estimates. Estimates for disturbance correlations, latent variable measurement, unanalysed relationships and control effects for age, gender and size of residence area are omitted for clarity. *p < .05, **p < .01.

increasingly anti-immigrant attitudes, we opt for a structural

equation model, estimated with full-information maximum like

lihood by Amos 6.0 (Arbuckle, 2005). This choice makes the use of the latent endogenous variable straightforward and exemplifies the hierarchical, direct and indirect role played by Orthodox commitment in the belief system of Greek citizens. This role would have been obscured by a single regression analysis of the kind used in most treatments of immigration attitudes (although it could be explored in a more complicated alternative by a bloc recursive model; for example, see Miller & Shanks, 1996, Appen dix D). Since the present analysis brings attention to the founda tional effect of religiosity in Greece, the relationships between

religiosity and economic evaluations, but also among ideology, media discourse and economic evaluations were left unanalysed, without assumptions made on the direction of causality (i.e. only correlations or error correlations between the aforementioned

variables were implied in the models). Goodness of model fit is evaluated with the following con

ventional criteria (Arbuckle, 2005): the %2 test / degrees of freedom ratio, in which values less than 5 are desirable with

large samples; the incremental fit index (IFI), which makes

adjustments for sample size and the complexity of the model,

taking into account degrees of freedom, should score close to .90 and above; the comparative fit index (CFI), which again accounts for small sample sizes and should be greater than

.90; finally, the root mean square error of approximation

(RMSEA), where values should be lower than .08.

Results from the baseline model The main results from the structural equation model are sum

marized in Table I. To begin, the model obtains an acceptable fit to the data from both rounds. It becomes immediately evi dent that the effect of religiosity does not vanish, even with the

application of standard demographic controls. Moreover, this

effect also works indirectly through ideological predisposition in both cases (p = .279 in 2002; p = .256 in 2004), providing a strong clue as to the constraining role of Greek Orthodoxy regarding socio-political attitudes.

However, we identify two differences between the two

rounds, which we consider as critical for further empirical ela boration. Although only two snapshots are available, the direct effect of religiosity on perceptions of immigrant threat seems to strengthen with time. For example, while the influence of

religious commitment on anti-immigrant feelings ranks fourth

in strength compared with other predictors in 2002 ((3 =

.070), it is the second most powerful in 2004 (P = .163), only lagging behind economic evaluations ((3 = ?.181).

Second, with the progress of time, the impact of religiosity also begins to work through exposure to media discourse, which in the case of immigration we take to reproduce political messages, as discussed earlier. In 2002, the effect of religious commitment is irrelevant to experience of political discourse

in the media (P = .037, p > .05). The effect of commitment on political exposure through TV is much stronger in 2004, with the more religious avoiding political discourse in the media ceteris paribus (P = -.107). Although we can only spec ulate at this point, our coverage of the divergence between reli

gious and political discourse in the period covered by the data directs us to a plausible conclusion. This divergence could somehow explain why the more religious begin to avoid

political messages. We examine this in more detail in the next

part.

The same argument serves as an explanation for the appear ance of the neutralizing (negative) effect of media exposure on

anti-immigration concerns, as we move from 2002 to 2004.

The absence of this relationship in 2002 should be related to the cancelling out of contradictory political cues from the

media (P = -.027 at p > .05). At the time, political pro grammes on TV must have been disseminating a mix of




Table II. The effect of church discourse

Mean (standard deviation) Variable -

Round 1 Round 2

Freq. Infreq. Difference Freq. Infreq. Difference attenders attenders (t-test) attenders attenders (t-test)

Immigration: cultural threat 6.79 (2.29) 6.35 (2.55) -.43**a 6.53 (2.27) 6.24 (2.50) -.29**a

Immigration: economic threat 6.83 (2.55) 6.21 (2.72) -.62**a 6.48 (2.37) 6.02 (2.59) -.46**a

Immigration: life threat 6.87 (2.24) 6.58 (2.25) -.29** 6.85 (2.19) 6.57 (2.33) -.28**a

Source: ESS Rounds 1 & 2. * p < .05,

** p < .01. aUnequal variances assumed (Levene's statistic significant at p < .05).

securitizing and de-securitizing speech acts, confounding the

effect of political discourse on attitudes towards immigration.

However, with political discourse in the media becoming dif ferentiated from religious discourse, its hypothesized causal

path should make a difference regarding public attitudes. It seems that this divergence is what lies behind the softening (negative) effect of political discourse on anti-immigration concerns ((3

= -.053).

The conditional effect of religious discourse Results hitherto provide a general impression of the power of Greek Orthodox commitment to affect immigration attitudes

through a securitizing perspective, above and beyond other influences usually considered more important in cross

national research. Our attention to generic religious commit

ment, however, has so far ignored the specific emphasis placed by the Copenhagen School on the discourse of relevant actors.

We now present a more direct test of our expectations for the

role of the church in the Copenhagen School framework, argu ing for an empirically overlooked process. Specifically, while the security logic was an endemic pattern in the discourse of Greek political actors on migration for most of the 1990s,

qualitative evidence presented above suggests that this frame

work is initially supplemented but later perhaps overrun by the

securitizing religious discourse since the ascendance of Arch

bishop Christodoulos. As noted, religious discourses typically make claims of abso

lute and infallible principles that demand supremacy over all other discourses, including political ones. Based on indications

presented in the above analysis, we contend that exposure to

discourse within a hierarchical religious organization like the Orthodox Church is a very potent mechanism, essentially making the church a political community alternative to the actual political sphere (Wald, Owen & Hill, 1988, 1990). Our

expectation is that with the distancing between religious and

political messages on migration, religious communications

should take over the minds of exposed citizens to the detri ment of external political cues.

In detail, during the period when the security logic was still

dominating political rhetoric, the religious discourse would

not clash with the political message transmitted through the media. As the political discourse becomes more neutral, we

would anticipate organized religion to inculcate its congrega tion from this - literal -

political correctness. The contextual

phenomenon described here represents a moderating effect,

which we base on frequency of church attendance, a practice

well established in the relevant literature (Leege & Kellstedt, 1993). Frequent attendance will be taken as an indication of

wider exposure to the church discourse, which should 'protect' citizens from the opposing messages originating in the political sphere. To test this expectation, our baseline model will now

be re-estimated for two subsamples, frequent and infrequent

attenders, across the two time points (the cutoff point is

'monthly or more frequent attendance', which demarcates the

frequent attenders group; see Appendix A).6 The zero-order relationships in Table II suggest that expo

sure to church messages leads to consistently stronger anti

immigration feelings. Frequent attenders are clearly more

prone to internalize the securitization of migration compared with infrequent attenders, as evident by the higher means

among the former. This is the case for both rounds of data, even with a decreased sample size in Round 2. Furthermore,

inspection of the standard deviations for specific items across the two groups reveals that attitudes among churchgoers

are

persistently clustered around the group mean, more so com

pared to the second group (see also note a in Table II).

Although only at the bivariate level, this points towards the lower diversity of attitudes within this group, that is, towards

greater cohesiveness around the group norm. In other words, we interpret this as illustration of the 'homogenizing' effect that church discourse has on congregant anti-immigration attitudes.

6 The religiosity variable is removed from these models, as a precaution against

selection bias when adding the attendance variable. For instance, the average score of religiosity for frequent attenders in Round 1 is 8.4 (scale 0-10). This also breaks the normality assumption required for structural equation

modelling (Bollen, 1989). Subsamples contain the following number of cases: 1,324 frequent and 1,051 infrequent attenders in Round 1; 1,210

frequent and 1,025 infrequent attenders in Round 2.




Media

discourse

Ideology ._. I-1 Economic

\v evaluations

Socio-demographic \ V

091\ \-.095

Figure 2a. The effect of exposure in 2002 among infrequent attenders

Source: ESS Round 1, using design weight. Estimates are standardized coefficients. Model assumes control effects on ideology, economic evaluations and exposure to media discourse. Also implied are

disturbance correlations between ideology and economic evaluations and between economic evaluations and the conceptually proximate economic indicator of

the latent threat variable. Only statistically significant paths are shown {p < .05). Same specification applies to all subsequent diagrams. N = 1075; X2 I df = 1.884; IFI = .991; CFI = .991; RMSEA = .029.

Building on our initial findings, results from the main multivariate analysis are presented in diagrammatic form in

Figures 2 and 3. As expected in 2002 (Figures 2a, 2b), at a

period without marked antagonism between religious and polit ical rhetoric, messages from the pulpit make no difference as to

individual exposure to political discourse (the effect of exposure to political discourse in the media is insignificant for both sub

samples). All causal paths are roughly similar for both groups, a

potent indication of the trivial influence of church exposure in this phase. We attribute this absence of effects from political

media discourse on immigration attitudes to the overlap between religious and political messages; that is, at a period of similar religious and political messages there would be no expec tation for the latter to make an independent contribution.

A different picture is produced with inspection of the esti mates from the 2004 data. In this period, we already estab lished that official political discourse had departed from the

security rationale maintained by the Church. Figures 3a and

3b show that messages from the pulpit make a perceptible dif ference regarding exposure to political discourse. While infre

quent attenders seem to partly base their attitudes on the

political messages they receive from TV (p = -.092 in

Figure 3a), it becomes clear that the church environment shields congregants from the softening effect of political cues

on anti-immigration attitudes ((3 = -.057, p > .05, omitted

as insignificant in Figure 3b; intergoup differences are signifi cant). These results reveal the supremacy of religious discourse over political discourse, a phenomenon connected to the

divergence in the content of the two. On the whole, findings highlight the crucial role that religious elites play in the sur vival of the security logic on migration in Greece.

Conclusion

Our study revisited securitization theory, by addressing an

empirical overemphasis on political actors and offering a quan titative supplement to typically qualitative assessments of the

theory. The melding of national and religious identity in Greece served as the starting point for arguing that religious elites can be a significant stakeholder in securitization pro cesses, when the church's defence of faith is equated with its

defence of national identity. Using qualitative methods, we

documented a divergence between political and religious dis courses in the early years of the millennium. This, in turn, offered us an invaluable opportunity to gauge the relative

impact of each discourse. The evidence from the quantitative analysis of public threat perceptions reveals that, as

expected, religiosity is one of the strongest predictors of




Media discourse

Ideology ,-. I_I Economic

\ evaluations Socio-demographic \ V

075\ \ _.Q89

Figure 2b. The effect of exposure in 2002 among frequent attenders

Source: ESS Round 1.

N = 1288; x2 I df = 3.140; IFI = .983; CFI = .983; RMSEA = .041.

Media discourse

Ideology : i / I | Economic / \ evaluations /

Socio-demographic \ V /

\ \ ~'?92 \ 134X\ \_.171 /

(Immigrationthreat )

Figure 3a. The effect of exposure in 2004 among infrequent attenders


N = 1027; X2 I df = 2.847; IFI = .987; CFI = .987; RMSEA = .042.




Media discourse

Ideology _

I-_-1 Economic

\ evaluations Socio-demographic \ X

.149\ \ -.207

-.096 (educatic^

Figure 3b. The effect of exposure in 2004 among frequent attenders


N = 1194; x2 / df = 1.914; IFI = .993; CFI = .993; RMSEA - .028.

anti-immigration attitudes in Greece. Significantly, the revival

of nationalism in the discourse of the Orthodox Church under Christodoulos appears to have been detrimental to the domi nance of more moderate interpretations of immigration, even

when political elites supported such alternative discourses.

Admittedly, our ability to establish definite causation between religious discourse and anti-immigration attitudes is

hampered by limitations in the available data. The ESS data used do not allow us to robustly test whether church atten

dance actually causes exposure to religious cues on immigra

tion, although past studies have shown this to be true

(Wald, Owen & Hill, 1988; Jelen, 1992). Future, more

sophisticated research should provide one of two types of pri mary contextual information: survey questions asking parishi oners whether the sermon contains messages on immigration,

and what the nature of those messages is; alternatively, data collected through direct observation of such messages in a qua litative sample of carefully selected congregations. The same

limitation applies to our use of media exposure, which in the absence of better data, only serves as a proxy for exposure to

political discourse.

The qualitative-quantitative synthesis advocated in this

article represents our main contribution to securitization

research. One objection to this could be that we cannot

combine the two methods because securitization is

essentially a discursive process. Securitization theory typi

cally utilizes discourse analysis to identify how existential threats are rhetorically constructed. Evidence that the rele

vant audience has accepted such securitization moves is then

solely sought at the policy level, as the unopposed adoption of exceptional measures. However, despite its limitations,

the incorporation of survey data to supplement discourse

analysis allows for a more comprehensive investigation of

outcomes stemming from competing discourses in at least

two significant ways.

Firstly, by measuring the relative impact of the discourse of different actors on public attitudes, we can assess who the most

influential securitizing actor is. Our analysis highlights the overlooked ability of a societal actor, in this case the Orthodox

Church, to act as an important discourse entrepreneur, even

outweighing political rhetoric in influencing public attitudes on immigration. This finding, albeit perhaps not surprising given the intertwined nature of Greek religious and national

identity, calls for greater attention and empirical research on

the role of non-political actors in securitization processes. Var

iations of our methodological strategy could be used to exam

ine the involvement of other actors, not just religious elites, in the securitization of an issue.

Secondly, by investigating the impact of elite discourse on

public attitudes, we can also partly evaluate why certain




policies materialize over others. Public threat perceptions could indicate the success/failure of discursive securitization

or desecuritization moves, even before policy change is initi

ated. In the Greek case, the relative strength of religious over

political messages on immigration among churchgoers offers a plausible explanation for the persistence of high anti

immigration attitudes. This in turn, should prepare us for

opposition to policy change, even as political elites begin to move towards the desecuritized pole of the continuum. Nev

ertheless, since the temporal interval examined here is rather

short (only two time points), further analysis is needed as data become available, before establishing the existence of a clear trend. The enthronement of the moderate Archbishop Ierony

mos as the Head of the Church, following Christodoulos' death in January 2008, may herald a new era, since Ieronymos, unlike his predecessor, has pledged to avoid interfering in

politics.

Finally, from the above derives a more persistent, theoreti

cal question about the Copenhagen School's understanding of desecuritization. As shown, desecuritization may not depend on a rhetorical deconstruction of existing threat perceptions

by those actors that constructed the security discourse in the

first place (i.e. the political elites). This is because, as with the Orthodox Church, other influential actors in the field may find the maintenance of the 'threat' frame beneficial and thus continue to reproduce it. Additional empirical research and theoretical study is therefore required to explore the full

dynamics of desecuritization as a process.

Replication data The dataset, codebook and do-files for the empirical analysis in this article can be found at http://www.prio.no/jpr/datasets

Acknowledgement An earlier version of this article was presented in the Sympo sium on 'Security and migration: Challenges to the liberal state', at Edinburgh University, 27 August 2007. The authors

would like to thank participants for their feedback, as well as the four anonymous reviewers for the extensive and construc

tive comments.

References

Adorno, Theodor, Else Frenkel-Brunswik & Daniel Levinson (1950)

The Authoritarian Personality. New York: Harper.

Anastassiadis, Anastassios (2004) Religion and politics in Greece: The

Greek Church's 'Conservative Modernization' in the 1990s.

Sciences Po, Research in Question 11.

Antoniou, Dimitris (2003) Muslim immigrants in Greece: Religious

organization and local responses. Immigrants and Minorities

22(2-3): 155-174. Arbuckle, James (2005) Amos 6.0 User's Guide. Chicago, IL:

SmallWaters.

Baldwin-Edwards, Martin & Rossetos Fakiolas (1999) Greece: The

contours of a fragmented policy response. South European Society and Politics 3(3): 186-204.

Bigo, Didier (2002) Security and immigration: Toward a critique of

the governmentality of unease. Alternatives 27(1): 63?92.

Bollen, Kenneth (1989) Structural Equations with Latent Variables.

New York: Wiley.

Buonfino, Alessandra (2004) Between unity and plurality: The

politicization and securitization of the discourse of immigration in Europe. New Political Science 26(1): 23-49.

Buzan, Barry; Ole Waever & Japp de Wilde (1998) Security: A New Framework for Analysis. London: Lynne Rienner.

Ceyhan, Ayse & Anastasia Tsoukala (2002) The securitization of

migration in western societies: Ambivalent discourses and policies. Alternatives 27(1): 21-39.

Chandler, Charles & Yung-Mei Tsai (2001) Social factors influen

cing immigration attitudes. Social Science Journal38(2): 177-188.

Chrysoloras, Nikos (2004) Why Orthodoxy?: Religion and national

ism in Greek political culture. Studies in Ethnicity and Nationalism

4(1): 40-61.

Citrin, Jack; Donald Green, Christopher Muste & Cara Wong

(1997) Public opinion toward immigration reform: The role of

economic motivations. Journal of Politics 59(3): 858-881.

Dimitras, Panayote (1999) Greece: Racist government answers to

racism on the rise. Greek Helsinki Monitor, Athens, 1 August.

Erjavec, Karmen (2003) Media construction of identity through moral panics: Discourses of immigration in Slovenia. Journal of Ethnic and Migration Studies 29(1): 83-101.

European Parliament (1985) Report of the Committee of Inquiry into the Rise of Fascism and Racism in Europe. Strasbourg: Com

mission of the European Communities.

Fakiolas, Rossetos (2003) Regularising undocumented immigrants in

Greece: Procedures and effects. Journal of Ethnic and Migration Studies 29(3): 535-561.

Fetzer, Joel (2000) Public Attitudes Toward Immigration in the United

States, France and Germany. Cambridge: Cambridge University Press.

Foucault, Michel (1986) The Archaeology of Knowledge. London: Tavistock.

Gatopoulos, Derek (2000) School yard row turns into national

dispute. Athens News, 31 October.

Gilson, George (1998) Inspiring Archbishop Christodoulos enthroned.

Athens News, 10 May.

Gilson, George (2000) Archbishop lashes out against arrogance of

'intelligentsia'. Athens News, 15 June.

Hayes, Bernadette & Lizanne Dowds (2006) Social contact, cultural

marginality or economic self-interest? Attitudes towards immi

grants in Northern Ireland. Journal of Ethnic and Migration Stud

ies 32(3): 455-476.

Hemes, Gudmund & Knud Knudsen (1992) Norwegians' attitudes

toward new immigrants. Acta Sociologica 35(2): 123-139.

Hunsberger, Bruce & Lynne Jackson (2005) Religion, meaning and

prejudice. Journal of Social Issues 61(4): 807-826.

Huysmans, Jef (2000) The European Union and the securitization

of migration. Journal of Common Market Studies 38(5):

751-777.



56 journal of 'Peace Research 47(1)

Huysmans, Jef (2002) Defining social constructivism in security studies: The normative dilemma of writing security. Alternatives

27(1): 41-62.

Ibrahim, Maggie (2005) The securitization of migration: A racial dis

course. International Migration 43(5): 163-187.

Jackson, Nicole (2006) International organizations, security dichoto

mies and the trafficking of persons and narcotics in post-Soviet Central Asia. Security Dialogue 37(3): 299-317.

Jelen, Ted (1992) Political Christianity: A contextual analysis. Amer

ican Journal of Political Science 36(3): 692-714.

Jowell, Roger & the Central Co-ordinating Team (2005) European Social Survey 2004/2005: Technical Report. London: Centre for

Comparative Social Surveys.

Karydis, Vasilis (1996) The Criminality of Migrants in Greece: Issues of

Theory and Criminal Policy. Athens: Papazisis [in Greek].

Karyotis, Georgios (2007) European migration policy in the

aftermath of September 11: The security-migration nexus.

Innovation: The European Journal of Social Science Research

20(1): 1-17.

Kiprianos, Pandelis; Stathis Balias & Vaggelis Passas (2003) Greek

policy towards immigration and immigrants. Social Policy and

Administration 37(2): 148-164.

Kyriazopoulos, Kyriakos (2001) The prevailing religion in Greece: Its

meaning and implications. Journal of Church and State 43(3):

511-538.

Laustsen, Carsten Bagge & Ole Wsever (2000) In defence of religion: Sacred referent objects for securitization. Millennium: Journal of International Studies 29(3): 705-739.

Leege, David & Lyman Kellstedt, eds (1993) Rediscovering the Reli

gious Factor in American Politics. Armonk, NY: Sharpe.

Lesser, Ian; Stephen Larraee, Michele Zanini & Katia Vlachos-Dengler

(2001) Greece's New Geopolitics. Santa Monica, CA: RAND.

Mantouvalou, Katerina (2005) Immigration as a security concern:

The case of Greece. Slovo 17(2): 119-136.

Martin, David (1978) A General Theory of Secularization. Oxford:

Blackwell.

Mavrogordatos, George (2003) Orthodoxy and nationalism in the

Greek case. West European Politics 26(1): 117-136.

McLaren, Lauren (2001) Immigration and the new politics of inclu

sion and exclusion in the European Union. European Journal of

Political Research 39(1): 81-108.

McSweeney, Bill (1999) Security, Identity and Interests: A Sociology

of International Relations. Cambridge: Cambridge University

Press.

Miller, Warren & Merrill Shanks (1996) The New American Voter.

Cambridge, MA: Harvard University Press.

Mitropoulos, Dimitris (1999) Immigrants ignite a media maelstrom

in Greece. Nieman Reports 53(2): 36?40.

Molokotos-Liederman, Lina (2003) Identity Crisis: Greece, Ortho

doxy and the European Union. Journal of Contemporary Religion

18(3): 291-315. Roe, Paul (2004) Securitization and minority rights: Conditions of

desecuritization. Security Dialogue 35(3): 279-294.

Scheve, Kenneth &C Matthew Slaughter (2001) Labour market com

petition and individual preferences over immigration policy. Review of Economics and Statistics 83(1): 133-145.

Skordas, Achilles (2002) The new immigration law in Greece: Mod

ernization on the wrong track. European Journal of Migration and

Law 4(1): 23-48.

Smith, Helena (2003) Villagers try to block Athens mosque plan. The

Guardian, 16 September.

Stavrakakis, Yannis (2004) Antinomies of formalism: Laclau's theory of populism and the lessons from religious populism in Greece.

Journal of Political Ideologies 9(3): 253-267. Stritzel, Holger (2007) Towards a theory of securitization: Copenha

gen and beyond. European Journal of International Relations 13(3):

357-383.

Tajfel, Henri & John Turner (1986) The social identity theory of

inter-group behavior. In: Stephen Worchel & William Austin

(eds) Psychology of Intergroup Relations. Chicago, IL: Nelson

Hall, 7-24.

Triandafyllidou, Anna (2000) The political discourse on immigration in Southern Europe: A critical analysis. Journal of Community and

Applied Social Psychology 10(5): 373-389.

Triandafyllidou, Anna (2002) Greece. In: Jessica ter Wal (ed.)

Racism and Cultural Diversity in the Mass Media. Vienna: EUMC,

149-172.

Tzilivakis, Kathy (2003a) A diverse Greece in 2015. Athens News, 24

January.

Tzilivakis, Kathy (2003b) Migrants boost economy. Athens News, 7

March.

Tzilivakis, Kathy(2003c) Christodoulos joins chorus against Peania

Mosque. Athens News, 26 September.

Tzilivakis, Kathy (2004) Minister visits migration think-tank. Athens

News, 15 October.

Waiver, Ole (1995) Securitization and desecuritization. In: Ronny

Lipschutz (ed.) On Security. New York: Columbia University

Press, 46-86.

Wald, Kenneth; Dennis Owen & Samuel Hill (1988) Churches as

political communities. American Political Science Review 82(3):

531-548.

Wald, Kenneth; Dennis Owen & Samuel Hill (1990) Political

cohesion in churches. Journal of Politics 52(1): 197-215.

Yin, Robert (1994) Case Study Research: Design and Methods.

Thousand Oaks, CA: Sage.

GEORGIOS KARYOTIS, b. 1977, PhD Political Science

(University of Edinburgh, 2005); Lecturer, International Rela

tions, University of Strathclyde (2004- ); main research inter ests: international security theory, European migration policy, terrorism and South-Eastern Europe politics.

STRATOS PATRIKIOS, b. 1977, PhD Political Science

(University of Strathclyde, 2008); Lecturer, Quantitative Methods, University of Strathclyde (2009- ); main research

interests: politicization of religion and civic participation.




Appendix A

Question wording and coding of variables

Variable Wording [ESS code] Min. Max.

Religiosity Regardless of whether you belong to a particular religion, how 0 10

religious would you say you are? [RLGDGR] (not at all) (very religious) Attendance Apart from special occasions such as weddings and funerals, 1 7

about how often do you attend religious services nowadays? (every day) (never)

[RLGATND] Immigration: cultural threat And, using this card, would you say that [country]'s cultural life 0 10

is generally undermined or enriched by people coming to live (enriched) (undermined) here from other countries? [IMUECLT]

Immigration: economic threat Would you say it is generally bad or good for [country]'s 0 10

economy that people come to live here from other countries? (good) (bad)

[IMBGECO] Immigration: life threat Is [country] made a worse or a better place to live by people 0 10

coming to live here from other countries? [IMWBCNT] (better place) (worse place) Economic evaluations Which of the descriptions on this card comes closest to how 1 4

you feel about your household's income nowadays? (very difficult) (living comfortably) [HINCFEL]

Ideology In politics people sometimes talk of'left' and 'right'. Using this 0 10

card, where would you place yourself on this scale, where 0 (left) (right) means the left and 10 means the right? [LRSCALE]

Media discourse (exposure to) On an average weekday, how much of your time watching 0 7

television is spent watching news or programmes about (no time at all) (more than 3 hrs)

politics and current affairs? [TVPOL]

Source: ESS.



Documents

Religion, securitization and anti-immigration attitudes: The case of Greece