Everyday security threats: Perceptions, experiences, and consequences 9781526108999

Citation preview

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Everyday security threats Perceptions, experiences, and consequences

Daniel Stevens and Nick Vaughan-Williams

Manchester University Press

Copyright © Daniel Stevens and Nick Vaughan-Williams 2016

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

The right of Daniel Stevens and Nick Vaughan-Williams to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988. Published by Manchester University Press Altrincham Street, Manchester M1 7JA www.manchesteruniversitypress.co.uk

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data applied for ISBN   978 0 7190 9606 8  hardback First published 2016 The publisher has no responsibility for the persistence or accuracy of URLs for any external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Typeset by Toppan Best-set Premedia Limited

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Contents

List of figures and tables Acknowledgements Abbreviations Introduction

page vii ix xi 1

1  Perspectives on security threat politics

14

2  The 2012 study ‘Public Perceptions of Threat in Britain’

40

3  The scope of security threats and their causes

64

4  Security threats and their consequences

96

5  Government, perceptions and experiences of security threats, and citizen involvement in the risk management cycle

130

Conclusion

159

Appendix: coding of variables References Index

167 171 187

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Figures and tables

Figures 2.1 2.2 3.1 3.2 3.3 3.4

4.1 4.2 4.3 5.1 5.2 5.3

Stages of the research The models for estimating the origins and effects of perceptions of threats UK National Security Strategy (2010): priority risks Major threats at the global, national, community, and individual levels Maximum effects of main influences on breadth of threats The effects of mortality salience, authoritarianism, and education on threats from immigration and environmental degradation Maximum effects of breadth of threats Maximum effects of perceived threat from terrorism and immigration Maximum effects of perceived threat from the economy and environmental degradation Coverage of ‘National Security Strategy’ and ‘security policy’ in newspapers, May 2010–June 2012 Example pen portrait used in focus groups (Ajay Bhaskar/Shutterstock.com) Maximum effects of awareness of security strategies/ programmes and the NSS on perceptions of the breadth of threats

50 60 70 76 79

90 111 122 123 133 135

149

Tables 2.1 3.1

Survey sample profile Definitions of ‘security’

57 67

viii

Figures and tables

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

3.2

Influences on perceptions of the breadth of global, national, community, and personal/family threats 3.3 Comparisons of perceptions of the breadth of contemporary security threats at different levels 3.4 Influences on perceptions of the relative breadth of global, national, community, and personal/family threats 3.5 Expected influences on perceptions of specific threats 3.6 Perceptions of contemporary security threats 3.7 Causes of perceptions of threats 4.1 Perceptions of breadth of threats and aggression and intolerance 4.2 Perceptions of breadth of threats and attitudes towards minorities and outgroups 4.3 Perceptions of breadth of threats and attitudes towards spending and taxation 4.4 Perceptions of breadth of threats and voting 4.5 Perceptions of specific threats and aggression and intolerance 4.6 Perceptions of specific threats and attitudes towards minorities and outgroups 4.7 Perceptions of specific threats and attitudes towards spending and taxation 4.8 Perceptions of specific threats and voting 5.1 Awareness of government strategies or programmes to reduce threat 5.2 Influences on awareness of government strategies or programmes to reduce threat 5.3a The association between awareness of government strategies or programmes to reduce threat and perceptions of the breadth of threats (general) 5.3b The association between awareness of government strategies or programmes to reduce threat and perceptions of the breadth of threats (just Tier One) 5.4 The association between awareness of government strategies or programmes to reduce threat and confidence about handling a terrorist attack 5.5 The association between awareness of government strategies or programmes to reduce threat and perceptions of terrorism and the economy as threats 5.6 Muslims and the association between awareness of government strategies or programmes to reduce threat and perceptions of threats

78 81 83 86 86 88 105 106 108 110 112 114 118 120 144 146

148

148

151

152

155

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Acknowledgements

The intellectual origins of this book lie in conversations that the authors first had during 2007 and 2010 at the University of Exeter, Cornwall Campus, about the absence of any serious engagement between literatures produced by International Relations and Security Studies on the one hand and Public Opinion, Political Psychology and Behaviour on the other. Countless open-day events at the Cornwall Campus gave us opportunities to develop these conversations into more structured presentations, and we took the politics of threat as our central focus for discussion. Over the years our double-act was only encouraged by lively audience participation – among prospective students and their families alike – and we became convinced that there was something intellectually and politically valuable not only in working across disciplinary boundaries, but also in engaging in participatory forms of research in politics. The advent of the 2010 UK National Security Strategy and its renewed commitment to citizen engagement provided the catalyst for a joint application in 2011 to the UK Economic and Social Research Council entitled ‘Public Perceptions of Threat in Britain: Security in an Age of Austerity’ (grant number ES/J004596/1) and it is on the basis of this programme of research – and out of those early conversations and open days – that the book has gradually developed. Earlier versions of various chapters were presented at the following events and institutions: the 2012 Elections, Public Opinion and Parties Conference in Oxford; the 2013 International Studies Association Convention in San Francisco; the 2013 British International Studies Association Conference in Birmingham; the meeting of the 2013 Midwest Political Science Association in Chicago; the 2013 International Society for the Study of Political Psychology conference in Herzliya, Israel; the Frontex Second Global Conference and Exhibition on Future Developments of Automated Border Control (ABC) in October 2013 in Warsaw; the University of Queensland in May 2014; the January 2015 meeting of the Ammerdown Group in Somerset; the May 2015 workshop ‘Everyday Insecurities and Vulnerabilities’ at the University of Glasgow; and the June 2015 workshop

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

x

Acknowledgements

‘Strategizing British Foreign Policy’ at the Foreign and Commonwealth Office, London. We are especially grateful to Martin Boon at ICM, Lucy Evans at TNS-BMRB, and all our focus group and survey participants. The following colleagues offered valuable input at crucial stages in the research and writing process: Richard Aldrich, Katherine Allison, Claudia Aradau, Susan Banducci, Victoria Basham, James Brassett, Sarah Bulmer, Stuart Croft, Jenny Edkins, Madeleine Fagan, Chris Federico, Diana Francis, Jamie Gaskarth, Nehemia Geva, Kyle Grayson, Xavier Guillaume, Jef Huysmans, Richard Jackson, Rob Johns, Tom Lundborg, Matt McDonald, Celia McKeon, Cian O’Driscoll, Columba Peoples, Shirin Rai, Michael Saward, Ty Solomon, Erzsebet Strausz, Brent J. Steele, John Transue, and Matthew Watson. Thanks also to our commissioning editor Tony Mason, three anonymous reviewers, and all at Manchester University Press for their encouragement of this project from its inception. Finally, Dan would like to thank Kelly, Daisy, and Lola, and Nick would like to thank Madeleine and Helena for all their support and patience.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Abbreviations

ANES BES CSS DHS FGR IPS IR JCNSS LRF NRR NSS

American National Election Study British Election Study Critical Security Studies US Department of Homeland Security Focus group research International Political Sociology International Relations Joint Committee on the National Security Strategy Local Resilience Forum National Risk Register National Security Strategy

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Introduction

Citizens are increasingly enjoined to occupy a central and active role in the national security architecture of Britain. In the 2008 National Security Strategy (NSS) the UK’s then-Labour government stated its commitment to finding ‘new opportunities to seek views from members of the public’, which was presented as ‘the next step in a process of engagement designed to ensure that government thinking on national security constantly keeps pace with the rapidly evolving global security environment’ (Cabinet Office, 2008: 61). The Conservative– Liberal Democrat coalition government’s 2010 NSS, entitled A Strong Britain in an Age of Uncertainty, reiterated the ‘need to build a much closer relationship between government, the private sector and the public when it comes to national security’ and claimed that ‘we all have a part to play in keeping the country safe – be it from terrorists, cyber attack or natural disasters’ (Cabinet Office, 2010: 5). Continuing this familiar theme, the First Report of the Joint Committee on the National Security Strategy (JCNSS) called for greater ‘public engagement’ in the formulation of national security policy ahead of the next NSS (JCNSS, 2015: 10). Despite the centrality of the citizen in national security policy, however, successive governments have not sought actively to engage the views and experiences of diverse publics in the assessment and prioritisation of issues presented as security threats and risks (Hagmann and Dunn Cavelty, 2012; see also McCormack, 2015; Ritchie, 2011).1 Indeed, Jonas Hagmann and Miriam Dunn Cavelty (2012: 87) argue that the purportedly ‘scientific’ assessment and presentation of issues in the NSS and accompanying National Risk Register (NRR) reflect a ‘distinct security rationality that “depoliticises” security politics’ (see also Leander, 2013). By prioritising issues according to their probability of occurrence and scale of impact at the national level, the NSS presents a seemingly neutral and value-free basis for determining national security policy – one that closes off ‘debates about values, purposes, and formulations of security’ (Hagmann and Dunn Cavelty, 2012: 87).

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

2

Everyday security threats

While the methodology underpinning the NSS and NRR may be read as part of a wider crisis of representation and legitimacy (McCormack, 2015: 2; see also Ritchie, 2011), there are nonetheless various government-backed initiatives designed to enlist the support of citizens throughout society in the risk management cycle (Jarvis and Lister, 2012). For example, posters and announcements in public spaces – such as those associated with the London Metropolitan Police’s long-standing ‘If you suspect it, report it’ campaign – enjoin ‘citizen-detectives’ to be vigilant at all times and to report any behaviour that they deem to be ‘suspicious’ (Vaughan-Williams, 2008). At ports, airports, and international railway stations, a growing number of ‘trusted’ travellers are also expected to interact willingly with biometric technologies such as ‘e-passports’ and ‘e-gates’ in order to facilitate identity-based risk management and reduce waiting times (Amoore, 2009). In a similar vein, Local Resilience Forums (LRFs) invite individuals to feed into local community risk registers in order to identify the greatest risks in a given area and then help to plan and take part in exercises to mitigate against those risks (Adey and Anderson, 2011). Thus, in the 2011 Strategic National Framework on Community Resilience, citizens are called upon to ‘look after themselves and each other for a period until any necessary external assistance can be provided’ (Cabinet Office, 2011: 3). Yet, despite the rhetoric of the NSS and burgeoning expectations that citizens should become stakeholders in and indeed agents of national security, still relatively little is known about how citizens conceptualise and experience ‘threat’ and ‘(in)security’ in the context of their everyday lives, whether they are aware of, engage with, and/or refuse government attempts to enlist them in building societal resilience, and what the implications of these initiatives might be for social interaction among multiethnic publics. International Relations (IR) and Security Studies in both ‘traditional’ and ‘critical’ guises have for the most part privileged the rhetoric, speech acts, and (in)securitizing moves of politicians, policy-making communities, security professionals, private security companies, and other elites (Jarvis and Lister, 2015; McDonald, 2008). In the laudable attempt to make visible diverse governmental logics of (in)securitization, an unintended outcome of this scholarly focus is that the views, repertoires of knowledge, and testimonies of the political subject of (in)security have been rendered largely invisible (Booth, 2007; Gillespie and O’ Loughlin, 2009a; Hansen, 2000; Jarvis and Lister, 2013; Wibben, 2011; see also Walker, 1997). More recently, as we will go on to discuss in Chapter 1, there have been efforts to address this analytical deficit in the context of two so-called ‘turns’ within the literature produced by the subfield of Critical Security Studies (CSS) in particular – ‘the vernacular’ and ‘the everyday’.2 However, these nascent bodies of work – and other related traditions such as standpoint feminist approaches – have developed largely separately from rather than in conversation with each other, which has both perpetuated existing blind spots

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Introduction

3

and also led to new ones – gaps that we wish to identify and start to address in this book. A further lacuna in academic understanding of public perceptions and experiences of everyday security threats exists as a result of the absence of any serious engagement between the IR and Security Studies literature on the one hand and that in Political Psychology and Behaviour on the other. The most direct engagement with IR (e.g., McDermott, 2004) also has a strong focus on leadership, with a particular emphasis on explanations for (often defective) decision-making by elites, including theories of personality, analogical reasoning, groupthink and polythink, and prospect theory (Redd and Mintz, 2013), as well as the role of contextual factors like stress (Levi and Tetlock, 1980). When it comes to individual attitudes and opinion, political psychologists have been more interested in explaining manifestations of perceived security threats, such as group bias, obedience to authority, prejudice, and intolerance, than in perceptions of threats themselves (e.g., Kinder and Dale-Riddle, 2012; Marcus et al., 1995; Milgram, 1963; Stenner, 2005). When they have looked at perceptions of threats, for example from terrorism in the wake of 9/11, it has been of single salient threats rather than security threats writ large (Davis and Silver, 2004; Huddy et al., 2002). It would also be fair to say that research into political psychology and political behaviour tends not to focus directly on how individuals articulate threats as a source for understanding, preferring to draw inferences about understanding from techniques like varying frames or priming different identities. The aim of this book is to redress these lacunae in different subfields of the discipline, make them speak to each other, and in so doing gain a greater understanding of perceptions of security threats, with implications for elite understanding also, than either can accomplish on its own. Terrorism, ISIL (Islamic State of Iraq and the Levant), migration, immigration, weak border control, Ebola, swine flu . . . if we were to examine the British public’s perceptions of security threats at any time during 2005–16, these are the kinds of issues that we might expect to be uppermost in citizens’ minds. After all, they have featured prominently in the news, in some cases they present threats that are mortal, and in others to livelihood, and they may result in the kinds of consequences that are of great concern to individuals. In November 2014, the UK government’s Home Secretary Theresa May described the terrorist threat as greater than at any other time in the country’s history. Indeed, Clarke et al. (2009) argue that the ‘old issues’ of the National Health Service and education have declined in importance in British politics, to be replaced by new security issues pertaining to crime, immigration, and terrorism (see also Whiteley, 2012; Whiteley et al., 2013). Similarly, according to Willer and Adams (2008: 3), while ‘terrorism has become the most significant political issue of the past decade in the United States’, other issues such as immigration, health pandemics like Ebola, and ‘moral’ issues such as same-sex marriage have at times also been at

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

4

Everyday security threats

the forefront of public consciousness. Some argue that perceptions of threat from changing lifestyles have contributed to a restructuring of American politics (Hetherington and Weiler, 2009). It is not only these salient global and national level issues that may be perceived as security threats and that may influence political attitudes and behaviours. Other threats that are closer to home but that receive less attention – such as identity theft – may loom just as large, or larger, in citizens’ everyday lives. At the same time, and as mentioned above, citizens in countries like Britain and the US have been given an unprecedented role in the exercise of security, as in campaigns such as ‘If you suspect it, report it’, a function both of new threats and, we are told, of state security apparatuses’ inability to deal with them as they did with older threats, which is in turn a function of the threat and of austerity and the limits on the tools available. Thus, Mark Rowley, the UK police’s chief counter-terrorism officer, said, ‘the police “cannot succeed alone” in defeating the terror threat and that the public must be vigilant about reporting suspicious behaviour in their area’ (in Mason, 2014).3 It is such perceptions of security threats, from terrorism to identity theft, and the politics of these dynamics at the level of the everyday that are the subject of this book. We explore the full range of issues that members of the British public perceive as threatening to the security of themselves, their community, the nation, and the globe. We ask whether individuals see more threats at some of these levels than at others, what those threats are, their individual-level origins, and also the effects they have ‘downstream’ such as on preferences for more spending that might mitigate security threats as opposed to more spending on other areas such as education. The stakes in the answers to these questions about contemporary threat perceptions are high for governments and citizens alike. While liberal democracies attempt to ‘balance’ civil liberties and security, a threatened public skews the trade-off towards the latter, tending to favour repression, intolerance, and aggressive and exclusionist attitudes towards minorities, towards targets with different political ideologies, and to show a greater willingness to support war against external sources of threat (e.g., Burke et al., 2013; Motyl et al., 2010). A threatened public may also be more receptive to the enhancement of elite power to enact otherwise unpopular or illiberal policies (Bigo and Tsoukala, 2008; Chalk, 1998; Nacos et al., 2011). Indeed, the combination of threat and the belief that elites sanction punitive actions that combat threat is particularly dangerous to democracy. Instead of adapting levels of protection to the perceived existence of threats, it may lead to the modulation of threat perceptions in order to justify enhanced levels of protection, such that protection itself may become a threat (Esposito, 2011: 16). At the extremes, this can result in what Fromm (1941) referred to as ‘an escape from freedom’, which Heymann (1998) more recently described as the undermining of nations’ democratic traditions, and to the

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Introduction

5

temptation to support ‘highly charismatic leaders, such as Juan Peron and Adolf Hitler’ (Merolla et al., 2007: 30). Of course, when authors such as Clarke et al. (2009) refer to new security issues, much of the context is ‘9/11’. The events and aftermath of 9/11 have presented what some analysts consider to be a new world order in terms of threat perceptions: it has ‘resulted in chronic changes to schematic representations of the social world as a dangerous and threatening place for many people’ (Sibley et al., 2007: 368), giving a ‘new urgency to understanding the degree, origins, nature and consequences’ of threats (Huddy et al., 2002).4 The securitization of migration and its increasing association with transnational crime and international terrorism – particularly, though not exclusively, in Europe and North America (Huysmans, 2006; Sniderman et al., 2004) – is just one example of ‘“new politics” and “new security” issues’ since 9/11 (Lahav and Courtemanche, 2012; see also Brader et al., 2008). Key themes and arguments The central topic of the book – everyday perceptions and experiences of security threats among citizens – is among the most salient issues in contemporary politics, transnationally. It attracts significant government spending and animates national security policy, and yet academics and policy-makers alike admit to knowing very little about citizens’ attitudes towards, and experiences of, security. Our book seeks to contribute to a small but expanding body of literature that addresses this pressing knowledge deficit. We deliberately seek to bring different disciplinary perspectives and approaches to the subject-matter. We suggest that the complexity of the topic demands this kind of intellectual pluralism and therefore a certain degree of methodological pragmatism. Thus, we hope that the book demonstrates the enormous potential of this kind of collaboration, which remains unusual in the discipline of Political Science. Transgressing subdisciplinary boundaries means that we are able to cover more substantive ground than extant studies on the theme of security threat politics, which tend to adopt a single orientation and methodological approach and are therefore far more limited in scope and appeal. Indeed, our commitment is to methodological pluralism and a post-positivist epistemological stance beyond the tired and totalising dichotomy of ‘quantitative’ versus ‘qualitative’ perspectives. We argue that this distinction is ultimately unhelpful and that a more productive attitude is one driven by addressing substantive issues and problems in contemporary political life from a range of potentially apposite perspectives. By foregrounding theoretical and methodological challenges posed by the task of studying the politics of everyday perceptions and experiences of security threats, we insert our operating assumptions and their implications into the analysis rather than pretending that these can be ignored.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

6

Everyday security threats

This ethos acknowledges the inability of any perspective to grasp fully the complexities of social realities and proceeds instead by rigorously exploring and evaluating the insights and limitations of various starting points by building self-reflexivity into the research design. Aside from advancing understanding of everyday security threat politics, therefore, a key theme of this book is that of methodology and the practical challenges posed by our substantive topic. Studies on this topic in IR and Security Studies tend to refer to ‘the public’, ‘citizens’, and ‘the everyday’ without actually gathering significant bodies of original empirical evidence from these people or sites, or they tend to rely only on interview or focus group materials. This book, by contrast, combines qualitative and quantitative approaches and triangulates methods in order to gain a broader and deeper understanding of the relationship between public opinion and security threats. Studies on this topic in Political Psychology tend to focus on single threats, without having a broader perspective on threats in general. Again, this book examines multiple threats between and within subjects. We provide several new findings about how diverse multiethnic publics conceptualise, understand, and experience ‘threat’, ‘security’, and ‘security threats’ in their daily lives, and we examine the implications of these for an understanding of their origins and their consequences for a range of public attitudes. Recent attempts to bring ‘vernacular theorizing’ (Bubandt, 2005; Gillespie, 2007; Gillespie and O’Loughlin, 2009a, 2009b; Gillespie et al., 2010; Jarvis and Lister, 2012, 2013, 2015) into the study of security have made significant inroads into addressing the otherwise elitist outlook of the sub-discipline – the excellent study by Lee Jarvis and Michael Lister (2015) is of particular significance for blazing this trail – and it is our hope to contribute to this burgeoning literature. However, the focus of these studies has been primarily on public attitudes towards the specific threat of terrorism, British anti-terrorism legislation, policy, and practice, and experiences of citizenship. By contrast, our study was explicitly designed to approach the question of citizens’ attitudes towards and understandings of ‘security’ and ‘security threats’ from a purposely open and expansive perspective. As such, the range of issues often construed as security threats that we cover in this book includes not only terrorism but also economic security, Islamophobia, racism, and hate crime, to name only a few prominent examples. In summary, this book provides new insights into everyday threat politics by focusing on: 1) the breadth of issues that members of the public identify as security threats – how many, and whether seeing one type of threat is associated with seeing others; 2) the extent to which perceptions of the breadth of threats vary as individuals move from global to personal security threats; 3) the individual-level influences on perceptions of the breadth of threats; and 4) the relationships between perceptions of threats and other political attitudes and

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Introduction

7

behaviours, such as attitudes towards minority groups and intention to vote in an election. In so doing, we make several novel contributions to understanding. First, this is, to our knowledge, the first study to examine perceptions of the breadth of threats rather than specific threats and their intensity. We contend that it provides both a broadened understanding of how threatened citizens think they are, and also a deeper exploration of contemporary understandings of security threats than heretofore. Second, by not confining ourselves to specific threats, we are able to generalise about how individual-level variables, from mortality salience – both a greater awareness of one’s own mortality and feelings of vulnerability – and authoritarianism to media habits, affect perceptions of the breadth of security threats and where and why there is variation. This provides a broader understanding of threats and a firmer foundation for understanding their origins. Third, we extend the levels at which threats are perceived from the national versus personal dichotomy that has dominated the literature to a continuum spanning the individual, family, community, nation, and globe, while also showing the extent to which perceptions of threat at each level have different causes. Fourth, we examine the effects that the breadth of threats at these different levels has on a range of political attitudes and behaviours. Previous research has speculated on why there is variation in such influence but has not examined the range of threats and outcomes that we do and that allow us to pinpoint the causes. Outline and map of the book The book is organised into five chapters. The first chapter explores existing insights into the question of what ‘security threats’ are and how we can study everyday perceptions and experiences of them. In the IR and Security Studies literature the study of the concept of threat has evolved from analyses of the security dilemma between states under anarchic conditions (Waltz, 1979) towards a view that threats are not automatically produced as a result of those conditions (Wendt, 1999). The impact of the social constructivist turn, alongside the broadening and deepening of the security agenda, has meant that threats are now widely seen as produced through dialogue and interaction between states and non-state actors alike. As a result, threats are said not to simply exist independently of our knowledge and representations of them. Rather, as typified by the work of the Copenhagen School, they are brought into being by processes of ‘securitization’ whereby a particular issue comes to be framed in terms of an existential threat in, for example, political speeches and media representations (Buzan et al., 1998). Thus, the concept of threat has been shown to depend on contingent factors such as intersubjective interaction between states (Wendt, 1999) and the (re)production of the identity of an individual state vis-à-vis other

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

8

Everyday security threats

states and discourses of ‘dangerousness’ as inherently political categories of understanding (Campbell, 1992). What has tended to be overlooked, however, is precisely the role of public opinion and everyday views, stories, and experiences in shaping securitizing moves and conditioning their ultimate success and/ or failure (Balzacq, 2005, 2010 Eriksson, 2001; McDonald, 2008). This omission is also characteristic of approaches focusing on the relational economy of (in) securitizing moves (Bigo, 2008; Shapiro, 2013) and the affective politics of contemporary forms of neoliberal governmentality (Adey and Anderson, 2011, 2012; Massumi, 2005; see also Bröckling et al., 2010: 13). Political psychologists have conceived of threats as lying within two principal dichotomies: realistic versus symbolic threat and personal versus sociotropic threat. Realistic threat refers mainly to ‘potential harm to tangible or concrete objects (e.g., money, land, human life), whereas symbolic threat includes various potential threats to relatively abstract aspects of the collective, such as threats to the in-group’s identity, value system, belief system, or worldview (e.g., language, religion, morality)’ (Canetti-Nisim et al., 2009: 464). The distinction between personal and sociotropic threat is between threat to the individual as opposed to threat to a collective such as one’s community or nation. Thus, whereas for IR and Security Studies there has been a tendency to ignore the views of citizens altogether (Jarvis and Lister, 2012, 2013, 2015), we identify two main problems with psychological and behavioural analyses of threat: first, that research tends to focus on discrete security threats, such as from terrorism, immigration, or the environment – limiting understanding of threats in general – and, second, the predominant focus on threats at the national or personal level at the expense of other levels at which threats may be experienced, such as to the community in which a citizen lives. In Chapter 2 we outline the 2012 study ‘Public Perceptions of Threat in Britain’, which we designed in order to address the gaps in the literature identified in Chapter 1, along with our approach to the analysis of data. Our methodology is anchored in a post-positivist derivative understanding of what security means; that is to say, we start from the operating assumption that there are multiple ways of conceiving what security ‘is’, each of which derives from particular worldviews (Krause and Williams, 1997). Hence, in order to study everyday threats to security, it is necessary to ask how different citizens and groups portray various issues as threatening – what devices they use to do this, why, and with what consequences for others. In this way, we did not begin with preestablished criteria about what counts as a security threat or what might be considered threatening to a given understanding of security. Rather, our use of the concept was immanent to the way in which participants in the fieldwork stages of the project understood the twinned concepts of ‘security’ and ‘threat’. The programme of research adopted an unusual and innovative combination of quantitative and qualitative approaches. The optimal research design would

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Introduction

9

combine the richness and interactive dynamics of focus group research, the spontaneity and depth of individual semi-structured interviews, and a large sample survey to enable researchers to assess key relationships and subgroups statistically. Any one of these on its own is insufficient in view of the complexities of the research problem; our 2012 study incorporated each method in order to combine representative macro-level insights into public opinion across key variables, with non-representative micro-level thick descriptive accounts of individuals’ everyday stories, experiences, and (de)constructions. We set out how we conducted an initial tranche of ten mini-focus groups, or ‘triads’, of three people each in conjunction with the social survey company TNS-BMRB in April 2012, in which we explored questions such as how participants conceptualise ‘security’ and ‘security threat’, the way that they think about different security threats in their daily lives, and whether they agree or disagree with and/or are affected by a range of government messages about security. The rationale for these small groups was to combine the kind of group interaction in traditional focus groups – for example, getting participants to work together on sorting different threats into categories based on what they regard as salient criteria and sharing their individual stories of (in)security – with the depth of insight that can be more easily obtained in individual interviews. We used our observations of the mini-focus groups and analysis of the transcripts to reflexively inform the development of a twenty-five-minute online survey that was administered to 2,004 respondents from ICM’s internet panel in June 2012 (ICM, a leading polling organisation in the UK, has a panel of roughly 100,000 respondents from which it samples for online surveys). This included a booster sample of 251 British Muslims. We were interested in Muslims as a subgroup that faces threats such as racism, Islamophobia, and hate crime, and is also regarded as uniquely threatening by members of other groups. We describe the sample we obtained, which, other than the booster, was representative of the British population on demographic criteria such as region and age. We also outline the questions we asked, and why, with a particular emphasis on our measurement of perceptions of twenty-two different security threats, which focused on breadth rather than intensity, and asked about threats at the world, national, community, and personal levels. We then conducted a second wave of ten mini-focus groups in September 2012, in which we concentrated on more specific areas of concern to us in the light of the first two stages of research – such as how citizens think and articulate specific security threats more than how they group different threats – and explored some of the themes emerging from initial analysis of the survey data. The last part of the chapter discusses our approach to analysis. With the qualitative data, this is based on our observation of particularly salient and striking interchanges in groups, and also the organisation of the transcripts into recurring themes, i.e., how certain ideas, modes of communication, and cultural

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

10

Everyday security threats

devices for expressing understanding and marshalling evidence in support of individual views were repeated in different groups. Our method of presenting and analysing the views, anecdotes, and stories we co-produced proceeds largely through juxtaposition (Shapiro, 2013): by comparing and contrasting what different participants said (and did not say); and by setting these diverse opinions against the backdrop of dominant and otherwise homogenous (and elitist) national security frames. With the survey data, the analysis progresses from univariate statistics, through bivariate correlations and tables, to structural equation models in which we simultaneously estimate the influences on perceptions of threats and then the influence of threats on other variables such as spending preferences. Chapter 3 brings together insights from our focus group and survey data in order to examine the scope of threats and their origins. We look at threats in two ways. We begin by summarising how participants in group discussions defined and understood the key concepts of ‘security’ and ‘threat’, the primary referent object that they invoked in doing so, and the vernacular methods of perception, measurement and categories of understanding drawn upon in response to open-ended questioning about what security means to them in the context of their everyday lives. Crucially, we show that a recurring scale of understanding consisting of four primary levels – namely, personal, community, national, and global – structured all of our group discussions of security threats in one way or another. We therefore focus on the breadth, or number, of security threats that individuals identified in total at the global, national, community, and personal levels. We show that on average our sample identified the most threats at the global and national levels – roughly seven and four respectively – with a drop when we moved to the community and personal levels, with roughly two. We examine what those threats were: issues like terrorism, religious extremism, and the economy at the global and national levels, immigration and crime at the community and individual levels, while the economic crisis was identified as a threat at all levels. We then explore the influences on the breadth of threats individuals identify, drawing on previous research to focus on the key dispositions of authoritarianism and mortality salience, media use, level of education, and race and religion. Chapter 2 includes a path diagram that depicts the model for the lay reader. The second part of the chapter then focuses on specific threats that featured in discussions of security threats in the focus groups, the ones that were among the most salient threats identified by survey respondents and that vary in their characteristics, such as symbolic versus realistic, or physical versus economic threats. We concentrate on perceptions of threats from terrorism, immigration, the economy, and the environment and their origins, hypothesising which threats and at which levels we would expect to see variation. Among the findings of this chapter is that mortality salience and authoritarian attitudes are particularly

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Introduction

11

strong predictors of the number of threats that individuals identify and of the identification of specific threats such as terrorism and immigration. However, our analysis clarifies the nature of that influence. For example, mortality salience is a powerful influence on the perception of threats at the global and national levels, especially where the dangers are physical (terrorism) or pertain to identity and labour competition (immigration), but mortality salience has no influence on perceptions of community- or personal-level threats. Authoritarianism, on the other hand, is strongly related to heightened perceptions of threat from terrorism and immigration at all levels, but negatively related to seeing the economic crisis or environmental degradation as threats. Thus, this chapter debunks previous claims that there are few systematic influences on threats, and it goes further in clarifying the variation in the origins of different threats at different levels. It demonstrates that perceptions of threat are the result of the interaction of the characteristics of the threats themselves – e.g., physical versus symbolic threat – with characteristics of individuals. From here Chapter 4 pertains to the consequences of identifying both more or fewer security threats, and also specific security threats such as from immigration, for other political attitudes and behaviours. Previous research has examined several consequences, from whether or not an individual votes and for whom they vote, to attitudes towards minorities or preferences for specific counterterrorism policies. But these have generally been explored in isolation – single dependent variables or with respect to single threats – rather than by looking at threats writ large as we do here. Using the survey data, we look at the effects of the breadth of global, national, community, and individual threats identified, and then with respect to specific security threats individuals identify from terrorism, immigration, the economy, and the environment, on voting behaviour, attitudes towards immigrants and minorities, and policy preferences. The methodological approach is an extension of the structural equation models used in the previous chapter. Among the findings are that voting is unique in that only global and national considerations appear influential – thus the fact that it is often the focus of studies of, for example, economic threat is misleading. We also demonstrate and explain differences in the effects of threats such as terrorism and immigration from the economy and the environment. Chapter 5 takes as its starting point the various ways in which elite responses to security threats such as the NSS both summarise government perceptions of the most salient threats and are also intended to send messages to the public and shape their behaviour, perhaps directly – individuals encounter messages on public transport and have leaflets posted through their doors – but also indirectly via the media. For example, a search of Nexis UK shows that the terms ‘National Security Strategy’ or ‘security policy’ appeared more than fifty times in the Daily Telegraph between the formation of the Conservative–Liberal Democrat coalition government in May 2010 and June 2012, when our survey was in

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

12

Everyday security threats

the field, i.e., about twice a month. We probed awareness and perceptions of the NSS and other government messages about security in our focus groups and survey. This allowed us to examine three dimensions of the relationship between elite and non-elite perceptions and experiences of security threat politics: 1) the extent to which the British public are aware of the NSS or of other government messages and efforts to mitigate security concerns; 2) whether such awareness is associated with heightened or reduced levels of threat perception; and 3) what citizens think of such messages, whether they feel that their behaviour has been affected as a result, and how they express their knowledge and experience of being enlisted as ‘citizen-detectives’ in public spaces in particular (VaughanWilliams, 2008). The focus groups attempted to find out how participants viewed government messages and whether they thought that these altered their everyday behaviour or that of others in society. Discussions were animated by a range of stimuli such as pen portraits of various fictional characters faced with decisions about how to behave in specific circumstances. Other group stimulus material included a 2004 booklet distributed to all UK households by the government called ‘Preparing for an emergency’, a 2006 poster from the London Metropolitan Police’s ‘You are that someone’ anti-terrorism campaign, and a radio advert from the same campaign about being prepared on the city’s transport network. Participants were asked about their awareness of these various campaigns, whether or not they felt these initiatives were effective in changing their behaviour and that of the public more generally, and if they had any suggestions about how security-related communication of this nature could be changed in the future. By investigating narratives of, inter alia, threat, safety, and belonging, the analysis considers how it might be possible to study the politics of (in)security from the perspective of subjects produced by those apparatuses of security. The various stories people tell – of economic insecurities, fear of crime, and Islamophobia – problematise the narrow and homogenising imperatives of the NSSs, and open up alternative narratives about identity, border-production, and multiple overlapping (in)securities. We argue that everyday attitudes towards and stories of (in)security offer a powerful counter-archive to the dominant national frame. While many of the views, anecdotes, and stories reproduce dominant governmental logics, it is also possible to identify political discourses that challenge these logics, repoliticise the grounds on which national security agendas are authorised, and reveal actually existing alternatives to cultures of suspicion and unease. In the survey we asked about awareness of any government security strategies or programmes (and what they were) and then about the NSS in particular. Awareness of any government security programme and of the NSS is low, about 10 per cent of the sample for each, with surprisingly little overlap between the two: fewer than half the respondents who said that they had heard of a government security programme also said that they had heard of the NSS. Beyond

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Introduction

13

awareness per se we examine the relationship between awareness of government security strategies and perceptions of threat. There are obvious endogeneity problems here, so we are not claiming there is a causal relationship, but it is nevertheless striking that awareness of government strategies for security is associated with perceptions of more threats. We also look to see whether awareness is associated with perceptions of fewer threats in the future, but there too the association is with more threats. We then repeat the analysis but focus on British Muslims in particular, showing that these relationships are often stronger, which reinforces the evidence from the focus groups that contemporary security policy appears to intensify Muslims’ fear that they are objects of suspicion. We end the chapter by discussing the implications of our qualitative and quantitative data for government messages about security. Finally, the Conclusion sums up the research, explores its implications, and draws lessons for the future for both academic and policy-making communities. The implications of the research are several, spanning: government and its understandings of how the public views security threats and how the public perceives, experiences, and responds to messages about security threats; academic research in IR and Security Studies and how it conceives of public opinion and the role of the citizen in the risk management cycle; and academic research in Political Psychology and its understandings of the origins and consequences of threat perceptions. Notes 1  While the 2010 NSS used the language of ‘risk’ more so than the 1998 NSS, the former continued to mobilise the concepts of ‘risk’ and ‘threat’ interchangeably; for a useful genealogical analysis of the deployment of these terms in the context of UK national security policy see Hammerstad and Boas (2015: 480-2). 2  Here and throughout the book we refer to the field of Critical Security Studies in the expansive sense as paradigmatically outlined by Krause and Williams (1997) – see also Peoples and Vaughan-Williams (2014). 3  In R. Mason,‘Britain will be at heightened risk of terrorism for years, says police chief ’, Guardian, 24 November 2014. 4  For a critical commentary on the chronopolitics of 9/11 and the discourse of beginnings and endings see Closs Stephens and Vaughan-Williams (2008).

1

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Perspectives on security threat politics

Introduction As governments – particularly, though not exclusively in the global North – responded to what they commonly framed as a ‘new threat’ from terrorism after 9/11 (Croft and Moore, 2010; Thrall and Cramer, 2009), they felt compelled in turn to outline the security strategies that this shift and other perceived threats in the post-Cold War world necessitated. Thus, whatever continuities we may seek to delineate before and after 9/11 – including the political move to claim the novelty of a given era for particular policy ends – the international political landscape in which Britain and other liberal democratic states operate is presented by policy elites as having been transformed dramatically. No longer are interests at ‘home’ and ‘abroad’ portrayed as being under threat from particular states, but rather from a complex web of security issues that undermine the domestic/international distinction and are commonly said to include: international terrorism; biological and chemical weapons of mass destruction; conflict and state ‘failure’; migration and immigration; pandemics; and transnational crime (Cabinet Office 2008, 2010). A genealogical account of this perceived shift in the nature and location of security threats – and socioeconomic and ideational factors that may have given rise to this transformation – is beyond the scope of this study (for more on this theme, see Hammerstad and Boas, 2015). What matters more in view of our specific aims and objectives is the way in which governments in Britain and America in particular have pledged not only to develop a resilient security architecture designed to identify and mitigate against the effects of the emergence of these perceived threats but, as key policy objectives, to reassure their citizens, to heighten collective levels of security among populations, and to reduce subjective feelings of being ‘threatened’ on the one hand, while including citizens as agents of national security on the other: this constitutes the main problematique with which the book as a whole engages.



Perspectives on security threat politics

15

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Indeed, against this policy backdrop, diverse publics have been given unprecedented prominence in the formulation and exercise of national security policy. In the US, for example, Secretary of Homeland Security Tom Ridge said: Citizens are a necessary and absolutely irreplaceable asset in this fight. Since that day [9/11], we have come a long way to motivating our citizens to do their part to prepare and ready their families and friends for any potential disaster, whether natural or man made . . . They have all helped us to engage and empower citizens to embrace a direct role to accept the responsibility to secure your family, your freedom, and your community. (quoted in Jarvis and Lister, 2010: 178)

In Ridge’s articulation, citizens are active participants with the state in threat preparedness and are also ‘empowered’ as resilient subjects by performing this role. In accepting, indeed ‘embracing’, the challenge, Ridge suggests that citizens can enhance their own individual security, that of their communities’, and by implication the nation’s. Similarly, successive governments in the UK have published US-style NSS documents that outline the principal threats facing the nation, divided into three tiers from most to least pressing, and discuss the roles of government and citizens in mitigating them. Jarvis and Lister (2010: 174) describe this approach as ‘conscripting “ordinary” citizens into the state’s security apparatuses’ and refer to citizens as ‘stakeholders’, while Vaughan-Williams (2008) employs the term ‘citizen-detectives’ to describe their role in maintaining vigilance in public spaces (see also Malcolm, 2013). The term ‘conscription’ suggests that the participation of citizens in national security architectures is not entirely voluntary. There are, in addition, obvious ambiguities in terms of both the effects of these new demands on citizens and the extent to which publics are reassured or made to feel more anxious as a result, as Jarvis and Lister (2010, 2015) among others acknowledge (see also Marshall et al., 2007; Massumi, 2005; McDermott and Zimbardo, 2006). Moreover, liberal democracies also rely on citizens to limit state responses to threat and to hold governments accountable for the illiberal choices they may make in the name of protecting society as a whole from threats and so there is a fundamental political ambivalence surrounding the relationship between vigilance and threat perception (Chalk, 1998). This contemporary focus on security threats and citizens’ roles in threat preparedness and response begs questions of how diverse multiethnic publics conceptualise, experience, and narrate their understandings of security and threat. While spending on national security in the UK since 2001 has more than tripled to £3.5 billion (Cabinet Office, 2008), it remains unclear how threats are conceived by and affect the British public, whether they are aware of and/or understand government security strategies and objectives, and whether citizens feel more or less ‘secure’ as a result. Despite changes in discourses surrounding the role of citizens in the formulation and implementation of national security

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

16

Everyday security threats

policy and increases in government security budgets, little is actually known about public attitudes towards and experiences of security threats, what sorts of issues citizens find threatening, whether everyday security concerns comport with those of government, and the connection between security concerns writ large and other political attitudes and behaviours. In short, we do not know in any empirical depth what the ‘broadening’ and ‘deepening’ of the concept of security to which IR and Security Studies scholars often refer might mean at the level of the everyday. As we suggested in the Introduction to this book, our knowledge of public perceptions and experiences of threats tends to be confined either to discrete policy areas such as terrorism and anti-terrorism (Jarvis and Lister, 2012), or as they relate to specific areas of personality, predispositions, or attitudes such as authoritarianism (Hetherington and Weiler, 2009; Stenner, 2005) and tolerance (Gibson and Gouws, 2003; Marcus et al., 1995). We know relatively little about the range of issues that citizens regard as security threats, their causes, or the levels at which such issues are perceived as threats – for example, as global or national threats. One aspect of this lacuna is a broader lack of social scientific research, including a tendency within IR and Security Studies – and across the so-called ‘traditional’ and ‘critical’ divide (Peoples and Vaughan-Williams, 2014) – to focus on elite perceptions and constructions of security threat rather than public opinion, non-elite knowledge and experience, and the issue of audience reception of acts of securitization (Balzacq, 2010; McDonald, 2008). In this context, a national frame for understanding security threats is still predominantly shared by national governments and, though increasingly to a lesser extent, academia, but the extent to which members of the public share this framing is largely unknown. Equally, extant research has yet to offer any real depth of insight into convergence and/ or divergence between ‘official’/‘elite’ and ‘popular’/‘non-elite’ knowledge, constructions, and understandings of the concept of security, public encounters with and negotiations of security threats in everyday life, and the sorts of factors affecting citizens’ perception of threat – let alone the possible consequences of such divergence between official and popular understandings as they relate to government projects of enhancing societal resilience. Another aspect is a lack of understanding of the political psychology of different threat perceptions as opposed to singular threats, such as from international terrorism, immigration, or environmental degradation, and of the consequences of different threat perceptions for other political attitudes and behaviours. Research has tended to be either on discrete threats when, as work in IR and Security Studies tells us, individuals deal with multiple threats simultaneously. Beyond authoritarianism, we know little about how individuals construct and make sense of the range of potential threats they face on a day-to-day basis, and even among authoritarians it is unclear whether their disposition to

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

17

panic encompasses both sociotropic and personal risks or whether sociotropic concerns, about the fate of society and the groups with which authoritarians identify, may dominate. Against this backdrop, this opening chapter has three purposes. First, we discuss previous research on these issues in the otherwise discrete literatures produced in the fields of IR and Security Studies on the one hand and Political Psychology on the other. These works are rarely – if ever – brought into conversation with each other and yet, we argue, the complexity and political significance of the concept of threat demands a triangulation of approaches in order to explore it from a range of perspectives. Second, in so doing, we situate our own research by identifying what we consider to be some of the limitations of previous scholarship both within each subdisciplinary context and, crucially, as a result of their mutual neglect. Our approach is to synthesise and push further existing insights without seeking to prioritise one perspective over another. Rather, we are interested in exploring the implications of how different perspectives on security threat politics might be mobilised. Third, we describe the ways in which our research seeks to overcome past limitations and set the scene for our post-positivist mixed-methods research design. We end by presenting our own theory and discussion of the origins and consequences of threats from the foregoing analysis of the current state of the art. From there, Chapter 2 goes on to discuss our research design in greater depth. The concept of threat in International Relations and Security Studies The argument that we develop throughout the course of the book as a whole is that the study of security threats has become vulnerable to a series of ‘turns’ in the literature – to the vernacular, to the everyday, and to an analysis of public opinion – because of the hitherto persistence of elitist perspectives and their limitations. We begin here by offering a brief historiography of the treatment of the concept of threat – first of all in the literature produced by IR and Security Studies and then by Political Psychology – in order to contextualise these turns and ultimately our contribution to them. In the context of the Cold War, for IR and Security Studies scholars the concept of ‘threat’ was once operationalised largely as an objective condition concerning the physical safety and survival of the state in the context of conflict within the anarchical state system (Dunn Cavelty, 2008; Gillespie and O’Loughlin, 2009a, 2009b; Hammerstad and Boas 2015; Meyer 2009; Morgan, 2000). Thus, perceptions of ‘threat’ were dominated by the prospect of nuclear war, and the concept pertained almost exclusively to the military sector (Bjereld, 2001). Indeed, Miriam Dunn Cavelty (2008: 8) argues that for most of the twentieth century, ‘threat images’ in IR were largely taken ‘as a given’ and understood to be ‘measurable’, and therefore states’ security policies were interpreted as

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

18

Everyday security threats

responding to ‘an objective’ understanding of threats. For these reasons, Dunn Cavelty points out that the connection drawn between the distribution of military capability within the state system and the concept of threat meant that there was a high degree of certainty, which lent itself to approaches that claimed replicability and predictability. However, according to what is by now a familiar historiographical narrative, the dissolution of the USSR, the rising significance of non-state actors, and the broadening of the security agenda beyond state-centred militarism set the conditions for new understandings of security threats (Buzan et al., 1998; Buzan and Hansen, 2009). Since the end of the Cold War, policy and academic discourses concerning security threats have ‘broadened’ to encompass sectors other than the military – including environmental (i.e., threats to earth as biosphere), societal (i.e., threats to notions of community), economic (i.e., threats to citizens’ welfare), and political (i.e., threats to particular national identities) (Buzan, 1991; Buzan et al., 1998; Peoples and Vaughan-Williams, 2014). Moreover, many discourses associated with new security challenges have also ‘deepened’ to include not only the state as the referent object of security (i.e., that which is threatened), but also the individual (‘human security’) (Booth, 1991), the planet (‘ecosecurity’) (Dalby, 2002), financial architectures and critical infrastructure (‘network security’) (Coward, 2012; de Goede, 2007; Lakoff and Collier, 2010), and populations and particular ways of life (‘biopolitical security’) (Dillon and Lobo-Guerrero, 2008; Dillon and Reid, 2009; Vaughan-Williams, 2015). Thus, in the IR and Security Studies literature, the study of the concept of threat has evolved from analyses of the security dilemma between states under anarchic conditions (Waltz, 1979) towards a more commonplace view that threats are not automatically given as a result of those systemic features (Wendt, 1999). The impact of the social constructivist turn, alongside the broadening and deepening of the security agenda as above, has meant that the perception and understanding of threats are now widely seen as being produced through interaction between states and non-state actors alike. As such, from this perspective security threats are said not simply to exist independently of our knowledge and representations of them. Rather, in the language of the Copenhagen School, a security threat is socially constructed according to a process of securitization whereby a particular issue comes to be framed as posing an existential threat to a given referent object – via a speech act, for example, in the context of political speeches and media representations (Buzan et al., 1998). This account, moreover, emphasises a degree of urgency associated with the construction of threats which parallels Carl Schmitt’s (2005) paradigmatic definition of sovereignty as the ability to decide firstly that emergency conditions exist and secondly that extraordinary measures – such as martial law – are required in response. Despite the prominence and popularity of the Copenhagen School approach, however, various interlocutors have pointed out a number of its shortcomings

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

19

vis-à-vis the study of threat perception. With its prioritisation of the speech acts of elite securitizing actors, in particular the understanding of threat in existential terms, and the emphasis on emergency conditions, securitization theory has been critiqued for downplaying the role of audience reception (Balzacq, 2010; McDonald, 2008; Meyer, 2009). According to Thierry Balzacq (2010: 19), securitization theory has tended to ‘skirt the distinctive role of the audience’ because of a prior focus on the utterance of a speech act (the illocutionary act) rather than its effects (the perlocutionary act). Likewise, Christoph Meyer (2009: 650) argues that securitization theory does not adequately address ‘different levels of risk perception and fear among different types of audiences’. These critiques feed into a wider set of concerns about the narrowness of the securitization frame and its inability to locate the social construction of threat and (in)security within broader cultural contexts. For Johan Eriksson (2001: 9), the securitization frame is limited in its focus on the ‘societal salience’ of issues seen as important in a given time and place: it does not account for the way in which some threat images acquire wider ‘societal salience’ whereas others do not or the way in which such ‘salience’ is unlikely to mean the same thing for everyone in society. As well as ignoring different meanings and struggles over the ‘becoming salient’ of a particular issue, Eriksson (2001: 9) argues that ‘security’ and ‘threats’ and their interrelationship can ‘be framed in other ways than the specific frame implied by securitization’. Thus, for example, while the Copenhagen School portrays ‘security’ negatively – as providing legitimacy for military mobilisation and emergency measures – Eriksson (2001: 13) argues that connotations of ‘threat’ and ‘security’ may well differ among diverse publics. From his perspective, the societal salience of particular issues is shaped by the framing of threat politics defined by ‘cognition and culture, in combination with some contingency of epistemic communities, bureaucratic politics, identity politics, and/or some set of referent objects of images at hand’ – in other words, a broader set of factors, characteristics, and conditions than the otherwise narrow ‘securitization’ approach allows for (Eriksson, 2001: 222). It is possible to trace elements of Eriksson’s (2001: 215) approach to security threat politics – one that pays particular attention to the ‘cultural context in which a threat image is identified’ – back to earlier currents within critical constructivist and poststructuralist scholarship. Jutta Weldes et al. (1999: 10), for example, draw attention to the cultural reproduction of insecurity, which they view as ‘the product of processes of identity construction in which the self and other, or multiple others, are constituted’. On their view, cultures of insecurity encompass individual subjectivities, communities, and states, and in order to understand the social construction and reproduction of threat it is necessary to pay attention to ‘the context within which people give meaning to their actions and experiences and make sense of their lives’ (Weldes et al., 1999: 1). In turn,

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

20

Everyday security threats

this means focusing on the ‘practices and categories through which people engage with each other’ – the narratives, memories, and imaginaries that they draw upon – in making sense of security threats and concomitant cultures of insecurity (Weldes et al., 1999: 13). Earlier still, in his seminal analysis of the role that difference and otherness play in constituting the identity of the US, David Campbell (1992: 2) argued that threats are never given and that danger is a socially constructed and inherently political category of understanding: ‘Danger is not an objective condition. It is not a thing that exists independently of those to whom it may become a threat.’ More recently, perspectives associated with International Political Sociology (IPS) attempt to offer a thicker understanding of dynamics of (in)securitization in ways that identify and seek to work with – rather than bracket off and sideline – cultural contexts otherwise occluded by securitization theory (Bigo, 2008). The IPS research agenda has given rise to studies of not only the (in)securitizing moves of national governments but also of the knowledge and practices of nonstate actors, including security professionals (Bigo, 2000) and private security companies (Leander, 2005, 2013). Diverse attempts to mobilise discourses of security threat in order to produce governable subjects have been investigated in the context of, inter alia, risk management and dataveillance (Amoore and de Goede, 2005); the cultivation of unease and suspicion in counter-terrorism measures since 9/11 (Bigo, et al., 2007; Bigo and Tsoukala, 2008); and the use of exercises and resilience training in civil contingency planning (Anderson and Adey, 2011; Aradau and van Munster, 2012). But while research associated with IPS has problematised the range of actors involved in (in)securitizing moves, the ability to draw neat lines around the cultural contexts that delimits them, and the attempt to produce certain types of political subjects for the purpose of governmental logics, these dynamics and the question of the social and cultural (re)production of understandings of threat and (in)security have been least explored from the embodied perspectives of those subjects (Bröckling et al., 2010 13; Gillespie and O’Loughlin, 2009a: 681; Jarvis and Lister, 2013: 1). In this regard, some IPS approaches as well as securitization theory suffer from a prevalent elitist bias, which ultimately pays scant attention to the political subject of threat and (in)security that is both produced by and helps to shape attempted acts and processes of (in)securitization. The insistence in a range of feminist and gender perspectives on the importance of the embodied experiences of political subjects offers a tradition of thought that recovers the individual as the referent object of threat and (in) security and hence offers critical resources for countering elitism (Enloe, 1990). However, as Annick Wibben (2011: 7) has argued, the maxim ‘the personal is political’ – made famous by Cynthia Enloe’s classic ‘bottom-up’ study of globalisation – has in large part not been ‘perceived as contributing to the debate on security’ beyond gendered violence. More recently, the ‘vernacular’ and

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

21

‘everyday’ turns produced by critical literatures in Security Studies have sought to work with and negotiate this maxim and it is alongside this latest research – together with a growing body of thought associated with ‘ontological security’ – that we position the present study as a whole. A small but growing literature has sought to explore what it might mean to begin the study of threat and (in)security from the perspective of popular – or ‘vernacular’ – constructions. Inspired by an eclectic mix of ethnographic (Bubandt, 2005; Gillespie, 2007; Gillespie and O’Loughlin, 2009a; Gillespie et al., 2010), emancipatory (Jarvis and Lister, 2013), cosmopolitan (Gillespie and O’Loughlin, 2009a), and constructivist perspectives (Moss and O’Loughlin, 2008), this work has typically focused on how particular individuals and groups articulate their attitudes and understandings especially – though not exclusively – in the British context. For example, the 2004–6 ‘Shifting Securities’ project drew upon audience ethnography in order to explore how the media shapes multiethnic publics’ perceptions of threat and (in)security (Gillespie, 2007; Gillespie et al., 2010; Gillespie and O’Loughlin, 2009a, 2009b; Moss and O’Loughlin, 2008). Similarly, in 2010 Lee Jarvis and Michael Lister organised fourteen focus groups varied according to ethnicity and geographical location in order to investigate the relationship between anti-terrorism policy, experiences of citizenship, and human security (Jarvis and Lister, 2012, 2013, 2015). Both studies found that public perceptions and constructions of threat differed according to identity, ethnicity, religion, class, gender, location, and generation, and they make normative arguments for ‘bottom-up’ qualitative research agendas that recover ‘marginalized voices’ (Jarvis and Lister, 2013: 162) and politically progressive forms of ‘everyday cosmopolitics’ (Gillespie and O’Loughlin, 2009b: 110). Running in parallel with the ‘vernacular turn’ is that of the ‘everyday’. The everyday turn in the humanities and social sciences has its roots in French cultural thought of the 1980s (Sheringham, 2006). Influenced by the diverse works of thinkers such as Henri Lefebvre, Michel de Certeau, and Roland Barthes, this heterogeneous body of work finds common ground in giving greater importance to the quotidian: to the ‘spaces, rhythms, objects, and practices’ around us (Sheringham, 2006: 2). In CSS a number of scholars have embraced these insights in order to develop IPS beyond an elitist focus (Guillaume, 2011; Guillaume and Huysmans, 2013; Huysmans, 2014; Noxolo and Huysmans, 2009; see also Bajc and de Lint, 2011). Conceptually, as Huysmans (2009: 197) has argued, much of this work has shown that security is not only about exceptional politics, such that a distinction can be drawn between ‘high’ and ‘low’ politics: the everyday is fundamentally ambiguous and has a ‘double analytical status’ in which it is possible to trace both (in)securitizing moves and arenas in which these moves are negotiated and resisted by individuals and communities. Empirically, work on the everyday has sparked interest in more sociologically oriented studies of the

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

22

Everyday security threats

international via treatments of phenomena not normally considered to be ‘proper’ to IR and Security Studies such as the securitization of neighbourliness in multiethnic communities (Buonfino, 2013). Lastly, the question of threat and the subject of (in)security also connects in potentially productive ways with recent work on ‘ontological security’ (Croft and Vaughan-Williams, 2016).1 According to R. D. Laing’s (1990: 39) paradigmatic definition, the notion of the ontologically secure person ‘will encounter all the hazards of life, social, ethical, spiritual, biological from a centrally firm sense of his own and other people’s reality and identity’. This is not to say that the ontologically secure self is divorced from wider intersubjective interaction – a point later emphasised by Anthony Giddens (1991) – but that for Laing the referent object is the individual. More recently, Stuart Croft (2012) has argued that there are four key elements of ontological security for the individual subject in the context of wider social relations: biographical continuity (a storyline that the individual can communicate to others); a cocoon of trust relations (a network that helps to filter out a sense of threat in everyday life); self-integrity (an ability to make predictable decisions based on their biographical self-understanding); and, importantly, dread (an anxiety about the sense of self). It is this final element – the fear of precariousness and dread management via the development of biographical narratives of the self – that is arguably at the heart of ontological security studies (Croft and Vaughan-Williams, 2016). While some ontological security studies in IR have extrapolated from Laing’s focus on the individual subject and transposed his theory to the level of the state in order to make it more acceptable in disciplinary terms (Mitzen, 2006; Zarakol, 2010), others have not sought to fix the referent object in quite the same way or to the same extent, but precisely take the referent object as the site of investigation in the context of international politics (Huysmans, 1998; Kinnvall, 2006; Steele, 2008). Following this latter track in particular, there is considerable hitherto untapped potential for bringing ontological security studies into closer conversation with both the ‘vernacular’ and ‘everyday’ turns in order to better understand the range of articulations of dread, how such anxieties are managed by diverse publics via biographical narratives of self and other, and the different ways in which routine attempts at negotiating and managing fear and dread may lead to dynamics of insecuritization in society as a whole (Croft and Vaughan-Williams, 2016). As well as ‘vernacularising’ ontological security studies, the latter body of work also offers a theoretical framework for investigating the relationship between dread and particular identity claims, the use of contrastive others in creating distinctive biographies of the self and networks of trust, and other devices for managing anxiety to be found at the level of the everyday. By bringing together the ‘vernacular’ and ‘everyday’ turns in IR and Security Studies – in conversation with work on ontological security and threat perception in the Political Behaviour and Political Psychology literatures – we hope



Perspectives on security threat politics

23

then to mobilise a range of commensurate perspectives in order to advance this nascent research agenda and take the study of security threats beyond its extant elitist bias.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Threats and their origins in Political Behaviour and Political Psychology Research on mass political behaviour was also heavily influenced by the Cold War and its precursors. Studies initially sought to understand mass behaviour and the temptations of dictatorships and illiberal regimes (e.g., Fromm, 1941), such as through Adorno et al.’s (1950) F-scale, Levinson’s (1949) account of ethnocentrism, and Stouffer’s (1954) study of political tolerance. At the core of these theories was that the more individuals are threatened by outgroups – threats arising from a combination of negative affect and assessments of outgroups’ power – the less tolerant they become. While theory and evidence in each of these areas has since evolved considerably (e.g., Altemeyer, 1981, 1988, 1996; Gibson and Gouws, 2003; Kam and Kinder, 2007; Marcus et al., 1995; Stenner, 2005) the role of threat remains paramount in them. Thus, Marcus et al. (1995) and Gibson and Gouws (2003) argue that a normative threat from disliked groups renders individuals less tolerant. Threatened individuals are also more responsive to information about threats (Marcus et al., 1995). Altemeyer (1996), Stenner (2005), and Hetherington and Weiler (2009), while disagreeing about the relationship between threat and authoritarianism, also all have perceptions of threat at the core of their theories, e.g., authoritarians ‘stand about ten steps closer to the panic button than the rest of the population’ (Altemeyer, 1996: 100). Other research examines threat in particular contexts. For example, political scientists and psychologists continue to look at areas such as identity threat (Fischer et al., 2010), policy threat (Miller and Krosnick, 2004), threat from immigration (Lahav and Courtemanche, 2012), environmental degradation (Baldassare and Katz, 1992), cyber-crime (Speer, 2000), religious threat (Campbell, 2006), and racism, sexism, and stereotype threat (Huguet and Regner, 2007; Steele and Aronson, 1995). In addition, much of the context and focus since 9/11 has been on threats associated with international terrorism, both as a consequence of 9/11, 7/7, and other attacks or foiled attacks, and because of the ongoing Palestine–Israel conflict. There is, however, uncertainty about the causes of individual perceptions of threats. Some research suggests that they are highly unsystematic: Gibson and Gouws (2003: 198) refer to ‘the failure of earlier research to account for any variability in threat perceptions’. Marcus et al. (1995: 37) find threats from groups to be ‘exogenous to measures of social background, personality, ideology, and support for the general norms of democracy’. Similarly, Feldman (2003), Feldman and Stenner (1997), Hetherington and Suhay (2011), and Stenner

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

24

Everyday security threats

(2005) uncover weak correlations between their measures of threat and authoritarianism.2 Yet, as the quote from Gibson and Gouws suggests, other research has identified systematic causes of threats. Lavine et al. (1999: 338) ‘view a chronic fear of a dangerous and threatening world to be a key component of the authoritarian personality’. In Gibson and Gouws’ own research in South Africa, while there is variation by race, closed-mindedness and harbouring democratic values are predictive of perceived threat from groups. Sniderman et al. (2004) find that perceptions of threats from immigration are linked to cultural identity, while Huddy et al. (2002, 2005) show that demographic factors such as education, gender and race, and the predisposition of authoritarianism,3 are linked to perceptions of threat (see also Goodwin et al., 2005; Green, 2009). The related concept to threat of ‘risk’ (Sjoberg, 2000) has been shown to be influenced by social trust (Viklund, 2003).4 The heightened sense of threat induced by ‘mortality salience’ also increases ‘the positivity of evaluations of people and ideas that support one’s cultural worldview and the negativity of evaluations of people and ideas that threaten it’ (Schmiel et al., 1999: 906). This may be more likely in contexts where there is prolonged conflict (Hirsch-Hoefler et al., 2014). Finally, Ridout et al. (2008) indicate that media exposure heightens perceptions of ‘modern’ and ‘postmodern’ threats.5 In sum, while there is disagreement, variables including dispositions such as authoritarianism, demographic characteristics such as education and age, exposure to information via media, and contexts in which there is prolonged conflict appear to be influences on threat perceptions. At the same time, previous research has also explored disparate consequences of threats, including political attitudes and behaviours (Davies et al., 2008; Davis and Silver, 2004; Hirsch-Hoefler et al., 2014; Huddy et al., 2002; Robbins et al., 2013), effects on other attitudes such as parenting (Fischer et al., 2010), the efficacy of government responses (Kerwin, 2005), and in the justification of illiberal policies and actions (Bigo and Tsoukala, 2008; Neal, 2009). In keeping with past research, post-9/11 security threats related to international terrorism have been shown to render individuals more illiberal and less tolerant of difference: making individuals more willing to trade off civil liberties for security measures (Sanquist et al., 2008); elevating antipathy towards the entry of immigrants (Green, 2009); elevating antipathy towards Muslims as a cultural and religious minority (Croft, 2012; Kalkan et al., 2009); altering the social identities of majority and minority populations and making them more salient (Aly and Green, 2010); and leading to a tendency to prefer leaders with particular traits such as strength and charisma (Merolla and Zechmeister, 2009). In addition, a heightened sense of security threat is associated with mortality salience and a greater tendency to stereotype outgroups, to dislike stereotype-inconsistent members of outgroups (Burke et al., 2013; Greenberg et al., 1990; Landau et al.,

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

25

2004; Pyszczynski et al., 2002), to view a peaceful accommodation with an outgroup negatively (Hirsch-Hoefler et al., 2014), and with aggression (McGregor et al., 1998). Studies from Fromm’s (1941) analysis of the rise of Hitler, through McCann’s (1997) of voting in American elections from 1824 to 1964, to more recent analyses of the effects of terrorism (Landau et al., 2004; Merolla et al., 2007; Merolla and Zechmeister, 2009) argue that there is a connection between threat and a desire for strong and charismatic leadership. That being the case, we might expect elections to be of heightened interest to threatened individuals. It is therefore unsurprising that studies have found a relationship between threat and increased turnout in elections (Berrebi and Klor, 2008; Robbins et al., 2013). Robbins et al., informed by affective intelligence theory (Marcus et al., 2000), argue that this is the result of heightened anxiety among voters. Thus, researchers claim to know more about the consequences of perceptions of threats than of their psychological origins. Yet there is a narrowness of approach in this work in two respects: first, too often lacking is any effort to link research on distinct threats in order to examine the putative generalisability of findings. For example, are threats, regardless of the issue, all experienced similarly such that their effects on political attitudes do not vary? Second, the focus is generally on specific threats and how they are experienced at the national and personal levels, precluding understanding of whether people who perceive a greater threat from terrorism are also more likely to perceive other issues as threats and of any accumulation of knowledge pertaining to perceptions of the breadth of security threats. An agenda for new research in security threat politics While there is already considerable work on security threat politics across the subfields of IR and Security Studies and Political Psychology and Political Behaviour, in this section we identify several opportunities for novel areas of research arising from limitations both within and between the bodies of literature surveyed above. We suggest that a new research agenda in the study of security threat politics might be oriented around four key challenges that we seek to address in the course of the book as a whole: 1) the prevalent elitist bias in analysing security threats; 2) the mutual neglect of ‘vernacular’ and ‘everyday’ perspectives that might otherwise provide an antidote to such elitism; 3) the current emphasis on individual ‘threats of the moment’ rather than an appreciation of different levels at which threats are perceived; and 4) the need to move beyond a dichotomous understanding of sociotropic versus individual threat levels and a more nuanced appreciation of movement between multiple levels. We now elaborate each of these points in turn; taken together, they constitute the basis for a new framework of analysis.

26

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Elitist bias

As a growing number of analysts have sought to emphasise from a range of vantage points in IR and Security Studies, the prevalent elitist focus on politicians, security professionals, and private security companies – even in the ‘critical’ study of the politics of threat and (in)security – is analytically and politically problematic in excluding the political subject of (in)security for a number of significant reasons. First, if meanings of (in)security are intersubjectively produced by, culturally embedded in, and politically contested via ‘processes of identity construction in which the self and other, or multiple others, are constituted’, then the role of diverse publics in shaping that field of understanding is of central concern (Weldes et al, 1999: 10). Second, to prioritise elites silences the voices of individuals and groups marginalised by their socioeconomic, gender, racial, and ethnic status – political subjectivities who may also be disproportionately affected by discourses and practices of security – which in turn serves to perpetuate their exclusion (Booth, 2007; Hansen, 2000; Jarvis and Lister, 2013, 2015). Third, as Jef Huysmans (2014: 59) has argued, elitism also has the effect of reducing ‘the visibility of the pervasive presence of exceptionalist securitizing in everyday life and intimate relations’ such that the latter are bracketed off from security politics, which is assumed to be the preserve of an ‘exceptional class’. Therefore, as the ‘vernacular’ and ‘everyday’ turns have sought to emphasise, there are both analytical and political reasons why a recovery of nonelite knowledge, understanding, and experience of security threats is desirable and, as Jarvis and Lister (2015) among others have pointed out, there is considerable work to be done in order to aid a recovery of this nature. However, for reasons we go on to outline next, more could be done to integrate the insights of these turns, which hitherto have developed largely in parallel rather than in conversation with each other: one of our ambitions in this study is to provide such a synthesis.

The mutual neglect of the ‘vernacular’ and ‘everyday’ turns

Although both currents seek to recover the political subject of threat and (in) security in potentially productive ways, it is possible to delineate several tensions between the ‘vernacular’ and ‘everyday’ turns. The first concerns the use of operating distinctions between high/low and elite/everyday politics: scholars associated with the former largely work within these dichotomies, whereas those working in the context of the latter tend to refuse them. The second follows on from this because, while the former align emancipatory and cosmopolitan potential with a ‘bottom-up’ perspective, the latter are more circumspect about the everyday as a site for progressive politics while at the same time emphasising that it is not a somehow passive or inert realm. The third relates to

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

27

methodological differences: those seeking to ‘access’ the views and experiences of particular groups of citizens draw extensively on ethnographic and focus group work and are sometimes critical of what they consider to be the absence of such empirical engagement in IPS-related work (Gillespie and O’Loughlin, 2009a: 681). Finally, whereas the vernacular turn has taken the linguistic constructions of citizens’ accounts of threat and (in)security in their daily lives as its primary object of analysis, those influenced by the everyday typically privilege security practices negotiated in the context of citizenship more generally (Guillaume and Huysmans, 2013; Noxolo and Huysmans, 2009). What we hope to achieve in this book is to bring together aspects of both the ‘vernacular’ and ‘everyday’ turns in order to produce a novel theoretical perspective. While we agree with approaches that stress the fundamental ambivalence of the everyday, we maintain the distinction between elite and non-elite knowledge in order to identify the specificity of the latter in containing the capacity to disrupt the former. This stress on the disruptive capacity of non-elite knowledge is elaborated upon in Chapter 5. Specific ‘threats of the moment’

Moving away from trends within the IR and Security Studies literature, the common emphasis of research in Political Psychology and Political Behaviour on perceptions of threats is on specific threats, such as terrorism or immigration, and their intensity. This raises two problems related to the focus on an ‘extreme stimulus’. First, just as Lupia and Menning (2009: 104) argue that ‘researchers tend to ask about people and events that they suspect have caused emotional reactions’ (italics in original), rendering the generalisability of their results questionable, the same is true of perceptions of threats, hence the predominant focus in the extant literature on terrorist threats since 2001. We almost certainly cannot extrapolate findings about intense threats of the moment to everyday threats in general (see Sjoberg, 1999 and Wildavsky and Dake, 1990 for similar arguments about risk perceptions). Second, there is a parallel to the debate in the political tolerance literature about measurement error and the conceptual limitations of focusing on a single ‘least liked group’. The argument in that literature is that it misses the extent to which individuals vary in the breadth of their intolerance. Some individuals may harbour an intense dislike for one group, others for multiple groups; concentrating on a disliked group lumps them together. The parallel with threats is that the tendency to examine the intensity of single threats of the moment misses potentially important and consequential differences between individuals for whom the threat of the moment is one of many and individuals for whom it is the only threat.6 While some have recognised these limitations (e.g., Malholtra and Popp, 2012: 44) and called for the generalisability of their findings to be put to the test with other types of threat, their calls have not been

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

28

Everyday security threats

followed through. In the tolerance literature, those who have argued for a focus on breadth have claimed that concentrating on an extreme stimulus limits understanding and can lead to erroneous inferences such as of the effects of education (Gibson, 1986; Gibson and Gouws, 2003; Sniderman et al., 1989). As we will demonstrate, perceptions of threats from the global to the personal range from terrorism, immigration, and racial or religious hate crime to the economy, environmental degradation, online fraud, and burglary. Some of these perceived threats are realistic, others symbolic, while some may contain elements of both. In addition, terrorism and immigration are threats from outgroups with implications that are physical, economic, and identity-related, whereas racial or religious hate crime is, for most, a threat to another group, while environmental degradation is a threat that goes beyond groups but also pertains to physical damage, though not as directly as terrorism, and livelihood. Yet some research suggests that the threat of climate change has similar consequences to those we would normally associate with perceptions of threats such as terrorism – associations with authoritarianism, intolerance of outgroups, and defensiveness, for example – suggesting that such variation in the characteristics of threats is immaterial and that the consequences of perceptions of threats are not necessarily confined to the sources of the threat (Fritsche et al., 2012). As well as the possible effects of such distinctions, our interest is in the extent to which issues such as terrorism, immigration, and environmental degradation are seen as threats, by whom, and at what level – global, national, community, or family/ individual. Sociotropic versus personal threat

A second issue with extant research in Political Psychology and Political Behaviour is what we go on to outline and develop as the ‘level’ at which threats are perceived. Whether referred to as normative threat, economic threat, or racial threat, research has largely conceived of threats as sociotropic or national-level concerns, versus personal, individual-level concerns (e.g., Huddy et al., 2002; Joslyn and Haider-Markel, 2007; Maoz and McCauley, 2009; Schildkraut, 2009).7 Most of the research, albeit of single threats that are often threats of the moment, such as terrorism after 9/11, rather than of threats in general, finds that sociotropic threat is a more powerful influence on attitudes and behaviours than personal threat.8 Indeed, these are the findings regarding terrorism (e.g., Huddy et al., 2005; Joslyn and Haider-Markel, 2007; Maoz and McCauley, 2009), environmental degradation (Homer-Dixon, 1999; Klare, 2000), immigration (Burns and Gimpel, 2000; Givens et al., 2009; Huysmans, 2006), racism (Kinder and Sears, 1981), and the economy (Clarke et al., 2004; Kinder and Kiewiet, 1981). For example, research on immigration has shown how societal threats are largely symbolic and sociotropic as ‘outsiders’ are presented as posing an existential

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

29

challenge to a supposedly unified sovereign political community (Givens et al., 2009; Huysmans, 2006). Similarly, Kinder and Sears (1981) find that the impact of racial threat on voting behaviour is due to perceptions of collective and symbolic threat rather than concerns at the individual level. Political tolerance is lower in nations that are faced with external threats to national territory (Hutchison and Gibler, 2007). Huddy et al. (2002) argue that national threat is far more influential on perceptions of the economic consequences of terrorism, while personal threat is more likely to alter behaviour designed to mitigate threat, such as changing air travel habits to avoid the possibility of harm from a terrorist attack. They conclude that, as in other policy areas, the influence of the personal on political judgements is limited. Furthermore, the social conformity preferences that some argue are at the heart of authoritarians’ political intolerance and willingness to restrict civil liberties are related to threats to social cohesion at the national level (Feldman, 2003), hence Stenner’s (2005: 32) description of authoritarians as ‘relentlessly sociotropic boundary maintainers’. These differences in the effects of sociotropic versus personal judgements on political attitudes and behaviour have attributions and information as their core explanations. Although sociotropic factors are frequently described as more remote and less vivid than personal concerns (e.g., Huddy et al., 2002; Lavine et al., 1996), individuals are said to be more able to make the connections between societal conditions and government than they are between their own circumstances and the actions of public officials. It may also be the case that individuals simply view sociotropic factors as more reliable indicators of the likely personal impact of government policies than personal indicators (Kinder and Kiewiet, 1981), and, relatedly, because they present the greater threat (Kinder and Sears, 1981; Stenner, 2005). Sociotropic and personal situations may also bring different values and considerations to the fore: sociotropic threat may call to mind a value like freedom of expression, whereas personal threat renders the value of safety more salient (Chanley, 1994). Feldman (2013) suggests that the personal threat of terrorism raises issues of individual autonomy, whereas national threat is more likely to render salient considerations of group conformity, hence authoritarians’ greater concern about the latter kind of threat than the former (though see Asbrock and Fritsche, 2013). Another argument is that there is more media coverage of national than local circumstances, more contextualising of national conditions, and media coverage tends to prime – that is, bring to the front of people’s minds – sociotropic rather than personal concerns (Mutz, 1992). Finally, in many policy areas individuals may lack the direct experience necessary for personal considerations to weigh heavily (Clarke et al., 2009). Yet there is also evidence to suggest that national considerations are not as dominant as previous research would have us believe. Some studies have found that there are attitudes and judgements for which personal considerations are more salient than national (Chanley, 1994; Huddy et al., 2002; Joslyn and

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

30

Everyday security threats

Haider-Markel, 2007; Schildkraut, 2009), including authoritarianism (Asbrock and Fritsche, 2013). Galea et al. (2002: 985) show, for example, that ‘Persons directly affected by disasters have higher rates of post-event psychiatric disorders than persons indirectly affected’ (see also Hirsch-Hoefler et al., 2014). We might extrapolate from this that perceptions of personal security threats similarly have stronger psychological effects than more distant threats to the nation or to the world. Indeed, Canetti-Nisim et al. (2009) and Hetherington and Suhay (2011) argue that it is physical threats to the individual – ‘the self as vulnerable to victimization’ (Canetti-Nisim et al., 2009: 369) – that are particularly salient with regard to terrorism, while Fernandez and Kuenzi (2010) find that perceptions of personal safety from crime are of more importance than national economic performance to support for democracy in Africa and Latin America. Jackobsson and Blom (forthcoming: 9) contend, ‘How people perceive their own risk of being a victim of a terror attack and the motive of the terrorists seem to play an important role in determining how people will respond to an attack’ (emphasis added). Gomez and Wilson (2001, 2006) argue that perceptions of personal economic circumstances are more salient to those high in political knowledge than are national economic circumstances. In addition, Maoz and McCauley (2009) contend that sociotropic and personal influences can be opposite in effects: Jewish-Israeli support for compromise with Palestinians was negatively affected by perceptions of collective threat but positively associated with personal fear. Burns and Gimpel (2000) also find that the impact of national and personal perceptions of the economy on attitudes towards immigration sometimes appear to vary in direction, with national perceptions associated with more hostility and personal perceptions with less hostility. Other research also suggests that the national–personal distinction is too limited in two respects. For example, some studies of economic effects claim that globalisation has dampened the influence of national factors while making international economic indicators more salient to individuals (Burden and Mughan, 2003; Hellwig, 2001),9 i.e., ‘the global’ is of growing importance relative to the national. Furthermore, while the consensus is that personal considerations tend to carry less weight, there is plenty of evidence that both personal and subnational considerations can matter. Chong et al. (2001) argue, for example, that personal considerations exert influence on reasoning when an individual’s ‘stakes in the policy are clear’. Clarke et al. (2009) argue that ‘pocketbook issues’ are central to the kinds of valence judgements about the way alternative governments would ‘manage’ the country that influence British voting behaviour. Moreover, Huddy et al. (2002) note that perspectives on the influence of personal threat could be limited by dependent variables that are frequently related to national consequences. Jones et al. (1992) demonstrate the importance of local context and Johnston et al. (2000) of local unemployment, in particular, to

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

31

voting behaviour in Britain. Similarly, studies in the US have shown subnational influences on economic perceptions (Niemi et al., 1999), support for social welfare spending (Kam and Nam, 2008), immigration (Hopkins, 2010), opinion on the Vietnam War (Gartner and Segura, 2000), and voting behaviour (Glasgow, 2005), while Berrebi and Klor (2008: 288) contend that ‘the consequences of terrorism are mostly felt and manifested at the local level’. Even research that shows a more pervasive influence of sociotropic than of personal considerations often finds that subnational considerations matter too, but, according to Lau and Heldman (2009: 535), only ‘sporadically, here and there and under particular circumstances’. Indeed, as we have already outlined, Schildkraut (2009) finds pervasive effects of personal perceptions on more specific counter-terrorism policies. Tyler’s (1982) work suggests that such effects may be a consequence of ‘defensive attribution’: the more personally threatened individuals are, the more they hold national actors responsible for their mitigation. The postulated mechanisms for the salience of non-national considerations are similar to those described above for the dominance of national-level concerns (but with different results). Thus, one factor is said to be clarity of responsibility – globalisation blurs responsibility for the economy, for example (Burden and Mughan, 2003) – while personal considerations may be more important when threats are specific and when there is a clear link between personal concerns and government responsibility (Joslyn and Haider-Markel, 2007). Another is the news media: international economic indicators have become more salient as the media have given them attention. Personal threats, particularly physical threat from external forces such as terrorism, may be associated with different emotions to national threats and thus lead to different attitudes and behaviours (Hetherington and Suhay, 2011; Maoz and McCauley, 2009). In addition, elites have framed contemporary security threats such as terrorism as global, as when former British Prime Minister Tony Blair said, ‘September 11 changed the world  . . . Bali, Beslan, Madrid and scores of other atrocities that never make the news are part of the same threat’ (quoted in Croft and Moore, 2010: 825). In sum, the limitations of the national–personal dichotomy are twofold. First, it unnecessarily restricts understanding of the effects of ‘levels’ of threat. There is no logical reason why sociotropic threat should be confined to the nation or subnational threat to the individual or family rather than to the community. Second, measurement of the national–personal dichotomy provides little theoretical or empirical guidance as to whether perceptions of national, let alone global threats have the same kinds of determinants as personal threats because the assumption that these are the salient levels of considerations leads to a focus on their effects rather than their causes. Yet Chanley’s (1994) research suggests that for an issue such as global warming, conceiving of it as a concern to the planet rather than as a threat confined to individual nations may connote a

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

32

Everyday security threats

different political outlook and the salience of different value considerations. Lee et al.’s (2009: 435) research in Canada also provides a suggestive example in which one of their focus group participants says of terrorism: ‘We’re in Canada. It could happen, there, but it worries me in a global sense, the repercussions, international politics with the USA and other countries and the lifestyle change that it brings to us in North America.’ This participant thinks of terrorism in global terms, but another might see the ‘world’ primarily in terms of national boundaries or the physical threat it presents to them as an individual (see also, Reifler et al., 2011; Ridout et al., 2008: 579–80). Indeed, Huddy et al.’s (2005) research suggests that authoritarianism may be one factor in this. They find that authoritarianism is positively related to preferences for military action and limiting civil liberties in the wake of 9/11 but negatively related to the desire for the US to be active in world affairs, and unrelated to the desire to see the US take the leading role in solving international problems. This implies a view of the world that, far from simply being ‘ten steps closer to the panic button’, wants action to resolve national disorder and is unsympathetic to action designed to mitigate international disorder (see also Duckitt and Sibley, 2010). Kahan’s research on ‘cultural worldviews’, drawing on the ‘cultural theory’ of Douglas and Wildavsky (1982; see also Wildavsky and Dake, 1990), while not directly about tendencies to see problems as sociotropic or personal, is also suggestive here in that it points to systematic differences between individuals over the meaning of various issues and thus about societal dangers (Kahan, 2012; Kahan and Braman, 2006). Although there are some gender differences, these worldviews have reliable impacts on perceptions: hierarchists tend to be sceptical about issues like environmental risk and the virtues of gun control. According to Kahan (2012: 727, quoting Douglas and Wildavsky, 1982: 8, 85), ‘Each way of life and associated worldview “has its own typical risk portfolio”, which “shuts out perception of some dangers and highlights others”, in manners that selectively concentrate censure on activities that subvert its norms and deflect it away from activities integral to sustaining them.’ Cultural predispositions also affect responses to messages about danger: the same set of facts may be rejected or accepted contingent on whether a message threatens or confirms cultural identity – ‘culture is prior to facts in individual cognition’ (Braman et al., 2005: 2). Our contention is that such differences may also be reflected in tendencies to see threats at one level – as global, for example – rather than another, e.g., as personal. Our claim here is not that cultural worldviews and perceptions of risk tell us about the levels at which individuals will experience threat. But Kahan’s theory does lend itself to the notion that there may be systematic influences, partly due to different ways of thinking about the world, that mean that some individuals are more likely to experience a threat or threats as global while to others a threat or threats are less likely to be identified as global than as personal.



Perspectives on security threat politics

33

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Theorising the origins and consequences of threats It is possible to identify several common trends among existing qualitative research on public perceptions of security threats in Britain associated with the ‘vernacular turn’. Owing to the nature of this work – its specificity in terms of place and time, its in-depth and non-generalisable nature, and its non-replicable research design – we do not seek to ‘test’ these insights against our own findings. Rather, we outline here several key themes for comparative analysis, which will allow us in subsequent chapters to relate to and build upon existing arguments. This approach has a multiplier effect in that it adds cumulative value and validity to what may otherwise be considered idiosyncratic conclusions based on smallscale projects. Narratives of security

Jarvis and Lister (2013: 164–5) offer a typology of six images of security arising from their focus group work across London, Birmingham, Oldham, Swansea, Llanelli, and Oxfordshire, namely: 1) survival (basic needs for living); 2) belonging (‘feeling situated in a particular spatial or human community’); 3) hospitality (‘organised around a need for others to recognise one’s own right to belong in a social space’); 4) equality; 5) freedom; and 6) insecurity (scepticism about the possibility of achieving security). Other studies have stressed the ambivalence, contradictory and messy nature of security narratives concerning threat perception (Gillespie and O’Loughlin, 2009a: 681). Nevertheless, a point of convergence is that among minority groups, particularly those who self-identify as Muslims, narratives of (in)security and injustice in the face of heightened societal vigilance are relatively commonplace (Gillespie et al., 2010: 246; see also Jarvis and Lister, 2012; Gillespie and O’Loughlin, 2009a: 668). Referent object

In mobilising security as a category of understanding, previous qualitative studies have found that participants in focus groups rarely think in terms of the state. Rather it is more common among the literature to find the individual as the referent object, thereby associating ‘vernacular’ theories with ‘human security’ approaches (Jarvis and Lister, 2013) and the question of ‘ontological security’ (Gillespie and O’Loughlin, 2009a). Terrorism

Despite the significance attached to the perceived threat of terrorism by Western governments in particular, past studies in the UK context after 9/11 have

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

34

Everyday security threats

suggested that this is not a pressing issue at the level of the everyday (Gillespie and O’Loughlin, 2009a: 668; see also Gillespie et al., 2010). The ‘Shifting Securities’ project, for example, found that, even around the time of the London bombings in 2005, terrorism was not considered to be a major threat to daily routines (Moss and O’Loughlin, 2008: 712). Rather, threats at that time were found by Marie Gillespie and her team of researchers in Swansea, Oldham, Edinburgh, Belfast, Bradford, Leeds, London, and Oxford to be more likely connected to ‘job and financial insecurity, environmental problems, and local crime’ (Gillespie and O’Loughlin, 2009a: 668). Political agency

Previous research has emphasised that many participants in ethnographic studies characterise themselves as passive political actors who ‘depict a world in which forces “out there” control events’ (Gillespie and O’Loughlin, 2009a: 675; see also Gillespie et al., 2010: 246). In this regard, Jarvis and Lister (2012: 15) have drawn a link between the rise of increasingly intrusive counter-terrorism measures on the one hand and heightened narratives of political alienation and incidences of extremism among minority groups on the other hand. Similarly, Moss and O’Loughlin (2008: 705) refer to a ‘new security dilemma’ in which ‘the government pursues actions designed to increase the security and well-being of all citizens, but these actions simultaneously appear to victimise, alienate and add to the insecurity of significant sections of the British citizenry’. Previous research in the Political Psychology and Political Behaviour literatures has provided mixed perspectives on the origins of threat perceptions. However, the kinds of threats that have been examined have been limited by virtue of the dominant national–personal dichotomy in measurement. Nevertheless, drawing on this literature, we analyse a range of potential influences on perceptions of the breadth of threats and on perceptions of specific threats such as terrorism and environmental degradation, from dispositions such as mortality salience and authoritarianism, through media habits, to individual characteristics such as education, age, and religion at the four levels (global, national, community, and personal). Mortality salience

Existing research tells us that individuals for whom thoughts of mortality are most accessible are prone to manifest feelings of threat and danger by defending their cultural worldviews against perceived challenges from outgroups. The heightened sense of threat induced by ‘mortality salience’ increases ‘the positivity of evaluations of people and ideas that support one’s cultural worldview and the

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

35

negativity of evaluations of people and ideas that threaten it’ (Schmiel et al., 1999: 906). Why? Because humans want to ‘survive’ after death, the continuation of life as they know it is a form of survival, and awareness of their own mortality coupled with potential threat to their cultural worldview therefore provokes anxiety (Das et al., 2009). However, this research is usually based on manipulation of mortality salience rather than directly assessing the influence on threat itself (e.g., Canetti-Nisim et al., 2009; Greenberg et al., 1990); our interest is precisely in the relationship between ‘dispositional’ mortality salience, i.e., unmanipulated, and perceptions of threats. Given that mortality salience is influential under conditions of mortal physical danger or threats to cultural worldview and is not aroused by mere insecurity, uncertainty, or anxietyproducing events (Schmiel et al., 1999), we would expect that mortality salience enhances perceptions of the breadth of global and national threats, where terrorism and immigration loom large, but not perceptions of subnational threats, which appear from Figure 3.1 (see Chapter 3) most strongly related to economic insecurity and the anxiety produced by crime. Authoritarianism

Claims such as Altemeyer’s (1996) about authoritarian sensitivity to threat would lead us to expect both that authoritarianism will affect perceptions of threats and that the influence will be consistent across levels. However, other research on authoritarianism argues that it is not a cause of elevated threats but is activated for those higher in authoritarianism when social order is threatened (Stenner, 2005),10 or that it is a cause of perceptions of threat but only in relatively benign societal conditions (Hetherington and Weiler, 2009: 110). We are agnostic, but clearly the latter theories suggest we may not find a direct relationship between authoritarianism and perceptions of threats. Media habits

In arguing, contrary to Galea et al. (2002), that post-traumatic stress disorder symptoms caused by 9/11 spread much further than those people directly affected in New York, Marshall et al. (2007: 309) view media exposure as a primary cause: ‘Because media channels focus on novel information and novelty increases the fear response, they inevitably promote the distortion of signal intensity, thereby amplifying it throughout the communication-saturated environment of the United States.’ Ridout et al. (2008) indicate that media exposure heightens perceptions of ‘modern’ and ‘postmodern’ threats, while Gillespie and O’Loughlin (2009a: 677) refer to interviewees who ‘exhibited a strong mistrust of media, yet simultaneously expressed a fear to threats such media conveyed or represented’. Indeed, media coverage of terrorism in particular ‘may increase

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

36

prejudice against different outgroups (i.e., Arabs, Europeans), regardless of whether they are linked to news content’ (Das et al., 2009: 458). Previous research also suggests that the news media may exacerbate perceptions of threats (Boomgarden and de Vreese, 2007) through such tendencies as focusing on conflict and catastrophe – on raised but not lowered threat levels and on incidents of threat rather than threat preparedness, for example (Nacos et al., 2007) – and limiting appearances of minorities to stories about crime (Gilliam and Iyengar, 2000; Hurwitz and Peffley, 1997). Education

Education has been associated with perceptions of lower threat levels as a result, it is argued, of mechanisms such as highly educated individuals having a greater grasp of probabilistic information (Huddy et al., 2005), e.g., the likelihood of a terrorist attack being low or the probability that immigration, far from being costly to the economy, may be necessary for future economic prosperity given low indigenous birth rates. Race, religion, age, and gender

Gender and religiosity may also heighten threat perceptions (Goodwin et al., 2005; Norris et al., 2002), while age may have a negative effect on perceptions of threat (Canetti-Nisim et al., 2009; though see Huddy et al., 2005). Finally, British Muslims ‘suffered disproportionately greater levels of stress than respondents from other faiths’ following the 7/7 bombings in London (Rubin et al., 2005). Thus, we might expect different perceptions of the breadth of global threats from British Muslims because of a different perspective on issues such as religious extremism. Adopting a similar logic about possible differences due to ethnicity, we might also expect an influence of white ethnicity. Consequences

We mentioned above that research on specific threats, from terrorism to immigration or environmental degradation, has examined a disparate array of consequences. In our reading, however, we can group these into four categories. Aggression and intolerance Research on appraisal-tendency theory (Lerner et al., 2003) and mortality salience (McGregor et al., 1998) in social psychology suggests that fear is associated with more thinking about one’s mortality and with support for more punitive

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Perspectives on security threat politics

37

approaches from governments. Studies of perceptions of threat from terrorism have confirmed this relationship (Fischer et al., 2007; Malholtra and Popp, 2012; Merolla and Zechmeister, 2009). However, Finseraas and Listhaug (2013) find that while the Mumbai terrorist attacks raised the fear of terrorism in Europe – the fear of what we are calling ‘terrorism at the global level’ – they did not prompt a desire for more illiberal interrogation techniques. Voors et al. (2012) demonstrate that exposure to the violence of civil war in Burundi was associated more with altruistic behaviour rather than hostility towards others (though see Hirsch-Hoefler et al., 2014). Thus, while there appears to be a relationship between perceptions of security threats and aggression and intolerance, past research also suggests variation that is not well understood. Stereotyping minorities and the desire for stricter criteria for immigration and citizenship Similarly, previous studies have indicated a relationship between threats and attitudes towards minorities and outgroups, including enhancing the stereotyping of, and antipathy towards, immigrants and cultural and religious minorities (Burke et al., 2013; Croft, 2012; Green, 2009; Greenberg et al., 1990; Kalkan et al., 2009; Landau et al., 2004; Pyszczynski et al., 2002). The Madrid bombing in 2004 resulted in greater hostility not only towards Arabs, who shared the ethnicity of the perpetrators, but also towards Jews, suggesting that threats may lead to a broader lashing-out against outgroups (Echebarria-Echabe and Fernandez-Guede, 2006). Finseraas et al. (2011) also find some effects on attitudes towards immigration in Western Europe following the Theo Van Gogh murder in the Netherlands in 2004 (see also Das et al., 2009). On the other hand, more distant, perhaps more global threats, may not affect attitudes to immigration (Finseraas and Listhaug, 2013). Public policy preferences Several studies have shown that the threat from terrorism renders individuals more willing to exchange civil liberties for security measures (Bozzoli and Muller, 2011; Davis and Silver, 2004; Sanquist et al., 2008). Perceptions of security threats are also associated with support for military action in the US (Huddy et al., 2005; Kam and Kinder, 2007) and for stricter border control (Kam and Kinder, 2007; Merolla and Zechmeister, 2009). The question is whether threats more generally lead to what Burke et al. (2013) describe as a ‘conservative shift’, in which policies directed at mitigating security threats are prioritised by individuals over other policy areas such as education or other spheres, such as international aid, that may have fewer direct effects on mitigating certain security threats.

38

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Leadership preferences Perceptions of threats may lead to a general conservative shift or to a preference for right-wing parties (Berrebi and Klor, 2008; though see Willer and Adams, 2008),11 resulting in a tendency to prefer leaders with particular traits such as strength and charisma (Merolla and Zechmeister, 2009). They may also polarise the electorate (Berrebi and Klor, 2008), which would imply that the stakes in voting become higher, rendering individuals more motivated to vote in an election (Robbins et al., 2013). Conclusion In this chapter, by reviewing previous research in both IR and Security Studies and Political Psychology and Political Behaviour on the origins and effects of threats, we have sought to make clear our own contributions and their rationales. First, this is, to our knowledge, the first study to examine perceptions of the breadth of threats rather than specific threats and their intensity. In so doing, we argue that this measure both provides an additional understanding of how threatened people are, and also allows a deeper exploration of contemporary understandings of security threats than heretofore. Second, by not confining ourselves to a single threat such as terrorism we are able to generalise about how individual-level variables from mortality salience and authoritarianism, to media habits, affect perceptions of security threats and where and why there is variation. Such analysis provides broader knowledge of threats and a firmer foundation on understanding of their origins. Third, we extend the levels at which threats are perceived from the national versus personal dichotomy that has dominated the literature to a continuum spanning the individual, community, nation, and globe, and we also show the extent to which perceptions of threat at each level have different causes. Fourth, in triangulating the insights of qualitative work associated with the ‘vernacular’ and ‘everyday turns’ – and ontological security studies more generally – with existing work produced by quantitative analyses of threat perception, we are able to offer a broader basis and richer intellectual context in which our study as a whole might be located. Indeed, one of the distinctive features of this book is our post-positivist mixed-methods research design, and the critical self-reflexivity that this affords, which is precisely the substantive focus of Chapter 2. Notes 1 We are grateful to Stuart Croft for his input on this topic. 2 Stenner (2005: 69–71) distinguishes between ‘normative threat’ – the threat of a changeable society – and ‘perceptions of a dangerous world’. Authoritarians are no more likely to be normatively threatened, she argues, but they are more likely to be generally fearful



3

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

4 5

6 7 8 9

10

11

Perspectives on security threat politics

39

of ‘disorder’, ‘chaos’, and ‘anarchy’, or a dangerous world. Hetherington and Weiler (2009: 8) also argue that authoritarians are more likely to see threats. There is disagreement about whether authoritarianism is a personality trait or a predisposition. To ease the flow of discussion, we refer to it as a predisposition from now on. Indeed, according to Slovic (2000), ‘risk’ and ‘threat’ are synonymous. By which, borrowing from work such as Beck (1992) and Inglehart (1997), they mean a distinction between ‘traditional threats’ based on physical survival or material scarcity – such as those from war and economic deprivation – and those characteristic of postmodern concerns about quality of life, like global warming and population growth. A focus on extreme stimuli also precludes analysis of the extent to which perceptions of threat change over time. A rare exception is Ridout et al. (2008), where sociotropic threat is global threat. This is also true of risk perceptions (Sjoberg, 1999). The causes are related to those usually given for greater national than personal effects – globalisation blurs responsibility for the economy, and international economic indicators have become more salient as the news media have given them more attention. Stenner (2005: 69) refers to ‘perceptions of a dangerous world’ – ‘a persistent fear of societal “disorder”, “chaos”, and “anarchy”’ (italics in original) – as distinct from normative threat. Our measure of security threats appears to be broader than these perceptions. Notwithstanding that the Madrid bombings appeared to have the opposite effect, with the incumbent conservative government being replaced by the socialists just a few days after the event (Montalvo, 2011). The reasons seem to have been the government’s initial hasty and inaccurate explanations for the bombings and that the bombings brought the government’s unpopular support of the war in Iraq into sharper focus (Bali, 2007).

2

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

The 2012 study ‘Public Perceptions of Threat in Britain’

Introduction The intellectual puzzle animating this book as a whole is not simply the tension, as discussed in Chapter 1, between the heightened prominence of the citizen in national security policy discourse and the persistence of elitist perspectives across ‘mainstream’ and ‘critical’ IR and Security Studies. In identifying this gap scholars seeking to better understand non-elite knowledge, experience, and perceptions of the contemporary politics of security threats at the level of the everyday face a significant challenge. This challenge, we argue, is primarily methodological because vernacular constructions of the concept of security, public encounters with and negotiations of security, and factors influencing the way citizens perceive and narrate threat are both ubiquitous and private, everywhere and nowhere, and visible and invisible. Moreover, this methodological challenge, as we see it, is more than the truism that diverse publics’ views, perceptions, and experiences are difficult to ‘access’: in any case, this notion relies on a problematic assumption that the subject of research holds fixed opinions across time and space and that the role of the researcher is merely to operationalise the ‘correct’ method in order to gather these pre-existing data. By contrast, our understanding of the nature of the task presented by our research aims and objectives – anchored in a post-positivist approach to the philosophy of social science and a derivative understanding of the concept of ‘security threat’ (see below for a fuller discussion) – is one that seeks to appreciate the performative effects of our own methodological enactments (see Aradau and Huysmans, 2014; Booth et al., 1996; Krause and Williams, 1997). From such an approach, any attempt at researching vernacular (in)security at the level of the everyday – as is ultimately the case with all forms of social inquiry – is going to reflect certain philosophical assumptions, which inevitably mean that data are always already co-produced between ‘researcher’ and ‘researched’ to the extent that the ability to make that very distinction is challenged (Kurowska and

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

41

Tallis, 2013). Importantly, however, to acknowledge these inescapable dimensions of social scientific research is not to say that any attempt to investigate vernacular security empirically is somehow fruitless or destined to remain caught within a hermeneutic trap and thereby a pointless exercise: methodologically and politically, this would be to support and reproduce the very elitism that continues to pervade the subfield of IR and Security Studies on the issue of threat politics. Rather, our pragmatic stance is one that readily admits that underpinning epistemological and ontological positions – conscious or otherwise – inescapably inform, shape, and curtail methodological positions. In turn, these positions govern the use of possible methods, which enact particular social realities and power relations and may therefore have significant political implications (Aradau and Huysmans, 2014). Ultimately, this leads to a disposition that seeks at the very least to demonstrate an awareness of these performative effects and to project them into a field of study rather than pretending that they have no bearing on conclusions that are drawn (Connolly, 1993). Nevertheless, what we believe makes these generic issues and dilemmas particularly challenging in the specific case of the attempt to investigate contemporary public perceptions and experiences of security threats is that the scale and complexity of the research problem demands the triangulation of a range of methods not normally used together in this field of research. Whereas mixed methods approaches are increasingly common in other areas of Political Science, such as Comparative Politics and Public Policy, scholars seeking to better understand the politics of security threats have tended to work either with small-scale qualitative methods such as focus groups in the context of IR and Security Studies (Gillespie, 2007; Gillespie et al., 2010; Gillespie and O’Loughlin, 2009a; Jarvis and Lister, 2012, 2013, 2015; Moss and O’Loughlin, 2008) or with large-N data sets based on quantitative methods such as surveys and survey experiments in Political Psychology and Political Behaviour (Hirsch-Hoefler et al., 2014; Huddy et al., 2005; Sniderman et al., 2004), but not both.1 While we agree with Aradau and Huysmans (2014: 10) when they argue that ‘using statistical analysis to identify social patterns often excludes the worlds of individual experiences and the lives that are lived but deviate significantly from the pattern’, the reverse is equally true. That is to say, focusing only on the experiences of individuals or small groups potentially misses the insights that can be gained only from attempting to map social relations, non-elite repertoires of knowledge, and patterns in how and what diverse publics think and perceive. We argue that a new approach that triangulates methods in the study of everyday security threat politics is therefore necessary. One of the possible reasons for the reproduction of this fissure between the respective literatures in the study of security threats is what we consider to be the problematic assumption that qualitative and quantitative methods cannot be used alongside each other for methodological reasons. For us, however, the key

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

42

Everyday security threats

distinction here is not between qualitative and quantitative at the level of methods understood as ‘material devices that enact worlds’ (Aradau and Huysmans, 2014: 9), but instead the underlying philosophical division between positivist and post-positivist approaches to social science (Booth et al., 1996). Once the automatic alignment is challenged between ‘positivist methodology’ and ‘quantitative methods’ on the one hand and ‘post-positivist methodology’ and ‘qualitative methods’ on the other, it becomes possible to see how a mixed methods approach can be mobilised reflecting a particular methodological stance without risking meta-theoretical or philosophical incoherence (Smith, 1996). As we shall go on to outline in greater detail over the course of this chapter, our study operated on the basis of a post-positivist methodological research design that sought to mobilise a mixed methods approach in response to five key aims arising from the limitations of previous research: 1 to gauge everyday understandings and experiences of the concepts of ‘threat’ and ‘security’; 2 to investigate what diverse publics consider to be the most pressing threats to (their particular conceptions of) security; 3 to explore the individual-level influences on perceptions of security threats; 4 to examine how and whether perceived threats to security – both in terms of perceptions of the breadth of security threats and perceptions of specific security threats such as terrorism – influence other political attitudes such as tolerance of outgroups; and 5 to assess whether non-elite knowledge and views coincide with or diverge from what the NSS presents as the greatest threats to British security and as effective ways to mitigate them. The term ‘mixed methods’ has a variety of meanings across the social sciences. In recent work in Political Science, for example, it often refers to combining quantitative data with detailed case studies: the large-N quantitative analysis informs the choice of small-N cases, which provide additional context and explanation (Lieberman, 2005; Weller and Barnes, 2014). Mixed methods may also simply mean the use of different techniques to examine the same set of qualitative or quantitative observations (Brannen, 2005). A definition of mixed methods from educational research is: ‘the class of research where the researcher combines quantitative and qualitative research techniques, methods, approaches, concepts or language into a single study’ (Johnson and Onwuegbuzie, 2004: 17). What we mean by a mixed methods approach is precisely this combination of qualitative focus group interviews and interpretive analysis with quantitative survey data and statistical modelling.2 Even within this kind of qualitative-quantitative mixed methods research design, however, there is a variety of possibilities. Brannen (2005) describes mixed methods research designs in which either the qualitative or quantitative

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

43

component is dominant, or in which the two play an equal part (see also Creswell, 2003). In our research they played an equal part in a three-stage project. We used each stage of the project to reflect and build on the previous one, enhancing our understanding of the data by using the strengths of one method to help us address the deficits in understanding inherent to the other. This point is of particular significance not only for the sake of ‘balance’ between methods, but also for the way in which it speaks precisely to the kind of self-reflexivity and awareness of inevitable limitations that is a hallmark of post-positivist research design (Smith, 1996). Because the purpose of the study was to ask how and why diverse publics view and construct various issues as threatening, at no point did we impose our own understanding of what ‘security’ or ‘security threats’ might mean on our participants in either the focus groups or survey work. Rather, throughout the project, our use of these concepts was immanent to the way in which participants understood and talked about them at the level of the everyday and in their own language. In the first stage of the research the rich and interactive dynamics of openended focus group discussion gave rise to a series of understandings, definitions, evidence-bases, and scales concerning security threats, which we took as the starting point for the programme of research as a whole. During the second stage, these vernacular meanings and measurements were then used to inform the design of the large sample survey, which enabled the kind of statistical assessment of key relationships and subgroups that qualitative work on its own cannot provide. Finally, the third stage meant that we could take the results of the survey work and use our findings to inform a revised moderation guide for a second round of focus group work in an iterative fashion. In this way, and commensurate with the complexities of the research problem, our 2012 study incorporated both methods in order to combine representative macro-level insights into public opinion across key variables with non-representative micro-level thick descriptive accounts of individuals’ everyday anecdotes and stories. This chapter is organised into four main sections. First, we begin by outlining and discussing the post-positivist methodological principles underpinning our research design in greater depth. Second, we present our rationale for triangulating quantitative and qualitative methods in the light of our research aims and objectives. The third and fourth sections offer a detailed synopsis of how we went about the fieldwork phase of the project, focusing on each method, respectively. Finally, we conclude with a commentary on our approach to data analysis, which frames subsequent chapters of this book. A post-positivist methodology Bernstein (1983) discusses ‘tyranny of method’, while Tashakkori and Teddlie (1998: 3) talk about ‘paradigm wars’. Johnson and Onuwuegbuzie (2004) refer to ‘purists’ who, regardless of research puzzles and questions, advocate the

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

44

Everyday security threats

exclusive use of either qualitative or quantitative methods. Whether tyrants or purists, the point is the same: ‘purists’ are prone to caricature the epistemological assumptions underlying the other approach, drawing on false contrasts and misinterpretations. We view both ‘sides’ of the debate as fundamentally problematic and unproductive in the light of the complexity of the demands of research into contemporary security threat politics. The tendencies referred to above are often reflected in the ‘positivist’ label given to quantitative researchers, ‘that through overuse and misuse, has become an almost meaningless term of abuse’ (Scott and Marshall, 2009: 581). Indeed, paraphrasing Bryman (1984), Brannen (2005: 7) argues that, ‘qualitative researchers have spent more time defining quantitative methods than quantitative researchers have themselves’. In the theoretical literature produced in the field of IR a landmark publication in this regard was the volume International Theory: Positivism and Beyond edited by Booth et al. (1996). Smith’s (1996: 11) seminal tour d’horizon of epistemological debates in IR portrays positivism as ‘a commitment to a unified view of science, and the adoption of the methodologies of the natural sciences in order to explain the social world’. On this view, during the latter half of the twentieth century the discipline of IR was largely ‘positivist’, characterised by four specific features: 1) a belief in the unity of natural and social science; 2) a distinction drawn between facts and values (and the notion that facts are theory-neutral); 3) a belief in the existence of regularities in the social world; and 4) an adoption of an empiricist epistemology (Smith, 1996: 15). While this textbook characterisation of ‘positivism’ is not entirely deceptive – if somewhat devoid of nuance, as we shall go on to argue – it has often been translated into rather more extreme portrayals. In her work on causation, for example, Kurki (2008: 15, 209) depicts quantitative researchers as ‘straitjacketed’ – restricted by their concentration on ‘observable-based variables’ and ‘the “superficial” level of empirical observation and generalisation’. As a result, she argues that quantitative analysts are unable to ‘theorise the complex “underlying” structures, relations and processes of world politics’ (Kurki, 2008: 15) and are primarily interested in prediction above explanation. On her account, such scholars find it uniquely difficult to ‘evaluate the complex interactions of various causal forces’ and do not recognise that ‘there might be other, non-statistical, non-Humean ways of dealing with causal complexity’ (Kurki, 2008: 118, 119). They stand accused of measuring causal effects ‘on the basis of measures of statistical significance’ and do not ‘recognise the causal role of unobservables – such as ideas, rules and discourses’, which they conceive of simply ‘as individual “mental states”’ (Kurki, 2008: 194, 170, 223). By the same token, advocates of a quantitative approach to research such as King et al. (1994) simplify and underestimate the potential of qualitative methods (Brady and Collier, 2004; Eby et al. 2009; Flyvbjerg, 2006; Maxwell, 2013). This frequently derives from misunderstandings about what ‘data’ are for qualitative

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

45

researchers, e.g., ‘Often quantitative researchers are surprised when listening to qualitative researchers describe their findings, based on “data”, given that frequently there are few numbers discussed in the study (“Where are the data?”) . . .  Words are of utmost importance to qualitative researchers’ (Given, 2008: 190). Similarly, ‘data’ can be less easy to collect than quantitative researchers assume, with implications for some of the touchstones of the quantitative method, such as hypothesis testing and replicability. For example, Ragin (2004: 126) responds to the advice that qualitative researchers should collect more data in order to test revised versions of their theories with the objection that ‘When the number of relevant cases is limited by the historical record to a mere handful, or even to several handfuls, it is simply not possible to collect a “new sample” to “test” each new theoretical clarification’. Finally, quantitative researchers are prone to assume a particular logic of causal inference – what Mahoney (2010: 133) describes as ‘the average effects of independent variables’, that is often not the relationship(s) of interest to qualitative researchers. These tendencies towards caricature allow tyrants and purists on both sides of the debate to ignore developments and debate within allegedly separate paradigms. Over the past twenty years, the ‘positivist’ research paradigm outlined by Smith (1996) has evolved considerably in the context of Political Science and IR. Indeed, many researchers using quantitative methods do not subscribe to the first two of Smith’s characteristics of positivistic forms of social inquiry. Moreover, while statistical analysis is certainly interested in regularities (and irregularities), this is not predicated upon an unproblematised empiricist epistemology influenced by a Humean understanding of causation – the idea that causal relationships are invariant is not subscribed to by a majority of quantitative researchers (Smith, 1996: 15). Equally, although for Smith behaviouralism entails a ‘disregard for what goes on inside actors’ heads’, many researchers associated with that approach today see the purpose of their work as being driven precisely by that challenge (Smith, 1996). In general terms, we are often told that ‘those writing about quantitative research typically give much less attention to epistemological and ontological assumptions’ (Brannen, 2005: 7). But by now, however, the vast literature on the meaning of survey response, differences in survey questions, interviewer effects, non-response bias, measurement equivalence, latent variables, and modes of data collection are testament to greater degrees of selfreflexivity among many of those using survey methods (e.g., Billiet et al., 2007; Bollen, 1989; Collier et al., 2004; Cor et al., 2012; Davidov et al., 2008; Lacy, 2001; Sanders et al., 2007; Tourangeau et al., 2000; Tourangeau and Smith, 1996; Yeager et al., 2011; Yeager and Krosnick, 2012; Zaller and Feldman, 1992). Indeed, there is far greater critical reflection about the role of theory, evidence, and the question of ‘objectivity’ than writers such as both Smith (1996) and Kurki (2008) otherwise imply. For this reason, the label ‘post-positivism’3 would seem to us to be apt to describe much quantitative work and it is certainly a term that we

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

46

Everyday security threats

would use to characterise our own approach to both the survey and focus group design. Using the term ‘post-positivism’ also allows for greater acknowledgement that the epistemological differences between qualitative and quantitative research have greatly narrowed and that in many cases the boundaries between the two paradigms are in fact somewhat blurred. For instance, as Brady and Collier (2004: xvii) write in the preface to their response to King et al.: Henry E. Brady, who is primarily a quantitative survey researcher, repeatedly finds that he must come to grips with interpreting the meanings conveyed in survey responses and with comprehending the qualitative complexity of the political behavior he studies in various national contexts. David Collier, who is primarily a qualitative comparativist, recognizes that it is sometimes productive to quantify concepts such as corporatism and democracy, the historical emergence of labor movements, and the international diffusion of policy innovations.

In addition, like Brady et al. (2004: 7), who say that ‘quantitative and qualitative methods are founded on essentially similar epistemologies’, Johnson and Onwuegbuzie (2004: 16) argue that many qualitative and quantitative researchers agree that, for example, ‘observation is not a perfect and direct window into “reality”’, recognise, ‘that we only obtain probabilistic evidence, not final proof in empirical research’, and accept ‘the social nature of the research enterprise’ (see also Mahoney and Goertz, 2006; Tashakkori and Teddlie, 1998). These ‘postpositivist’ methodological commitments best characterise our perspective on research design; by recognising the strengths, weaknesses, and political implications of both quantitative and qualitative approaches, we believe that it is both possible and desirable in view of our research aims to work pragmatically with a mixed methods approach. Before we further outline our research design in respect of the focus group and survey work, it is worth reviewing briefly what the strengths and weaknesses of qualitative and quantitative research methods are. The strengths of qualitative research including focus group work are commonly identified as, inter alia, that it allows its subjects to construct and reflect upon their own understandings and usages of concepts, its ability to examine cases in depth and to handle complexity in the social world, and that it is sensitive both to the relationship between investigator and participant and also to the extent to which knowledge is mutually constructed by the interaction between them. However, qualitative research also tends to lack generalisability due to small and unrepresentative samples, it is difficult and/or undesirable to strictly test theories and hypotheses, and the interpretation of qualitative data requires considerable sensitivity to the biases of researchers (Johnson and Onwuegbuzie, 2004). Quantitative methods properly employed should permit generalisation, the testing of theories and hypotheses, interpretation should be less affected by the biases of the researcher (though

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

47

not eliminated entirely, of course) and is more easily replicated. However, validity may be compromised by the researcher imposing his or her own understandings on participants, and the focus on hypothesis testing can lead to ‘confirmation bias’. Moreover, quantitative research ‘addresses much less successfully [than qualitative research] . . . concept formation and fine-grained description’ (Collier et al., 2004: 49). Johnson and Onwuegbuzie (2004) describe the combining of qualitative and quantitative methods in a single research design as the ‘third research paradigm’,4 while Tashakkori and Teddlie (1998: 5) call it ‘paradigm relativism’. The idea is the same: with the methods’ strengths and weaknesses in mind, the researcher can design a project in which the strengths of one avoid the pitfalls of the other because ‘pragmatists consider the research question to be more important than either the method they use or the worldview that is supposed to underlie the method’ (Tashakkori and Teddie, 1998: 21–3; see also Creswell, 2003: 11–12; Friedrichs and Kratochwil, 2009; Tarrow, 2004). For example, in quantitative research, ‘the researcher’s categories that are used may not reflect local constituencies’ understandings’ (Johnson and Onwuegbuzie, 2004: 19). In our research design, we could have designed a survey without the prior qualitative wave of mini-focus groups or ‘triads’ described below, and we might then have listed security threats as those identified in the NSS plus one or two additions that occurred to us or that have appeared in other surveys, such as immigration. However, the triads, which began by asking participants in an unprompted and open-ended manner to talk about their own understandings, experiences, and measurements of ‘threat’, ‘security’, and ‘security threat’, revealed a number of pressing security threats to individuals, such as online fraud and economic hardship, that we may not have thought of in these terms.5 Importantly, we are not saying that by triangulating methods it is somehow possible to cancel out the weaknesses of the respective approaches we adopt. As we will go on to discuss, there are inevitable pitfalls and we are well aware of the limitations of the study. One of these is the time-bounded nature of the data collection period – while this was staggered in three stages throughout 2012, inevitably we lack a more longitudinal perspective and the comparative dimension this would bring. The effects of the economic crisis and the London Olympics loom large over the content of our focus group discussions in particular – and the absence of any publicised terrorist incident in Britain since the attempted attack on the Giraffe Café in Exeter on 22 May 2008 is of potential significance. That the study was confined geographically to Britain is obviously another important consideration when reflecting on the broader implications of our findings for the everyday politics of threat perception in a transnational context. A number of additional choices we made – concerning the composition of the focus groups, for example – also have an inescapable bearing on the data we co-produced. With these and other limitations in mind, the following

48

Everyday security threats

discussion intersperses a series of reflections on the possible drawbacks of some of the decisions we made in the context of an outline of the research design as a whole.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

A mixed methods research design Johnson and Onwuegbuzie (2004) present a qualitative-quantitative mixed methods 2x2 taxonomy in which the dimensions are ‘equal versus dominant status’ given to qualitative or quantitative approaches – quantitative research is often considered to be ‘dominant’ because qualitative research is used to develop survey questions, providing some clues about language and understanding, but it is the survey evidence that is used to draw conclusions about relationships – and whether or not the two approaches are used ‘concurrently or sequentially’. According to these categories, our research design belongs in the quadrant in which the two methods are given equal status and are sequential. While qualitative methods were used in the first stage of our research, and indeed informed the development of our survey questions, this was not their sole purpose. Indeed, as Chapter 5 demonstrates, a critical approach to focus group research can play an important role in its own right in terms of both better understanding vernacular theories of everyday (in)securities and, beyond a representative function, challenging official accounts and repoliticising the technocratic framings of national security policy by juxtaposing conventional with radically alternative meanings of the same concept such as ‘security’. Moreover, we employed a further stage of qualitative research to follow the survey, in order both to provide additional understanding of the survey findings and also to explore further issues and findings that had emerged from the combination of the survey and the first wave of qualitative research in a selfreflexive manoeuvre. Qualitative and quantitative approaches in mixed methods research may not, of course, point in the same direction in terms of their implications for the attitudes that people hold, the salience of those attitudes, or the nature of relationships between different variables (and thus for hypotheses). Jones and Sumner (2009: 36; see also Greene et al., 1989) point to four ways in which qualitative and quantitative data analysis may combine: • elaboration or expansion (‘the use of one type of data analysis adds to the understanding being gained by another’); • initiation (‘the use of a first method sparks new hypotheses or research questions that can be pursued using a different method’); • complementarity (‘together the data analyses from the two methods are juxtaposed and generate complementary insights that together create a bigger picture’);



The 2012 study ‘Public Perceptions of Threat in Britain’

49

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

• contradictions (‘simply juxtapose the contradictions for others to explore in further research’). A good example of complementarity is Williams and Stahl’s (2008) study of police traffic stops and searches in Kentucky. Their quantitative analysis showed that African-Americans and Hispanics were more likely to be stopped and searched than other groups. But only by conducting focus groups with police officers themselves were Williams and Stahl able to gain a deeper understanding of the meaning of this finding, concluding that the cause of the pattern of stops was not as straightforward as racial profiling; rather, many of the major problems in the community, such as driving under the influence and the use of crack cocaine, were both thought by the police to be best dealt with by traffic stops and were also most prevalent in African-American (crack) or Hispanic (driving under the influence) areas, i.e., the police were profiling the problem.6 The purposes of our research design relate primarily to elaboration and complementarity. This is not to say there is never any tension between the conclusions we draw from the qualitative and quantitative research nor that we consider this to be inherently problematic; our primary aim is not to illuminate contradictions for others to explore. Instead, where we see contradictions, we try to explain them in the context of our research design. We are neither exclusively seeking the qualitative criterion of ‘trustworthiness’, nor the quantitative criterion of ‘validity’, but what Tashakkori and Teddlie (1998) describe as the appropriate criterion for mixed methods research, namely ‘inference quality’. Our general approach in this project was to engage in discussions with people in order to elicit their perceptions and experiences of contemporary security threats in narrative form, as opposed to alternative ethnographic approaches such as observing the way people behave in security situations, for example at airports, or by conducting experiments in which the researcher attempts to manipulate perceptions of security and observes changes or differences in attitudes and behaviour. While it is true that, as Jarvis and Lister (2015: 110) note when reflecting on their own usage of focus groups, ethnographic and experimental approaches allow for greater appreciation of covert or unintentional actions, neither of these approaches would have allowed us to answer the range of questions we pose. The equal status, sequential, qualitative, and quantitative research design we adopted combined mini-focus groups of three people (known as ‘triads’) and a large sample survey. Figure 2.1 depicts the stages of the research. We now outline and critically reflect upon each stage in greater detail. Focus groups The typical choice for qualitative research that seeks to hear from people is either in-depth one-on-one interviews or focus groups with six to eight participants.

Everyday security threats

50 Stage 1: Triads 4 April–2 May 2012

Stage 2: Survey 6–15 June 2012

Triad 1: London, male, 18–25

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Triad 2: London, female, older parents

ICM internet panel 2004 British adults over 18, including 251 Muslims

Stage 3: Triads 8 August–3 September 2012

Triad 11: Glasgow, mixed gender, older Triad 12: Glasgow, mixed gender, younger

Triad 3: Leicester, older male, Muslim

Triad 13: Cardiff, mixed gender, older

Triad 4: Leicester, female, Muslim

Triad 14: Cardiff, mixed gender, younger

Triad 5: Leicester, female, 18–25

Triad 15: Oldham, younger male, Muslim

Triad 6: Leicester, retired, mixed gender

Triad 16: Oldham, female, young parents

Triad 7 Leicester, female, younger, Hindu

Triad 17: Oldham, male, older parents

Triad 8: Bristol, female, 18–25

Triad 18: Oldham, female, retired

Triad 9: Bristol, male, young parents

Triad 19: London, female, older, Muslim

Triad 10: Bristol, male, retired

Triad 20: London, male, older, Hindu/Sikh

Figure 2.1  Stages of the research

The advantage of the former approach is that it permits exploration of specific issues in great depth and an understanding of the views and experiences of participants in the context of individual biographies (Andrews et al., 2013; Riessman, 2008). There is the potential for the researcher to gain considerable understanding of an inevitably restricted number of individuals from which to draw inferences about ‘canonical narratives’ in a given social and cultural context (Phoenix, 2013: 73). A limitation, however, is that in-depth one-on-one interviews do not allow observation of peer dynamics, shared experience and wider social currents, or the way in which dominant cultural paradigms both enable and constrain what can and cannot be said (Barbour and Kitzinger, 2009; Phoenix, 2013; Rodriguez et al., 2011). Thus, if the researcher believes that intersubjective interaction is critical to individuals’ understanding of an issue, and is interested in the language and knowledge used to narrate that understanding, then focus groups are optimal. We opted for a hybrid approach that provided us with elements of individual depth and group interactions. This was achieved by limiting the size of our focus groups to three participants (hence the term ‘triad’). On the one hand, it might be argued that this smaller number of participants reduced the range of possible group interaction that is a hallmark of focus group research. On the other hand, however, we decided on the triad approach in view of the likely sensitivity of the topics under discussion – particularly in the context of our minority groups. Also, an added advantage of



The 2012 study ‘Public Perceptions of Threat in Britain’

51

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

triads is the ability of the moderator to better steer the discussion if it drifts radically off topic and/or if a dominant voice emerges within the group. Before outlining how our groups were organised and run in greater detail it is first instructive to consider the various contending perspectives within focus group research in order to situate our own critical and interpretive approach. Critical focus group research

Focus group research (FGR) has a controversial history in the context of social science. It was first developed in the 1940s by the US military as a method of assessing the impact of wartime propaganda on serving military personnel (Bloor et al., 2001). In the 1950s FGR became the preferred method for businesses and advertising agencies to test attitudes towards new products among consumers (Rodriguez et al., 2011). The use of FGR by social scientists since the 1970s and 1980s has become increasingly popular, but it has not gone without criticism due to these earlier associations and, more recently in the British context, connections with the ‘sofa-style’ government of New Labour and former Prime Minister Tony Blair (Barbour and Kitzinger, 2009). Traditionally, Coule (2013: 148) argues, there has been a tendency for social scientists who draw on FGR to focus on the practicalities of method ‘with little consideration for the variety of assumptions regarding the nature of knowledge and its production’. Beyond this focus on ‘pros’ and ‘cons’, she argues that FGR has been dominated by two perspectives, which, though ostensibly different, are united by their positivist methodology: social psychology, which privileges direct questioning and offers few opportunities for group interaction, as such; and clinical psychotherapeutic approaches, which prioritise group tasks and discussions to access deep-seated thoughts. Both work within an epistemological framework dominated by deductive theorising and hypothesis testing whereby participants are typically viewed as ‘passive subjects, who hold opinions and preferences that are considered to be objective “facts”, best expressed in group situations under the “control” of the moderator’ (Coule, 2013: 151). The problem with these approaches, to refer back to our earlier discussion, is that they are blind to the intersubjective production of meaning within the group setting. By contrast, post-positivist interpretive research agendas – with which we align our own approach – have opened up a range of critical approaches to FGR which typically explore the intersubjective, culturally embedded, and politically contested social construction of meaning (Barbour and Kitzinger, 2009; Coule, 2013; Rodriguez et al., 2011). Critical FGR does not aim to produce a consensus based around fixed attitudes, but to create a reflexive forum for investigating ‘how knowledge, ideas, story-telling, self-presentation and linguistic exchanges operate within a given cultural context’ (Barbour and Kitzinger, 2009: 5). Particular analytical attention is paid to: the use of ‘contrastive others’ in

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

52

Everyday security threats

identity-making claims (Wilkinson, 2009: 67); the articulation of views and experiences in relation to ‘wider societal debates’ (Waterton and Wynne, 2009: 130); and the ‘categories of experience’ that people use to discuss ‘feelings, responses . . . and world-views’ (Myers and Macnaughten, 2009: 174). The assumption in critical FGR work is not that the subject and his/her views preexist the situation in which the discussion takes place, but that it is via the interaction with others that this identity and knowledge are constituted. Because of these dynamics, critical FGR does not make any scientific claims to replicability or generalisability. Rather, the purpose is to analyse on a small scale the ‘way in which people actively establish and re-establish the meanings they use to organize their experiences’ (Coule, 2013: 152). This permits a form of discourse analysis that is sceptical of truth claims and is therefore more interested in how such claims are made, by which means, and with what effect (and for whom). The existence of multiple realities and different ways of narrating and interpreting them is not a problem from this perspective but is to be expected and is indeed the driver of research: such negotiations provide both a fascinating window into individual narratives of the self and broader socio-cultural currents of shared meaning-making and understanding at any given time. Aside from the co-production of talk around a specific issue and the narrative and thematic analysis that follows, it is also acknowledged within the literature that critical FGR performs a potentially important political role in that it constitutes a form of participatory research. In this regard, demographic diversity within and/or between groups is intended not for the purposes of sampling or representativeness of particular communities, but rather to actively include other­wise marginalised voices in the study of potentially sensitive social and political issues. For example, feminist and black rights campaigners have long drawn upon critical FGR methods in order to produce oral testimonies and share stories of oppression, which are sometimes claimed by participants to have empowering and even emancipatory effects (Rodriguez et al., 2011). In participating in the research as co-producers of knowledge, individuals can share ‘personal troubles’ as ‘public issues’ (Barbour and Kitzinger, 2009: 19) and ‘lay’ knowledge treated as a form of expertise, which, despite being ‘contingent and relational’, can be fed into academic and policy-making debates (CunninghamBurley et al., 2009: 194–6). Our triads

Informed by critical approaches to FGR, we convened twenty ninety-minute triads involving a total of sixty participants (thirty-one female and twenty-nine male) to discuss contemporary security threats in Britain. This meant that we could probe individuals in more detail than in standard-sized focus groups, but also allowed us to observe peer-to-peer exchanges both in the context of general discussion and group tasks such as categorising specific threats.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

53

We conducted an initial ten triads, mostly in April 2012, followed by a second wave of ten triads in August and September of 2012 that followed the completion and preliminary analysis of the survey data and was designed to give us additional understanding of the first two stages of research. Figure 2.1 shows the two stages in which we conducted the triads and also the profiles of each group. Discussions were held in a mixture of venues, including hotels, community centres, and residential houses. All sessions were recorded and subsequently transcribed verbatim in order to retain the distinctive voices of all our participants. The triads varied on criteria that we anticipated could affect perceptions of security threats and the kinds of discussion we would observe: gender, region, age and life-stage, and religion. They took place across six British cities (Bristol, Cardiff, Glasgow, Leicester, London, and Oldham), among young and retired men and women with and without young or mature children, primarily in single-sex but sometimes in mixed-sex groups, and with Muslims, Hindus, and Sikhs, as well as in groups for which religion was not a criterion. The rationale for constituting the groups in this way was that some issues, such as fear of knife crime or racism, were sensitive and would be more easily broached in groups of similar individuals who may share those perceptions – e.g., a young person’s fear of knife crime or anxieties about not being able to get a job – than in groups whose other members did not share key characteristics. On the one hand, the matrix of focus groups depicted in Figure 2.1 demonstrates that in certain cities we did not speak to groups with particular personal characteristics (e.g., no young white women in London). This means that the criteria that may affect perceptions and experiences of security threats were not evenly distributed across all groups. On the other hand, we would stress that these issues are only of scientific significance if the ultimate aim of the focus group work was to arrive at generalisations about what certain types of participants in particular locations think: to repeat, it was not. Rather, we sought to maximise overall diversity of participants and locations – against parameters set by levels of funding available – in order to involve as many different and otherwise excluded voices as possible. We felt able to make these bold choices against the backdrop of the work of the survey, which sought in a more systematic manner to explore the effects of different criteria affecting threat perception. In this way, while alternative decisions about the composition and structure of the focus groups may have resulted in different conversations in terms of content and style, the mixed methods design ultimately allowed for such complementarity. The triads were recruited and moderated by experts from the social research company TNS-BMRB. While we could have conducted them ourselves, that would have presented the twin dangers that participants would respond differently to academics than to professional moderators and also that as academics directly involved in the research we could be prone to steer participants in directions that confirmed our own ideas. On the other hand, the main pitfall

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

54

Everyday security threats

of not moderating groups ourselves meant that there were fewer opportunities to explore emerging themes and issues as they developed over the course of the focus group work in a more inductive fashion. It is also the case that our ability to comment reflexively on our own experience of co-producing knowledge with participants is rather more limited than some studies associated with the autoethnographic turn in IR and Security Studies would demand (see, e.g., Kurowska and Tallis, 2013). We did, however, work very closely with TNSBMRB in developing the moderator’s guides for the two stages of qualitative research and acted as observer-participants in four pilot-stage sessions. Because of the team-based nature of the research, the issue of ‘researcher effect’ therefore has different implications for our study as a whole (Jarvis and Lister, 2013: 163). In engaging with this issue we here reflect on the nature of the discussion guide. Our primary overarching principle was to maximise free-flowing discussion of participants’ views, experiences, and stories of threat and (in)security using their own language. Discussions began by exploring both the range of issues that individuals and groups felt were the most pressing ‘security threats’ to them in their everyday lives and how they approached these issues. We asked open-ended questions such as: What comes to mind when you think of ‘security threat’? Which issues are security concerns for you and your family and why? Why are certain issues more important than others? Tell us a story about how a security threat has affected you? To begin with we imposed no particular conceptualisation of ‘security’ or ‘threat’ as we wanted to find out how our participants thought about these issues in an unprompted way. We then juxtaposed their answers with the list of priority risks contained in the NSS, which was intended to stimulate critical reflection on the issues they had identified and the rationale for that selection, whether they agreed with the government’s prioritisation of risks and its underpinning methodology, and the extent to which the two converged or diverged. Towards the end of the sessions we focused in greater depth on participants’ awareness of and views about government efforts to enlist the support of citizens in the risk management cycle – particularly with regard to vigilance and surveillance – and the impact on their daily lives and society as a whole. During the first stage of research the moderator’s guide covered the following topics and was organised as follows: 1 Understandings and perceptions of ‘security threat’: adjectives, pictures, and feelings that come to mind when prompted with the words ‘threat’, ‘security’, and ‘security threat’; security threats in participants’ lives, other people’s lives, in the past, in the future. 2 Hierarchy and influence of ‘security threats’: spontaneous ‘mapping’ of sort cards of different threats, including those discussed in response to (1) above

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

55

and those mentioned in the NSS; prompted mapping of issues into high and low likelihood and discussion of rationale; and which issues are threats to self, community, country, world. 3 Impact of ‘security threats’: how the beliefs about security threats influence behaviour; which threats have most and least influence and in what ways people respond to the risks; stimulus materials with pen portraits of possible behaviour modifications, e.g., doubts about flying from the UK to the US because of a change in the government’s assessment of the terrorist threat, followed by discussion; awareness of and use/potential use of anti-terrorist hotline and other forms of participation in the risk management cycle. 4 Strategies to reduce security threats: level of security threat in today’s world compared to the past; responsibility to reduce threat and extent to which it falls on individuals and/or governments; awareness of any government-led campaign or advertising, now or in the past, asking the public to play a role; response to audio-visual materials such as ‘Protect and survive’ and ‘If you suspect it, report it’ TV/radio campaigns; types of communication that would encourage changes in behaviour and why; government’s role and communication in tackling security threats; response to stimulus of NSS Tier One, Tier Two, and Tier Three threats; citizen-detectives’ case study: can/ should the public be involved in assessing levels of security threat? The second stage of research using triads, which followed the survey (see below), retained many of these same themes, but incorporated some substantive changes and changes in emphasis in order to gain a deeper understanding of perceptions. In part (2) of the group work we no longer asked participants to sort security threats into different categories. Rather, we focused more on discussion of how different issues are seen as threatening, to whom, and why, and particularly on participants’ use of contrastive others to make identity-based claims – for example, by probing the meaning of references to ‘we’ and ‘they’. We also sought to concentrate far more on individuals’ own experiences of and autobiographical ‘stories’ about security threats in the context of their everyday lives, e.g., an issue or event that they thought of as a security threat that had affected them. In addition, we sought to focus discussion on contemporary threats, still allowing participants to talk about past and future issues spontaneously, but not prompting such discussion. This was both because we had found that such changes in references to time prompted little additional discussion in the first wave of triads and also because the survey had indicated very little difference in perceptions of current and future security threats. An alternative, perhaps simpler, mixed method qualitative-quantitative approach might have been to interview in depth a limited number of survey respondents. We eschewed this approach for two reasons. First, while post-survey in-depth interviews may enhance understanding of survey responses, they are

56

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

limited in their ability to generate additional knowledge, particularly given that respondents have already been subjected to the researcher’s own questions – rather than questions generated from other participants in the research – in the survey. Second, a random national sample almost precludes the kind of groupbased research that was an integral part of our design, and for this reason it is certainly not an efficient or productive way to assess such dynamics. The survey While the triads provided us with great depth of material about perceptions and experiences of security threats, we also sought to enrich our data with a large sample survey probing some of the same and also additional issues. For example, the survey allowed us to estimate the relationships between perceptions of security threats and individual characteristics such as age and sex, as well as between particular attitudes and security threats. In this, the second stage of our research design depicted in Figure 2.1, we conducted a twenty-five-minute internet survey of 2,004 respondents from 6 to 15 June 2012. It included a booster sample of 251 Muslims, a unique group that is both threatened and often seen as threatening by other groups in the context of contemporary British society (Gillespie and O’Loughlin, 2009a). The survey was administered by ICM and the sample was drawn from its online panel. Internet surveys from online panels remain an area of contention, with some prominent researchers sanguine about them (Sanders et al., 2007), while others argue that they produce less accurate surveys than random probability samples (Yeager et al., 2011). Table 2.1 shows that, other than ethnicity (due to the booster sample), survey respondents were representative of the British population on dimensions of gender, age, region and the party for which they voted in the 2010 general election.7 We asked about perceptions of a range of twenty-two issues as global, national, community, or personal security threats: Terrorism Knife crime Burglary Crimes against women Racial or religious hate crime Weak border control Nuclear weapons programmes in Iran, North Korea, and other hostile states A health pandemic (e.g., avian flu) Environmental issues (e.g., global warming or the greenhouse effect, pollution) Online fraud or identity theft UK foreign policy



The 2012 study ‘Public Perceptions of Threat in Britain’

57

Table 2.1  Survey sample profile

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Population (%)

Gender Male 49 Female 51 Age 18–24 12 25–34 16 35-44 19 45–54 17 55–64 15 21 65+ Race/ethnicity White 90 Asian 4 Other 6 Region London 13 South and East (outside London) 31 Midlands 17 North 25 Scotland 9 Wales 5 Vote in 2010 general election (among those voting) Conservative 36 Labour 29 Liberal Democrat 23 Did not vote 35

Survey (%) n = 2004

48 52 13 18 17 16 16 20 82 11 7 16 29 17 26 8 4 34 30 25 24

Note:  Figures for the population are from the National Readership Survey (www.nrs.co.uk/ interview.html), with the exception of vote in the 2010 election. The figures in this category for the survey exclude refusals and don’t knows for this question (3 per cent of the sample).

Religious extremism Immigration Resource scarcity (e.g., dependence on oil, water shortages) Economic depression/financial crisis/unemployment Increasing power of Russia and China Attacks on cyber-space and cyber-crime An international military crisis between states Severe disruption of critical infrastructure (e.g., information, energy resources such as oil or gas, food)

58

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

The far right (e.g., English Defence League) A nuclear accident like Fukushima Islamophobia Some were drawn directly from the Tier One, Two and Three threats identified in the UK government’s 2010 NSS, including terrorism, hostile attacks upon UK cyber-space, and disruption to oil or gas supplies to the UK, while others emerged as recurrent security threats identified by participants in the first wave of triads, including the economy, immigration, the far right, burglary and online fraud. The survey questions did not define ‘security threat’ for respondents, nor did the list of twenty-two issues privilege any particular type or level of threat; rather, the survey covered potential threats that were domestic and international, economic, political, technological, health-related, and group-based. Indeed, while we would expect identification of some of these threats to be confined to the national or international level, such as the increasing power of Russia and China, others in the list are likely to be seen only as subnational threats, including knife crime and burglary, while still others could span multiple levels, including the economic crisis, immigration, online fraud, and religious extremism.8 These also encompass the ‘new issues’ of crime, immigration, asylum seekers, race, and terrorism, identified by British Election Studies (BES) investigators as part of the ‘changing issue agenda’ after 2001 (Clarke et al., 2009; Whiteley et al., 2013). Our examination of threats runs from the broadest collective, the global, through national to community-level threats, which are still collective but at a much more intimate and personal level than the state or the globe. Thus we see community-level threats as closer to what is usually meant by personal rather than by sociotropic threat. It is possible that people think of their community as the nation or the world, but we consider it unlikely here. The survey asked about the ‘community in which you live’. Transcript materials generated by the focus groups and findings from the survey indicate very strongly that ‘community’ has subnational, indeed local, connotations for most people. We operationalise perceptions of the breadth of threats as the number of issues that people see as global, national, community, or personal threats, while perceptions of specific threats are simply whether or not an issue was identified as a threat at the global, national, community, or personal level. We did not ask about direct exposure to these threats, the intensity with which they were felt, nor about their content (e.g., whether a threat from the far right stems from it being viewed as dangerous or powerful):9 threats to social cohesion (Feldman, 2003) or from particular groups will be captured by our measure if they rise to the level of security threats. The survey also asked about other issues such as spending and taxation preferences in various policy areas, perceptions of the efficacy of actions and policies designed to mitigate terrorist threats, awareness of the NSS and other

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

59

government actions or programmes designed to address security threats, as well as questions to assess the antecedents of perceptions of security threats such as age, sex, education, and media habits, and dispositions such as mortality salience and authoritarianism, and the effects of perceptions of security threats on attitudes towards immigrants and other minorities. Some questions were original to the survey, while others, such as attitudes towards immigrants, used or adapted question wordings employed in the European Social Survey, World Values Survey, American National Election Study (ANES), British Social Attitudes Survey, and BES. We focus on the antecedents of threat perceptions in Chapter 3. Our measurement of the key antecedents of threat perceptions, based on our review of previous research discussed in Chapter 1, is as follows: • Mortality salience: three agree–disagree questions about thoughts of death (details are in the Appendix).10 • Authoritarianism: we assess authoritarianism through the child rearing values questions used in the ANES that have become standard by virtue of not conflating authoritarianism with conservatism or prejudice (e.g., Feldman, 2003; Hetherington and Weiler, 2009; Stenner, 2005). • Media habits: we gauge news media effects by examining the influence of the time respondents claimed to watch news and current affairs programmes on television as a proportion of their total time spent watching television. • Education: we operationalise education with a dummy variable for respondents with a higher education degree or above. We expect that where level of education has an impact it will be on the identification of threats at the global or national level rather than at lower levels, where perceptions of threats are likely to be less dependent on probabilistic knowledge and more on day-today experiences in the neighbourhood in which a person lives.11 In Chapter 4, we turn our attention to the effects of threat perceptions. We examine a host of possible effects, or consequences, of perceptions of threats based on our desire to obtain a broad understanding of security threats. • Consequences: we cover four areas that have been the focus of previous research: punitiveness and intolerance; attitudes towards minorities and immigrants; spending and taxation preferences; and the desired attributes of leaders and voting behaviour. We describe the specific measures in detail in Chapter 4 when we discuss the analysis. Figure 2.2 provides an illustrated summary of the model we employ with the survey data and the chapters that contain each stage of the analysis. Cross-sectional surveys such as ours have well-known limitations pertaining to the reliability of causal inferences (see Malholtra and Popp, 2012). A danger

Everyday security threats

60 Chapter 3

Chapter 4

Exogenous variables

Perceptions of threats

Outcome: attitudes towards outgroups, spending, certainty of voting

Mortality salience Authoritarianism

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Global Media exposure National

Education Sex

Community

Attitudes towards outgroups Attitudes towards spending Certainty of voting

Age Personal

Muslim Religiosity White Awareness of security strategies

Chapter 5

Figure 2.2  The models for estimating the origins and effects of perceptions of threats Note:  Figure 2.2 presents the basic approach to modelling without showing arrows for each individual relationship estimated. In the models, education is specified as an influence on global and national threat only and white on community and personal threat only.

to our analysis is that respondents may rationalise from policy preferences to perceptions of security threats, for example, meaning that the causal relationships are the reverse of those we posit. It is not clear, however, why respondents would be motivated to do this for twenty-two security threats, nor how such rationalisations would affect stated perceptions of security threats at different levels. In addition, the fact that we asked about security threats prior to questions about policy preferences should limit such rationalisations. Conclusion The complexity of researching everyday perceptions and experiences of contemporary security threats in Britain demands a methodological step-change in how

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

61

IR and Security Studies – in both traditional and critical guises – approach empirical research. We have argued that a mixed methods approach, drawing together the insights of in-depth focus group work with large-N survey data, is most apposite in view of this complexity. Puritanical or tyrannical attempts at segregating qualitative and quantitative methods are in our opinion at best misplaced and at worst unhelpful in the study of social and political phenomena. Undergirded by methodological principles associated with post-positivist epistemology, in this chapter we have outlined the overall approach adopted in our 2012 study ‘Public Perceptions of Threat in Britain’. Subsequent chapters offer a detailed discussion of our findings and the implications for future research that brings together the insights of the ‘everyday’ and ‘vernacular’ turns in IR and Security Studies on the one hand, and on threat perception and public opinion in Political Science on the other. We conclude the present chapter with several reflections on how the post-positivist methodology and mixed methods approach outlined thus far shape our approach to analysis of the focus group transcripts and survey data. With regard to the qualitative work, in the discussions that follow we treat the transcripts not as ‘background in which themes arise’, but rather as ‘interactive pieces of talk’ and as such a ‘site of constant negotiation’ (Waterton and Wynne, 2009: 133). Given our critical approach to FGR, we focused our coding and discourse analysis of the data on the following key themes: 1) the relationship between ‘official’ and ‘non-elite’ representations and measurements of contemporary security threats and the extent to which certain key themes were repeated across groups; 2) the specific language, range of emotions, and knowledge mobilised by individuals and groups when discussing such threats; 3) the narratives of threat and (in)security drawn upon in order to structure accounts of individuals’ experiences of wider social relations; 4) the logics of identity and difference, self and other, and other binary oppositions used to make claims about participants’ own subject positions vis-à-vis others. Following Michael J. Shapiro’s (2013) ‘post-hermeneutic’ approach, our method of analysis and presentation of findings proceeds by way of juxtaposition. Instead of treating data obtained as the fixed and otherwise hidden views of subjects, we juxtapose diverse data in order to mobilise ‘subjects with conflicting perceptual orientations’ (Shapiro, 2013: 29). By juxtaposing the views, experiences, and stories of diverse publics, we do not seek ultimately to arrive at a representation of particular groups’ worldviews, but to demonstrate the multiplicity of perspectives in a way that, as we shall argue in Chapter 5, challenges dominant governmental logics relating to ‘national security’. In this way, like Shapiro, we seek to disrupt idealised images of the ‘citizen-subject familiar in nation state-centric discourses’ and explore how different experiences of (in)security are productive of different forms of political subjectivity (Shapiro, 2013: 116). But whereas Shapiro chooses novels and art to mobilise such disruptions, we argue that everyday talk offers a

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

62

Everyday security threats

rich and otherwise untapped reservoir of knowledge that can have a similar effect. With the survey data, while our approach to analysis is conventional, it is also informed and influenced at all times by evidence from the triads. We proceed from initial assessment of how people think of security threats to analysis of the breadth of security threats identified at different levels – global, national, community, and personal – in Chapter 3. We ask whether individuals tend to see the same threats regardless of level, whether there is more compartmentalisation, or whether the picture is still more nuanced than this. We also examine whether survey respondents who identify more threats rank those threats similarly, i.e., see the same threat(s) as most pressing, to respondents who identify fewer threats and, in bivariate analysis, whether identifying more threats at one level is strongly correlated with identifying more threats at another. We then move from this fairly descriptive approach to the data to more complex statistical models that allow us to isolate the relationship between, for example, perceptions of a threat such as terrorism and level of education. In Chapter 4, the same statistical models allow us to examine relationships between perceptions of threats and a variety of attitudes and behaviours, such as attitudes towards minorities and likelihood to vote in the next general election (2015 when the survey was in the field) and thus to gauge the effects of perceptions of security threats and the extent to which these are contingent on the issue in question. Finally, analysis of the survey data in Chapter 5 is juxtaposed with the qualitative evidence regarding everyday talk about government approaches to security threats to assess awareness and effectiveness of the NSS and other government policies or programmes designed to mitigate security threats. We ask both about depth of awareness of government strategy and policies and whether they are associated with heightened or reduced perceptions of current threats now and in the future.

Notes 1 For example, Jarvis and Lister (2015: 6) distinguish their qualitative approach to focus group research against quantitative studies of public opinion. 2 Jones and Sumner (2009) point out that the terms ‘quantitative’ and ‘qualitative’ can refer to methodology, methods of data collection, the raw data, and techniques of analysis. Our use of the term is primarily intended to encompass methods of data collection. 3 Tashakkori and Teddlie (1998: 8) use the term ‘post-positivist’ paradigm to describe quantitative research that no longer makes the methodological assumptions or adopts the same approaches as the positivist paradigm. For a discussion of the term in the context of IR see Lapid (1989). 4 Albeit they are referring to educational research. 5 To an extent, pre-testing a survey may allow the researcher to accomplish some of what we discuss but the triads provided us with a wealth of additional data and conceptual detail that pre-testing of a survey could not.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The 2012 study ‘Public Perceptions of Threat in Britain’

63

6 Although this is still an ineffective method, according to Williams and Stahl (2008). 7 As tends to be the case in surveys, our sample contained a lower proportion of non-voters than in the population but this proportion compares favourably to the face-to-face BES, which contained fewer non-voters (22 per cent). 8 An alternative approach would have been open-ended questions about perceptions of threats but we preferred a list format for three reasons: 1) We were interested in the combination of threats that both may occur to respondents spontaneously and also be regarded as threats when prompted to think about them, i.e., that may be further back in consciousness; 2) Some of the issues listed, such as immigration and religious extremism, were sensitive and respondents may be less willing to identify them themselves than when looking at a pre-defined set of issues, and 3) We felt there was a greater danger of conflating the ability to articulate threats with genuine perceptions of the breadth of security threats with an open-ended question. 9 Asking about the intensity of each threat would have necessitated up to eighty-eight (22 x 4) additional questions, leading to respondent fatigue and questionable data quality. Our priority in the survey was to capture perceptions of the breadth of threats with a large range of potential security threats: in our qualitative work we allowed participants to compare and contrast issues, to rank them in terms of the immediacy of the threat posed to them in everyday life, and to discuss in depth an issue that they felt particularly threatened by. We are also deliberately ignoring possible relationships between threats, e.g., if perceptions of threat from the economic crisis are related to heightened perceptions of threat from crime, which are beyond the scope of this chapter. For our purposes, these are two distinct threats. 10 Half the survey sample was asked these questions at the beginning of the survey and the other half at the end, in order to guard against the possibility that the content of the survey would raise mortality salience or that asking about mortality at the outset of the survey would affect the answers to other questions. Neither appears to have been the case (e.g., a chi-square test of the mortality salience index by question order is statistically insignificant). 11 Similarly, we expect that white ethnicity is unlikely to affect perceptions of threats at the global or national level but could at the community or personal levels, e.g., as a result of ‘white flight’ to apparently safer neighbourhoods. Akaike and Bayesian Information Criteria confirm that models without these restrictions have an inferior fit to the data. The operationalisation of white ethnicity and the other control variables of sex, age, Muslim religion, and religiosity are described in the Appendix.

3

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

The scope of security threats and their causes

Introduction In this chapter we endeavour to redress some of the lacunae in the existing literature identified in Chapter 2, by going beyond specific ‘threats of the moment’ and the personality traits that are associated with heightened perceptions of them. Instead, drawing on the findings of both our focus group work and the survey, we aim to investigate perceptions of a range of security threats, the levels at which they are identified – transcending the national–personal dichotomy – and their causes, incorporating psychological attributes but also other considerations. We begin the analysis by summarising how participants in our group discussions define and understand the key concepts of ‘security’ and ‘threat’: the referent object they invoke as that which is to be secured; the ways in which particular issues are perceived, measured, and classified as threatening; and the key themes and concrete issues that arise in response to open-ended questioning about what security threat means to them in their everyday lives. Findings from the first stage of the research informed the design of the large sample survey and we identify several areas of convergence both internally to our own programme of research and with extant studies of vernacular security. Crucially, our group discussions established a scale for the perception of security threats comprising four different ‘levels’ – personal, community, national, and global – which then structured survey questions to allow for a more fine-grained analysis than in hitherto existing research. The fact that this scale derived from our qualitative fieldwork and participants’ own method of explaining threat perception underscores both our derivative approach to the concept of security and also the project’s post-positivist commitment to self-reflexivity. Following an outline of how British citizens view security threats and which issues they consider to be particularly threatening according to our study, the discussion proceeds by focusing on three primary issues surrounding

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

65

perceptions of the breadth of security threats: how many different issues are regarded as security threats and whether seeing one type of threat is associated with seeing others; the extent to which perceptions of the breadth of threats vary as individuals move from ‘global’ to ‘personal’ levels of security threat perception; and the individual-level influences on perceptions of the breadth of threats. We then examine two additional issues in order to provide further understanding of the scope of security threats: the nature of variation in perceptions of threats at different levels, such as individuals who identify more personal than national threats, and their causes; and perceptions of the breadth of threats identified as Tier One by the British government in the 2010 NSS. In the final part of the chapter, we home in on perceptions of four specific threats, each with different characteristics, and explore their origins, allowing us to generalise further about the individual-level influences on perceptions of threats. How do citizens define and understand security threats? We began by asking participants in our focus group discussions: ‘When I say the words “security threats”, what adjectives, feelings, and pictures come to mind?’ Whereas traditionally the dominant referent object of security is the nation state in both academic IR and Security Studies and UK government policymaking communities, our groups tended not to think in these terms until prompted to do so (Triads 1, 2, 5, 6, 7, and 8): ‘I think security is not – it’s not national security . . . it’s very local’ (Aysha, Triad 7). Indeed, many of our findings about how citizens perceive security threats support the conclusions of past studies associated with the ‘vernacular turn’ – like Jarvis and Lister (2013) and Gillespie and O’Loughlin (2009a), for example – we also found that the most common referent object of security was the individual and his or her family. But while Jarvis and Lister (2013) draw six ‘images of security’ from their group discussion work – survival, belonging, hospitality, equality, freedom, and insecurity (with survival being the most common) – our findings were less nuanced. The dominant understanding of ‘security’ among our respondents referred overwhelmingly to personal safety – in other words, the absence of physical threats to the safety and well-being of the self (Triads 1, 5, 6, 7, 8, 11, 12, 13, and 14) – as typified by the following exchange in response to our opening question in Triad 12: Ben:

For me it has to be personal security, you and your family because that’s who you are and that who you are all the time so you have to secure that first before anything else. Angela: Yes, I’d agree. I think like safety, it’s about myself, my family, what could impact on me. I would probably think of global stuff second, third, fourth. First and foremost would be personal security.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

66

Everyday security threats

Furthermore, the categories drawn upon by Angela in the excerpt above reflect the most popular scale with which participants across all our groups conceptualised security threats: all discussions featured the invocation of either personal, community, national, and/or global levels (Triads 1, 5, 6, 7, 8, 11, 12, 13). Based on the findings of the first stage of triads – and those of previous research in vernacular security studies – we provided respondents to the survey with the following descriptions of ‘security’ and asked for up to three that best described their own definition: Security is: Feeling physically safe where you are (where you live, work, etc.) The absence of threats to humans in a general sense The absence of threats to me (i.e., individuals) The ability to access healthcare, education, and sufficient resources necessary for human well-being The freedom to do what you want within the confines of the law The nation’s freedom from foreign dictation The absence of threats to my community The threat, use, or control of military force Being treated the way others are treated It depends. Some people’s security depends on other people’s insecurity

Table 3.1 shows the proportion of the sample that chose each description first, and then the proportion that chose each description as one of three criteria that best encompassed the meaning of ‘security’ for them (respondents could choose fewer than three if they wished). Clearly, the most common definitions of security centred on the absence of threats, most especially to the individual, then to humans in general, and lastly to the community in which people live. Access to the resources needed for wellbeing and freedom to do what you want also emerge as important considerations but ones that are secondary to physical safety and the absence of threats. Far lower down in citizens’ definitions of security are equality in terms of ‘being treated the way others are treated’, which was the first mention of just 1 per cent of respondents, and concerns about the nation, such as freedom from foreign dictation or issues surrounding military force. Thus, the survey definitions of security echo much of the discussion in our focus groups and serve to emphasise one of the categories that Jarvis and Lister (2013) identify in particular. Jarvis and Lister describe this as ‘belonging’ because so many of their focus group respondents articulated it in terms of personal comfort in one’s surroundings and being accepted by others, but while physical safety and ‘the absence of threats to me’ loom large in our responses too, the survey evidence also suggests that feelings of security go beyond a personal sense of belonging to a more collective absence of threats.



The scope of security threats and their causes

67

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Table 3.1  Definitions of ‘security’ (numbers are %)

Feeling physically safe where you are (where you live, work, etc.) The absence of threats to humans in a general sense The absence of threats to me (i.e., individuals) The ability to access health care, education, and sufficient resources necessary for human well-being The freedom to do what you want within the confines of the law The absence of threats to my community The nation’s freedom from foreign dictation It depends. Some people’s security depends on other people’s insecurity Being treated the way others are treated The threat, use, or control of military force

First mention

All mentions

45

77

14 9 8

51 40 28

7

28

3 4 4

26 18 14

1 2

7 6

Source:  ICM Survey, ‘Security in an age of austerity’, 6–15 June 2012.

A further observation on how citizens define and understand security threats relates to the recurrence of several geographical and historical framings running throughout our group discussions. Participants regularly spoke of varying landscapes of (in)security and in particular a close connection was made repeatedly between the topic of our overall research and London. The capital city was typically framed – by those who live there as well as those living elsewhere – as an ‘exceptional place’ that is home to terrorists, illegal immigrants, and criminals: Whenever I travel up to London [. . .] I was over there yesterday and it is a whole different feeling when you get out [. . .] I wouldn’t want it to be taken as a comment, anything other than what it is, but when you are on a tube you are lucky to hear an English voice [. . .] And it’s nothing to do with colour or race or anything, but it’s everybody and everything up there. (Mike, Triad 9)

By contrast, other regions, particularly Scotland and the South-West of England, were associated with greater levels of security: I think up here we were maybe sort of alienated from the whole terrorist thing: it was always down in London until the thing at Glasgow airport a few years ago. (Jackie, Triad 11) I worry about my two sons down south, because there’s more bombs and threats down there than there is anywhere else. (Sandra, Triad 11) Well I think personally we are very fortunate in the South-West as we’ve got a bit of a buffer away from the main cut and thrust of it. (Dave, Triad 9)

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

68

Everyday security threats

Equally, various historical narratives structured some participants’ discussions about the extent to which they felt more or less secure in contemporary British society. In general, our groups said that they felt less secure now than in the past. The invocation of the events of 11 September 2001 as a turning point was common: ‘9/11 changed everything. It was sort of like where being safe stopped and feeling unsafe started [. . .] It was very sudden and I remember just afterwards feeling very uncomfortable’ (Aysha, Triad 7). Many participants tended to view today’s climate of fear and anxiety as being higher than when compared to the eras of the Cold War and Troubles in Northern Ireland: I mean, I think you always have to be a bit more aware of who is behind you when you are walking back at night and things like that nowadays which you might not have done thirty years ago. (Sarah, Triad 2) I’ve grown up in this country. I’ve been here since I was a child. And I remember when we were children, we felt safe staying out by ourselves in the park, anywhere, whereas now I would never send my children to a park any later than a certain time, because I just would not feel safe. (Aysha, Triad 7) This country is a lovely country, you know it is a beautiful country, really lovely, but I don’t know what is happening now. I’ve been here since 1972 but it wasn’t racist, it wasn’t that much scary. (Aamal, Triad 19) Racism wasn’t a big issue in the ’70s but now it is like people constantly telling black people ‘Oh, go home, you don’t belong here’. (Fadwa, Triad 19)

In summary, our qualitative work found that the referent object of security was overwhelmingly that of personal safety; that participants think in terms of security threats on a scale encompassing personal, community, national, and global levels; and that geographical and historical devices structure the way in which perceptions of contemporary security threats are expressed in group conversation.

Which issues do citizens consider to be a security threat? In marked contrast to the popularity of academic debates about the meaning and definition of the concept of security (Baldwin, 1997; Buzan, 1991; Buzan and Hansen, 2009; Herz, 1950; Huysmans, 1998; Krause and Williams, 1997; Lippmann, 1943; Morgan, 2000; Peoples and Vaughan-Williams, 2014; Walt, 1991; Wolfers, 1952), our focus groups struggled when they were asked to discuss and define this concept in abstract, generalised and depersonalised terms. It was only against the backdrop of questioning and discussion about specific and concrete issues associated with threat that participants face in their everyday lives that we could begin to gain a deeper understanding of their perceptions and understandings of the core issues at stake. A significant proportion (thirty minutes) of each triad was therefore devoted to exploring which issues individuals and groups felt were the most pressing

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

69

‘security threats’ to them and why. At first, moderators gave participants an opportunity to think alone and write down a list of possible issues. This exercise was followed by a group discussion that compared and contrasted their respective individual lists. Later on, groups were shown the list of priority risks contained in the UK NSS (2010) (see Figure 3.1 below), which was used as a stimulus for further reflection on the issues they had identified. Once a long-list of security threats had been drawn up on the basis of the participants’ individual reflections and the content of the NSS, the moderator then invited participants to sort these into low, medium, and high clusters. This ‘prompted mapping’ method was guided by questions such as: Which of these security threats are more or less of a concern for you? What is the likelihood of these affecting you? How great an impact would they have? The objective was not so much to arrive at a consensus or a definitive hierarchy of threats, but to use this multi-stage discussion as a way of exploring how participants approached the question of security threats, what language and examples they used to illustrate their thinking, and the logics according to which they structured their conversations. Because of their own vernacular theories of everyday (in)security – predicated on the referent object and scale referred to above – the kinds of issues that groups discussed as threatening their security on a daily basis diverged significantly from the priority risks of the NSS, which commonly appeared strange and aloof. Take the issue of crime, for instance. While organised crime features across each of the three tiers of significance in the NSS, the type of criminal activities that our groups commonly spoke of as threatening differed markedly. On the one hand, ‘hostile attacks on UK cyber-space by other states and large-scale cybercrime’ is listed as a Tier One issue. On the other hand, our participants were much more concerned about online banking (Triads 1, 2, 5, 6, 7, 8, 11, 12, 13, 14, and 16) and cyber-bullying via social networking websites (Triads 1, 2, 7, 11, 12, and 16). Equally, local crime – especially that threatening personal safety – was typically prioritised. Thus, Ben and Angela in Triad 12 were fearful of ‘broken windows’, ‘being jumped or robbed’, ‘having your car stolen’, ‘home security’, or just walking the streets: ‘Well, you know if you walk down any high street in Scotland it’s just full of zombies, you know, junkies on drugs, and any one of these could just snap at you and pull out a knife’ (Ben, Triad 12). Given the timing of our study – two years after the nationalisation of Northern Rock, which followed the first run on a UK bank in 150 years, and in the context of austerity measures introduced by the Conservative–Liberal Democrat coalition government after 2010 that included a rise in VAT from 17.5 percent to 20 percent and a two-year public sector pay freeze and other efforts to sharply reduce public spending as a share of GDP, while average house prices remained at more than four times average annual income – the issue of economic security featured more prominently as a threat among our group discussions than those reported in previous research of a similar kind. For Kieran in London, for example, the fear of unemployment and financial insecurity – issues to which

Everyday security threats

70

Tier One: The National Security Council considered the following groups of risks to be those of highest priority for UK national security looking ahead, taking account of both likelihood

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

and impact.

•

International terrorism affecting the UK or its interests, including a chemical, biological, radiological or nuclear attack by terrorists; and/or a significant increase in the levels of terrorism relating to Northern Ireland.

•

Hostile attacks upon UK cyber space by other states and large scale cyber crime.

•

A major accident or natural hazard which requires a national response, such as severe coastal flooding affecting three or more regions of the UK, or an influenza pandemic.

•

An international military crisis between states, drawing in the UK, and its allies as well as other states and non-state actors.

Tier Two: The National Security Council considered the following groups of risks to be the next highest priority looking ahead, taking account of both likelihood and impact. (For example, a CBRN attack on the UK by a state was judged to be low likelihood, but high impact.)

•

An attack on the UK or its Oversees Territories by another state or proxy using chemical, biological, radiological or nuclear (CBRN) weapons.

•

Risk of major instability, insurgency or civil war overseas which creates an environment that terrorists can exploit to threaten the UK.

•

A significant increase in the level of organised crime affecting the UK.

•

Severe disruption to information received, transmitted or collected by satellites, possibly as the result of a deliberate attack by another state.

Figure 3.1  UK National Security Strategy (2010): priority risks Source:  NSS (2010).



The scope of security threats and their causes

71

Tier Three: The National Security Council considered the following groups of risks to be the next highest priority after taking account of both likelihood and impact.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

•

A large scale conventional military attack on the UK by another state (not involving the use of CBRN weapons) resulting in fatalities and damage to infrastructure within the UK.

•

A significant increase in the level of terrorists, organised criminals, illegal immigrants and illicit goods trying to cross the UK border to enter the UK.

•

Disruption to oil or gas supplies to the UK, or price instability, as a result of war, accident, major political upheaval or deliberate manipulation of supply by producers.

•

A major release of radioactive material from a civil nuclear site within the UK which affects one or more regions.

•

A conventional attack by a state on another NATO or EU member to which the UK would have to respond.

•

An attack on a UK overseas territory as the result of a sovereignty dispute or a wider regional conflict.

•

Short to medium term disruption to international supplies of resources (e.g. food, minerals) essential to the UK.

Figure 3.1, cont’d

the NSS is also oblivious given its alternative referent object, purpose and methodology – were of most significance: I think money is at the forefront, because [. . .] money is great security. Like, 20 years ago you would come out of school and you would kind of have a plan, rent a flat, and know that in five years’ time you could afford to buy a flat or a house. You have no sort of security nowadays, most people will rent for the rest of their lives now. (Kieran, Triad 1)

Everyday security threats

72

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

In seven of the groups (Triads 1, 2, 5, 8, 11, 12, and 15), fears relating directly to personal and family job security were raised immediately in unprompted discussions. Often these issues were fused with more general concerns with perceived levels of immigration: People from Europe are coming in and taking our jobs. (Sandra, Triad 6) My son was admitted for a couple of weeks in hospital and I think 80 per cent of the staff were foreigners. So I think job-wise it’s immigration that is a big threat to lots of people at the moment. (Yoshita, Triad 7) It is not necessarily that I take issue with the fact that there is a Polish builder in my country, I don’t really mind about that, my issue is that they come here to work because they can’t get work in their own country or it is a better standard of living over here [. . .] They will earn their money but then they will send it back home, and if it was an English person having that job the money that they earned would be spent in this country. I think that is an issue. (Steve, Triad 1)

One issue where there was a high degree of convergence between the priority risks of the NSS and the opinions of our participants was the related issue of threats perceived to be posed by weak border security (Triads 1, 6, 8, 9, 12, 15): I recently spent some time out in Australia and I believe that they’ve got the right way to do it now, as soon as they catch people they just turn around and send them back from where they have come from. I really do think that as a nation we need to harden [our border] up. We are too much of a soft touch. (Stuart, Triad 9) The Home Secretary better get our borders sorted out – they are going to get through because we don’t have enough people. (Pam Triad 6) There has been a problem for years and years, which has gone untackled and there are certain areas of Dover now where it is purely illegal immigrants waiting to have their application processed. (Shaun, Triad 1) I think that the government should be taking more control of who is coming into the country. I think we are far too lenient. Watch any border control programme. [. . .] I am worried about fanatics coming into the country, getting in and getting lost in the system and then meeting up, teaming up with others, online as we said, meeting up and joining together. (Angela, Triad 12)

Janet, Bob, and Sally in Leicester not only agreed with the government’s prioritisation of the issue of border security – a Tier Three risk in the NSS – but also expressed strong views relating to ‘illegal immigration’: Q:

What is the implication of weak border control, what is it that you’re worried about? Sally: Letting the wrong sort of person in. Bob: Yes, that’s a fair description, and I’m not racist for one second, but infestation of foreign bodies of people that don’t really need to be here, because their own country can’t look after them. We are too soft with them. Janet: Yes. Let anybody in. (Triad 13)

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

73

The group referred to a generalised fear of weak border control without articulating what exactly constitutes the ‘wrong sort of person’ or on what grounds ‘they’ might pose a security threat. Moreover, Bob’s use of medicalised imagery implies the existence of a pure body politic that ‘deserves’ to be ‘here’ and whose very security depends on the prevention of ‘contamination’ by those who, he says – without specifying any criteria – do not. Indeed, a strong desire for ‘tougher border security’ was a common theme running throughout our groups. Some of the most vocal and passionate calls for more rigorous border security came particularly from our Muslim and Sikh participants: I feel the same way as English blokes: we need to tackle this. (Pamet, Triad 3) What happens if a bunch of Al Qaeda comes from Europe and we don’t have our border security sorted? I think this is a massive security issue and I don’t know whether they will be able to deal with it or not. (Aban, Triad 3) It bothers me because to me this is my country. I live here, this is where my family is and I believe very strongly that immigration is going to affect everybody. Whether somebody’s a Muslim or not to me that’s irrelevant here whether all these people come from a Muslim country or not doesn’t make a difference – they’re illegal. (Daksh, Triad 20)

Abbas made similar identity-based claims to Bob in Triad 15 in Leicester about immigration as a security threat, but he introduced a further division within the category of the ‘immigrant’ between those (such as his parents) whom he considered to have made a positive contribution to British life and those (such as ‘the Romanians’), whom he deems to be ‘taking advantage’ of society and thus a threat: When our parents came over we contributed to society. These are different, they’re not here to do that, they’re here for a totally different reason. Purely to take advantage of our society. I don’t want to discriminate, but basically at the moment I think it is the Romanians. I think we’re the hardest working ethnic minority group in Britain right now. It’s a national security issue. I don’t know how many people have slipped under the radar. I don’t know what kind of security we are running here. (Abbas, Triad 15, emphasis added)

Steve in Bristol went further than both Bob and Abbas by making a performative association between the very presence of ‘immigrants’ in Britain today and the threat of international terrorism: Well obviously you know the biggest security threat is from all these illegal immigrants and immigrants that we’ve got in this country, I’m not going to say I’m racist, but I just wish they’d all go back home because they’re ruining this country, they have for many years [. . .] We just let it go on and on and on and therefore you’re causing a situation whereby there’s threats and terrorism. (Steve, Triad 10, emphasis added)

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

74

Everyday security threats

However, Steve’s associative securitizing (Huysmans, 2014) between immigration and terrorism was isolated. Indeed, while the issue of international terrorism appears at the top of Tier One of the NSS, only four of our groups framed their responses spontaneously in terms of ‘bomb threats, things like that’ (Triads 2, 12, 16, and 17). As one participant put it: ‘I feel less scared of a terrorist than a mugger or a burglar’ (Doug, Triad 11). Of the four groups in which terrorism was raised as a security threat, two consisted exclusively of younger (18–25-yearold) participants. For other groups it was not ‘terrorism’ per se that was dwelt upon, but more specifically the threat of particular religious and ethnic groups being stereotyped and associated with terrorist activity: ‘Everyday when you switch on your TV or radio there is always a story about Muslims, Islam, and terrorism. People hear this and draw their own conclusions’ (Syazwana, Triad 7). Islamophobia – another issue unlikely ever to be included in any of the official tiers – was overwhelmingly cited among our Muslim triads (Triads 3, 7, 15, and 19) as the most significant security threat facing participants in their everyday lives: Islamophobia is increasing day by day not only in the UK, all over Europe. When nationalism is linked to racism, of course, it is a danger. (Pamet, Triad 3) Before they even know anything they draw up the conclusion it is somebody Muslim [. . .] It does make you worry a bit in a sense that there is a lot of other people make a threat, nobody ever mention the word Christian [. . .] and that really, really makes me sad. (Shabnam, Triad 7)

More generally, racism and hate crime were also discussed as security threats among ethnic minority groups: People have called me a ‘fat slave’ in the past [. . .] Some people feel like because we have come over and taken their jobs and this and that [. . .] They feel threatened by us, while, at the same time, we too feel threatened by them. (Azza, Triad 19) All of our kids were in the park and this guy, a mixed race, constantly telling black people, ‘Oh, you, go home, you don’t belong here’. (Shazi, Triad 19)

These dynamics arising from our focus group work – in particular: 1) the impact of the securitization of neighbourliness and attempts to foster vigilance and suspicion in public spaces on minority communities; and 2) the relationship between government messages and public behaviour – are the focus of Chapter 5. For now, however, the primary findings of the qualitative discussions highlight that the main issues perceived to pose security threats in contemporary British society include: knife crime, burglary, online fraud, unemployment and the financial crisis, immigration and weak border security, UK foreign policy, and Islamophobia, racism, and hate crime. The identification of these issues was equally important for the purposes of the design of the survey, in informing us about how participants might define security – the results of which we begin to outline in the remaining sections of the present chapter.



The scope of security threats and their causes

75

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Perceptions of the breadth of security threats As described in Chapter 2, our examination of threats in the ICM survey ran from the broadest collective, the global, through national- and community-level threats, which are still collective but at a much more intimate and personal level than the state or the globe, to threats to the individual and to his or her family. We see community-level threats as closer to what is usually meant by personal rather than by sociotropic threat. We operationalise perceptions of the breadth of threats as the number of issues that people identify as global, national, community, or personal threats. We begin by exploring the scope and composition of perceptions of security threats from the global to the individual level. How threatened are individuals at these different levels? Previous research provides us with little guidance. One possibility is that individuals compartmentalise, identifying a single security threat such as terrorism as the principal global threat, the economy as the national threat, knife crime as the community threat, and so on. This would lead us to expect limited variation in the breadth of threats at different levels – basically to one – but variation in what those threats are. A second possibility is that in seeing security threats as predominantly about ‘feeling physically safe where you are’, individuals identify more threats that are ‘closer to home’, i.e., they are more sensitive to threats on their doorstep, which would lead to the expectation of perceptions of more threats at the personal than at the global or national levels. Figure 3.2 demonstrates that neither of these possibilities is accurate. It shows the proportion of respondents in our survey identifying particular threats as their perceptions of the total number of security threats increased from one to seven. The four graphs display the five main security threats identified at the global, national, community and personal levels. The subheadings for each graph indicate that the average number of perceived threats was greatest at the global level, with fewer issues identified as national level security threats, and the lowest average being personal threats, where perceptions were of roughly two on average.1 Figure 3.2 also demonstrates a pattern in perceptions of threats: the ‘rankings’ of particular threats tend not to change as the breadth of security threats increases. For example, with global threats, terrorism was the most frequently identified threat for respondents who identified one threat, at 30 per cent, with nuclear war the joint fifth most frequent choice with 6 per cent; for respondents identifying seven global threats, terrorism was still the most frequently mentioned, but by 80 per cent of these respondents, and nuclear war the fourth most frequently mentioned with 60 per cent. Thus, Figure 3.2 indicates agreement about what the most salient threats are, especially at the global and national levels: the differences between respondents perceiving more or fewer threats is not in the issues that are most threatening but in whether or not an issue crosses a threshold to become a tangible security threat. For example,

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Global (mean number of threats = 6.6) 90 80 70 60 50 % 40 30 20 10 0

National (mean number of threats = 4.0)

Terrorism Religious extremism Environment Nuclear weapons Economy %

1

2

3

4

5

6

90 80 70 60 50 40 30 20 10 0

Terrorism Economy Religious extremism Immigration Border control

1

7

2

Community (mean number of threats = 2.2) 90 80 70 60 50 % 40 30 20 10 0

Knife crime Online fraud Burglary Immigration

Economy

1

2

3

4

5

Number of threats

3

4

5

6

7

Number of threats

Number of threats

6

7

Individual (mean number of threats = 2.1) 90 80 70 60 50 % 40 30 20 10 0

Environment Economy Knife crime Burglary Online fraud

1

2

3

4

5

6

Number of threats

Figure 3.2  Major threats at the global, national, community, and individual levels

7

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

77

for respondents identifying only one threat, the threshold is not generally crossed for nuclear war as a global threat, but it is crossed for more than half of those identifying seven global threats.2 Figure 3.2 also shows that there is variation in the kinds of security threats that are salient at different levels. At the global and national levels, the concerns are terrorism, the economy, and religious extremism. However, whereas immigration and border control are key national security threats, issues like nuclear weapons are seen as more pressing global security threats. In addition, while perceptions of salient community- and personal-level security threats are similar to each other, they are often different from the issues identified as global and national threats. Thus, although the economy remains a threat at the community and personal levels – along with immigration – burglary, knife crime and online fraud loom much larger than at the global or national levels. Huddy et al. (2002) characterise the effects of national and personal threat from terrorism as distinct but related. This notion of threats at different levels as distinct but related appears to be true more broadly: the correlations in the breadth of threats are mostly in the 0.4 to 0.5 range, showing that while the specific threats may vary, individuals who see more threats at one level are likely to see more threats at another. At the same time, however, the breadth of threats perceived at one level is not entirely predictive of the breadth of threats perceived at another. As an additional check, we factor analysed perceptions of threats at the global, national, community, and personal levels to see whether there were patterns in the kinds of threats identified.3 The results are easy to summarise: there was little evidence of such patterns. The uniqueness scores of most issues were close to or above 0.5 for all four levels, showing that they share little variance. More­ over, the factors that did emerge were difficult to categorise. At the global level, there appears to be some link between identifying ‘crime’, broadly construed, as a security threat – knife crime, burglary, online fraud – and a second factor for global nuclear and military threats such as from nuclear war, Russia and China, while at the national level there is a similar ‘crime’ factor and a second factor that encompasses threats to health and the environment, but the factors are weak and these are not the issues of greatest concern to individuals, as shown by Figure 3.2. Having discussed the patterns in perceptions of the breadth of threats, we now turn to the influences on them.4 Influences on perceptions of the breadth of threats As discussed in Chapter 1, previous research has provided variable evidence of systematic influences on perceptions of threat. The influences examined belong in three principal categories: predispositions such as mortality salience and

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

78

Everyday security threats

authoritarianism; media habits; and individual-level attributes such as education.5 The question is whether we can go beyond influences on perceptions of specific threats in discrete issue areas such as terrorism to a broader understanding of when and why these influences on threats vary across different issue areas. We have few a priori expectations about whether and for which kinds of variables we might see more robust effects on community and personal threats than on global and national threats; we explore such variation in the analysis. We expect that many of the kinds of factors that affect perceptions of threat at the global and national levels will also affect perceptions of threat at the community and personal levels but that, as we move from sociotropic to personal threats,6 the influences on perceptions will become less systematic, as more localised, random factors, which our models do not pick up, come into play. Rather than separate models for perceptions of threats at each level, we examine the four levels of threat simultaneously in a structural equation model.7 Perceptions of the breadth of threats are a count of the number of perceived threats at each level. The predictors are those depicted in Figure 3.2. All of the independent variables, including age, are coded from 0 to 1, meaning that effect sizes are easily compared. Table 3.2 presents the estimates, with separate columns for each level of threat.

Table 3.2  Influences on perceptions of the breadth of global, national, community, and personal/family threats Global Authoritarianism Mortality salience Time spent watching current affairs TV Educated to degree level or higher Woman Age Muslim Religiosity White Constant

0.01 (0.03) 0.26 (0.05)* 0.10 (0.03)*

National 0.13 (0.04)* 0.28 (0.06)* 0.14 (0.04)*

0.06 (0.02)*

−0.01 (0.02)

0.13 (0.02)* 0.23 (0.04)* −0.12 (0.03)* 0.03 (0.03)

0.07 (0.02)* 0.35 (0.05)* −0.10 (0.04)* −0.06 (0.04) 0 1.02 (0.05)*

1.57 (0.04)*

Community

Personal/family

0.16 (0.05)* 0.16 (0.09)# 0.19 (0.06)*

0.16 (0.05)* 0.35 (0.09)* 0.28 (0.06)*

0.01 (0.03) 0.15 (0.07)* −0.03 (0.07) 0.22 (0.05)* −0.11 (0.06)# 0.56 (0.08)*

0.08 (0.03)* 0.66 (0.07)* −0.02 (0.07) 0.19 (0.06)* −0.24 (0.06)* 0.26 (0.09)*

Notes:  n = 1903. Log likelihood = −19031.186. *p < 0.05; #p < 0.10 (two-tailed test). Estimates are from a generalised structural equation model and assume the endogenous variables – breadth of threats – have Poisson distributions. Statistical significance may not always appear accurate due to rounding.



The scope of security threats and their causes

79

Authoritarianism

Mortality salience

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Global National Media habits

Community Personal

Age

0

0.5 1 1.5 Increase in Number of Threats

2

Figure 3.3  Maximum effects of main influences on breadth of threats Notes:  Simulations are with all other variables set at their mean or mode – a white, nonMuslim woman, without higher education.

We begin by discussing the influences that are common across levels of threats and then turn to those whose effects on perceptions are distinct across levels of threat. The results in Table 3.2 confirm that there are common influences on perceptions of the breadth of threats that mean that some individuals perceive systematically more threats than others. Indeed, several of the variables have pervasive effects on perceptions of the breadth of threats. One is mortality salience. Somewhat contrary to our expectation that the influence of mortality salience would be confined to global- and national-level threats, individuals whose mortality was more cognitively accessible perceived more threats at all levels.8 This suggests that awareness of one’s mortality not only affects perceptions of physical danger but also elevates perceptions of threats that induce uncertainty or anxiety – contrary to some previous research (Greenberg et al., 1994, 1995) – perhaps because mortality salience leads to a ‘conservative shift’ (Burke et al., 2013: 186) that elevates perceptions of threat in general, regardless of whether they are physical or uncertainty- or anxiety-inducing.9 To illustrate the substantive meaning of these relationships, Figure 3.3 shows the maximum effects of mortality salience (and three other variables discussed below) when all other variables are at their mean or mode. It indicates that, as expected, morality salience has relatively large effects on perceptions of global and national threats, adding more than one additional threat at each level. But Figure 3.3 also shows the impact of mortality salience on perceptions of personal threats, where although the effect is to raise the number of perceived threats by

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

80

Everyday security threats

less than half the amount it does for global threats, it should also be borne in mind that the average number of global threats is more than three times larger than that of personal threats. Returning to the estimates in Table 3.2, they also show that watching proportionally more television that covers politics and current affairs has a consistent influence on harbouring more threats at all levels. This echoes Ridout et al.’s (2008) finding and suggests either that the world presented on television news heightens the threat levels of its most dedicated viewers, or that threatened individuals monitor current affairs more closely. The effects of watching political and current affairs television are most consequential at the community and personal levels: the maximum effect of watching politics and current affairs programmes is to raise the breadth of personal threats by about 0.7, while the influence on global and national threats is marginally larger but less noteworthy given the higher number of threats identified on average at these levels. Age also has a consistent impact on perceptions of the breadth of threats, with older individuals more threatened, and effects that are larger than any of the other variables on breadth of personal threats. The influence of variables like mortality salience and age is consistent in terms of its statistical significance on perceptions of the breadth of threats at different levels. Authoritarianism is different, however. While authoritarianism has the expected positive relationship with perceptions of the breadth of national, community, and personal threats – in this sense authoritarians are chronically threatened – its influence does not extend to perceptions of global-level threats. Thus, while authoritarians may be ‘relentlessly sociotropic boundary maintainers’ (Stenner, 2005: 32), their fears about nonconformity or the undermining of approved authority figures do not extend beyond an individual’s national borders, according to this analysis. Figure 3.3 also shows that the impact of authoritarianism is relatively modest for perceptions of personal threats, which accords with Feldman’s (2013) claim that personal threats raise issues of autonomy that are less salient to authoritarians than the desire for group conformity. The estimates in Table 3.2 also show systematic variation in the effects of other variables at different levels. For example, there is an effect of gender on perceptions of threats at all levels other than community, with women identifying more threats, while religiosity primarily affects perceptions of community and personal threats. We cannot establish why religiosity influences perceptions of threats at these levels rather than global or national threats but it is noteworthy that more frequent attendance at a place of worship is associated with perceptions of more community threats; perhaps places of worship are refuge from the world that is on your doorstep. Education is associated with identifying more global threats but has no influence on perceptions of national threats, while white individuals identify fewer community and personal threats. Lastly, being

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

81

a British Muslim does not result in perceptions of more threats at the community or personal levels but Muslims see fewer global and national threats than others.10 These results beg a further question, however. We have shown that there is a general tendency to identify more global than national threats, and more national than community or individual threats, but clearly some individuals buck these trends: some people see more threats at the national than global levels, for example. We focus in Table 3.3 on contrasts between perceptions of threats at the global, national, and community levels.11 Table 3.3 shows that two-thirds of respondents identify more global than national threats and four-fifths more global than community threats; this leaves fairly large minorities who either identify more or the same number of threats at the national or community than the global level, however. This is equally true of the national–community comparison, where almost one in two respondents views the breadth of threats as either the same (36 per cent) or the breadth of community threats as somewhat greater (10 per cent). This begs the question of who these individuals are. Table 3.2 provided some clues – for example, authoritarians seem to perceive fewer threats at the global level – but we now seek to examine this statistically by looking at models in which we predict the difference in the number of threats identified in each comparison in Table 3.3 (with the exception of community versus individual threats). The dependent variables are the differences in the breadth of threats identified at each level, operationalised as the breadth of threats at the higher collective level minus the breadth of threats at the lower level. Thus, for example, in the global versus national comparison, a positive score implies perceptions of more global threats and a negative score implies perceptions of more national than global threats. A positive coefficient therefore shows an influence towards perceptions of more global than national threats and a negative coefficient

Table 3.3  Comparisons of perceptions of the breadth of contemporary security threats at different levels Global vs. national Global vs. community National vs. community Community vs. personal Note:  n = 2004. Numbers are %.

More global 68 More global 81 More national 54 More community 7

Same 20 Same 16 Same 36 Same 68

More national 12 More community 3 More community 10 More personal 25

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

82

Everyday security threats

implies the opposite. We initially examine the influences of the same independent variables as in Table 3.2. Table 3.4 displays the results. The estimates in columns 1, 3, and 5 of Table 3.4 confirm some of the differences suggested by Table 3.2: authoritarianism is an influence on perceptions of more national than global threats but does not affect differences in perceptions at other levels; mortality salience is associated with systematically greater perceptions of the breadth of global and national threats than of community (or personal12) threats. Older respondents also identify more global and national than community-level threats, perhaps because they are more likely to be readers of newspapers and to watch the news (although Table 3.4 shows no impact of watching proportionally more programming about current affairs). Other results, such as the fact that women identify systematically more global and national than community-level threats in comparison to men are intriguing but more difficult to explain. It is also important to note, however, that the variables explain very little variance, with adjusted R2 of no more than 0.03 – or 3 per cent of the variance in perceptions – indicating that the differences are either primarily a product of influences we have not accounted for or that they are simply idiosyncratic. One possibility is that the tendency to identify systematically more threats at one level than another, i.e., to an unusual degree compared to the rest of the sample, is related to different geographical identities. For example, an individual whose primary identity is with the nation might be more likely to identify national-level threats. In order to examine this possibility, we asked respondents whether they primarily saw themselves as citizens of the world (17 per cent), of Europe (5 per cent), the United Kingdom (49 per cent), another nation or region of the UK such as Scotland, Wales or the south-east of England (21 per cent), or their local area (5 per cent). We constructed three variables capturing primary identity with the UK, primary identity with Scotland, Wales, a region or locality, and don’t knows (3 per cent). The reference, or excluded, variable is therefore respondents who see themselves as citizens of an entity larger than the nation.13 The effects are shown in columns 2, 4, and 6 of Table 3.4, in which the models are identical to those in columns 1, 3, and 5 but include the additional variables for perceptions of citizenship. The results suggest that perceptions of citizenship matter but that their effects on the overall amount of variance explained are minor and have no clear direction. Thus, identifying as a citizen of the UK is associated with perceptions of more national than global threats but has no effects on any of the other comparisons, including national versus community threats. One might therefore draw from this that identity as a citizen of a certain geographic area leads to a focus on that area and thence on perceptions of more security threats. However, the effects of identifying as a citizen of a sub-UK entity do not follow this pattern: it is associated with perceptions of more national than global threats and of more national than community security threats, suggesting

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Table 3.4  Influences on perceptions of the relative breadth of global, national, community, and personal/family threats

Authoritarianism Mortality salience Time spent watching current affairs TV Educated to degree level or higher Woman Age Muslim Religiosity White Primarily identify with the UK Primarily identify with subnational region Don’t know primary geographic identity Constant Adjusted R2

Global > National

Global > Community

National > Community

(1)

(2)

(3)

(4)

(5)

(6)

−0.52 (0.27)# 0.50 (0.47) 0.10 (0.32) 0.14 (0.08)# 0.59 (0.17)* 0.14 (0.39) −0.28 (0.37) 0.53 (0.31)# 0.16 (0.33)

−0.36 (0.27) 0.53 (0.47) 0.11 (0.32) 0.13 (0.08) 0.60 (0.17)* 0.24 (0.39) −0.26 (0.37) 0.46 (0.30) 0.32 (0.33) −0.44 (0.22)* −0.86 (0.25)* 0.47 (0.57) 1.99 (0.51)* 0.01

−0.35 (0.29) 1.29 (0.50)* 0.25 (0.34) 0.10 (0.09) 0.83 (0.18)* 1.27 (0.41)* −0.74 (0.39) −0.26 (0.32) 0.20 (0.34)

−0.26 (0.29) 1.32 (0.50)* 0.24 (0.34) 0.09 (0.09) 0.83 (0.18)* 1.33 (0.41)* −0.73 (0.39)# −0.29 (0.32) 0.27 (0.35) −0.31 (0.23) −0.41 (0.26) 0.02 (0.61) 3.03 (0.54)* 0.02

0.17 (0.21) 0.79 (0.36)* 0.15 (0.25) −0.04 (0.06) 0.24 (0.13)# 1.13 (0.30)* −0.46 (0.29) −0.79 (0.24)* 0.04 (0.25)

0.10 (0.21) 0.79 (0.36)* 0.13 (0.25) −0.04 (0.06) 0.23 (0.13)# 1.10 (0.30)* −0.47 (0.29)# −0.75 (0.24)* −0.04 (0.25) 0.13 (0.17) 0.46 (0.19)* −0.45 (0.44) 1.04 (0.39)* 0.03

1.82 (.50)* 0.01

Notes:  n = 1903 * p < 0.05; # p < 0.10 (two-tailed test). Estimates are from ordinary least squares regression models.

2.91 (0.53)* 0.02

1.09 (0.39)* 0.03

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

84

Everyday security threats

that identity as a sub-UK citizen is associated with perceptions of fewer security threats at the subnational level. We cannot shed further light on these findings but this warrants further investigation in future research.14 Finally, we looked at whether there were any differences in the influences on perceptions of the breadth of Tier One threats as defined in the 2010 NSS, versus other kinds of threats. The rationale was that there may be differences in the influences on the kinds of threats identified by government elites as most important and other kinds of threats. Tier One threats are those regarded as most pressing by the UK government. If those concerns have seeped down to the public we might expect authoritarians, for example – who are particularly influenced by the views of established authorities – to be particularly likely to share them. We separated perceptions of the breadth of global, national, community, and personal threats into Tier One versus ‘other threats’ and re-estimated the models in Table 3.2. The threats in the survey defined as Tier One were: terrorism, attacks on cyber-space and cyber-crime, a health pandemic, and a nuclear accident such as occurred at Fukushima. We do not show the results from the models because they are easy to summarise: splitting threats into Tier One and others made little or no difference to the estimates displayed in Table 3.2. For example, while authoritarianism has a slightly stronger effect on perceptions of the breadth of Tier One global threats, with a coefficient of 0.08 and a standard error of 0.06, it still falls short of conventional statistical significance, and the size of the effect is dwarfed by that of other variables such as mortality salience (0.25 with a standard error of 0.11). Thus the influences on perceptions of the breadth of threats appear to be quite generalisable and do not vary by what government elites define as the most salient threats.15 We explore this issue of the efficacy of government messages about security threats in more detail in Chapter 5. Perceptions of specific threats We switch now from looking at perceptions of the breadth of threats to perceptions of particular threats. Our focus is again on the influences on perceptions. We expect to see variation here based on the interaction of the characteristics of different security threats and the characteristics of individuals. For, as Feldman observes (2013: 55), ‘a person could . . . be personally threatened by the impending loss of a job, which is likely a very different type of threat than one that involves mortality’. Our analysis therefore focuses on four threats that both vary in their characteristics and that were also consistently identified as security threats by respondents at one or more levels in the course of focus group discussions and in the surveys: terrorism, immigration, the economy, and the environment. Security threats such as terrorism pertain to mortal physical danger and social disruption (Berrebi and Klor, 2008), while the threat from environmental

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

85

degradation pertains to damage to the physical world and indirectly to livelihood. Economic threat, on the other hand, is more ambiguous, with implications for social order16 but also resulting from growing inequality (Duckitt and Sibley, 2010; Shaffer and Duckitt, 2013).17 For the characteristics of individuals, while our models include the same variables as for perceptions of the breadth of threats, we also focus on four that allow us to explore our theory about the effects of variation: mortality salience, authoritarianism, media habits, and education. The point about variation in the characteristics of individuals is that individual-level mechanisms for the influence of dispositions like mortality salience differ from those behind authoritarianism, which in turn differ from the mechanisms for the effects of education. Authoritarian concerns are rendered salient by ‘insecurity, uncertainty, and danger’ to the ingroup (Shaffer and Duckitt, 2013) and not – i.e., may be negatively related to – ‘competition, resource scarcity, and inequality’ (Duckitt and Sibley, 2010: 1874–5). This implies that the influence of authoritarianism on perceptions of different kinds of threat may vary in sign, being positive for terrorism but negative for environmental degradation.18 Duckitt and Sibley’s research also suggests that authoritarianism will be positively related to economic threat if the threat is from social disorder with implications for ingroups but negatively related if the primary threat stems from competition or inequality (see also Feldman, 2013; Wildavsky and Dake, 1990).19 In contrast, mortality salience is thought to be influential under conditions of realistic or symbolic threats to the self (Schmiel et al., 1999), i.e., by mortal physical danger or threats to cultural worldview, and is not aroused by mere insecurity, uncertainty, or anxiety-producing events (though see note 9). But there is no relationship rather than a negative relationship to threats with the latter characteristics (Greenberg et al., 1994, 1995). This suggests that mortality salience enhances perceptions of threat from terrorism and immigration and is unrelated to threats from the economy or environmental degradation. The bases for effects of media coverage, on the other hand, do not vary by type of threat. If, for example, news media heighten perception of threats from a ‘mean world’ (Gerbner et al., 1980; Hawkins and Pingree, 1981) variation in the influence of media exposure should be a function of what the media choose to emphasise, i.e., agenda setting and framing, rather than characteristics of the kinds of danger each threat poses. Similarly, if the influence of education on perceptions of threat is because highly educated individuals have a greater grasp of probabilistic information (Huddy et al., 2002), we would not expect to see variation in the effects of education because of variation in the kinds of danger they present, but rather because of variation in the availability of base rate information about the threats: where it is harder to grasp – ‘hard issues’ in the terminology of Carmines and Stimson (1980) – we would expect to see effects of education. Given that base rate information about the absence of a threat from

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

86

immigration or about the threat from environmental degradation (e.g., Gavin et al., 2011) is harder to come by than information on the threat from terrorism or the economy, we would expect to see effects of education for these issues.20 The expectations implied by these relationships are summarised in Table 3.5. How these influences may vary with perceptions of different security threats at different levels is an open question but we reiterate that previous research has emphasised the power of sociotropic over subnational considerations. Table 3.6 begins by showing how often the four issues were mentioned by survey respondents as a threat at each level and where they ranked among all threats. Echoing Figure 3.2, it shows considerable variation in the patterns of threat: the economic crisis is seen as a major threat at all levels but terrorism is clearly identified mostly as a global and national threat – individuals generally feel neither personally threatened by terrorism, nor that it is a threat to their community in some sense – while immigration is more of a threat at national and subnational levels. We do not display them here but it is worth noting that the correlations between different threats are generally very small. As would be expected, the highest correlations are among the same threats at different levels, i.e.., an issue as a threat at the global level and an issue as a threat at the national level, although the strongest is 0.52, between personal and community environmental

Table 3.5  Expected influences on perceptions of specific threats Threat from . . .

Terrorism

Immigration

Economy

Influence of: Mortality salience Authoritarianism Media exposure Education

None +ve +ve +ve +ve −ve +ve when there is media emphasis on the issue None None −ve

Environment

None −ve +ve

Table 3.6  Perceptions of contemporary security threats

Terrorism Immigration Economy Environment

Global

National

Community

Personal

69 (1) 26 (12) 46 (4) 44 (5)

48 (1) 33 (4) 45 (2) 22 (8)

8 (10) 16 (4) 36 (1) 11 (7)

10 (7) 11 (6) 38 (1) 12 (5)

Note:  n = 2004. Numbers are %. Figures in parentheses are ranks.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

87

threat and most of the correlations are much weaker than that. But there are some intriguing exceptions, such as a stronger correlation between personal threats from terrorism and immigration than between personal threat from terrorism and national or global threat from terrorism. In general, however, while we have already shown moderately strong correlations between perceptions of the breadth of threats at different levels, perceptions of specific threats at different levels are not highly correlated. Table 3.7 presents the estimates of the influences on perceptions of the four threats.21 It shows that there are effects of all the exogenous variables on at least some of these specific threats at some levels.22 The results confirm that heightened mortality salience and authoritarianism are associated with perceptions of threat for some issues – terrorism, immigration – but that mortality salience is unrelated and authoritarianism is negatively related to perceptions of economic and environmental threat. Thus, as hypothesised, mortality salience appears to enhance perceptions of threats that pertain to physical danger or, with immigration, to possible danger to one’s cultural worldview but has weaker effects on perceptions of threats that represent uncertainties about financial security or danger to the physical world. Authoritarianism actually reduces perceptions of these threats, perhaps because authoritarians are so unconcerned by issues that raise considerations of inequality or resource scarcity (Duckitt and Sibley, 2010). This also indicates why there was no relationship between authoritarianism and perceptions of the breadth of global threats in Table 3.2: it is not because authoritarianism has no influence on perceptions of global threats but because its effects are positive for perceptions of some threats and negative for others. These contrasting relationships are more or less equal at the global level, while at the national and subnational levels they are stronger in the direction of enhancing perceptions of threat. While this still suggests that national-level threats loom larger than global-level threats for authoritarians, it provides nuance to the relationship between authoritarianism and perceptions of global threats.23 The impact of current affairs television on perceptions of specific threats in Table 3.7 is spotty, but where it is statistically significant the effects are always positive, as theorised. Interestingly, given previous claims about the connection between the focus of media coverage and sociotropic effects, media habits have a more extensive influence on perceptions of personal threat than other levels. It may be that watching more current affairs television in Britain, with its public broadcasting tradition, intensifies a general sense that there are multiple security threats at all levels (Table 3.2) but helps rather than hinders individuals in making the connections between complex issues and the personal. Finally, Table 3.7 partially confirms our expectations about education. Education is associated with perceptions of more threat from the environment, with lower perceptions of threat from immigration, and is unrelated to perceptions of threats from terrorism. However, we also anticipated that the availability of information about

Table 3.7  Causes of perceptions of threats Terrorism

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Global Mortality salience Authoritarianism Media exposure Education Religiosity Muslim White Woman Age Constant n = 1903

National

Community

Personal

Global 0.92 (0.30)* 1.13 (0.18)* −0.02 (0.21) −0.32 (0.11)* −0.24 (0.20) −0.52 (0.21)*

Economy

Environment

−0.20 (0.26) −0.36 (0.15)* 0.19 (0.18) 0.32 (0.10)* −0.22 (0.17) −0.11 (0.17)

National

Community

Personal

0.07 (0.27) −0.16 (0.26) −0.15 (0.27) −0.41 (0.15)* −0.54 (0.15)* −0.47 (0.15)* 0.15 (0.18) 0.16 (0.18) 0.39 (0.18)* 0.37 (0.10)* 0.03 (0.18) −0.15 (0.17) −0.35 (0.17)* −0.10 (0.17) −0.38 (0.22)# −0.19 (0.21) −0.02 (0.19) −0.13 (0.18) 0.28 (0.10)* 0.23 (0.10)* 0.21 (0.10)* 0.14 (0.10) 0.45 (0.21)* 0.33 (0.21) 0.62 (0.22)* 0.24 (0.22) −0.34 (0.20)# −0.24 (0.20) −0.54 (0.26)* −0.36 (0.26) Log likelihood = −5070.25

National

Community

0.50 (0.35) 0.52 (0.29)# 1.08 (0.17)* 0.73 (0.20)* 0.21 (0.24) −0.16 (0.20) −0.39 (0.11)* 0.05 (0.23) −0.39 (0.19)* −0.58 (0.20)* −0.08 (0.30) 0.28 (0.26) 0.04 (0.10) −0.14 (0.13) 0.20 (0.11)# 0.87 (0.24)* 1.16 (0.23)* 0.82 (0.29)* −2.33 (0.24)* −1.72 (0.22)* −2.84 (0.37)* Log likelihood = −3645.10

0.70 (0.29)* 0.71 (0.27)* 0.36 (0.49) 0.57 (0.43) 1.06 (0.17)* 0.83 (0.16)* 0.44 (0.28) 0.77 (0.25)* 0.06 (0.18) 0.44 (0.31) 0.57 (0.27)* −0.30 (0.19) 0.12 (0.11) −0.04 (0.10) 0.51 (0.30)# 0.33 (0.26) −0.31 (0.18)# −0.30 (0.17)# −0.69 (0.17)* −0.84 (0.18)* −0.72 (0.38)# −0.24 (0.33) −0.50 (0.30) −0.35 (0.28) 0.33 (0.10)* −0.03 (0.10) −0.44 (0.18)* 0.15 (0.16) 1.37 (0.36)* 0.51 (0.23)* 1.02 (0.21)* 0.69 (0.40)# −0.23 (0.21) −1.07 (0.20)* −2.74 (0.46)* −3.45 (0.42)* Log likelihood = −3509.36

Global Mortality salience Authoritarianism Media exposure Education Religiosity Muslim White Woman Age Constant n = 1903

Immigration

Global

National

Community

Personal 0.63 (0.43) 1.09 (0.25)* 0.31 (0.29) −0.14 (0.29) −0.52 (0.41) 0.13 (0.32) −0.20 (0.15) 1.53 (0.35)* −3.74 (0.45)*

Personal

0.01 (0.27) 0.06 (0.32) −0.14 (0.42) 0.16 (0.41) −0.77 (0.15)* −0.60 (0.18)* −0.52 (0.23)* −0.65 (0.22)* 0.16 (0.18) 0.32 (0.21) 0.61 (0.26)* 0.48 (0.25)# 0.39 (0.10)* 0.30 (0.11)* 0.07 (0.26) 0.37 (0.25) −0.34 (0.17)# −0.19 (0.20) 0.08 (0.20) 0.06 (0.33) −0.34 (0.32) −0.04 (0.17) −0.08 (0.29) −0.45 (0.26)# 0.17 (0.10)# −0.00 (0.11) −0.07 (0.15) −0.04 (0.14) 1.45 (0.43)* 0.05 (0.21) 0.30 (0.25) 0.58 (0.34)# −0.05 (0.20) −1.26 (0.23)* −2.08 (0.40)* −2.09 (0.38)* Log likelihood = −3622.96

Note:  *p < 0.05; #p < 0.10 (two-tailed test). Estimates are from a generalised structural equation model using logit estimates. Statistical significance may not always appear accurate due to rounding.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

89

the economy would mean that education would be unrelated to perceptions of economic threat. But that is not the case: education is associated with a greater likelihood to see global and national economic threats. It may be that the economy is never the ‘easy issue’ we suggested it could be as a result of the post2008 financial crisis. Because of the coding, the coefficients in Table 3.7 can be read and compared as the maximum effects of each variable. Thus it is apparent that authoritarianism not only has the most statistically significant relationships with perceptions of threats but that its impact is relatively large across the board. Table 3.7 cannot tell us about the substantive meaning of a 0.2 increase in the probability of judging immigration to be a global or national threat, however – i.e., whether the impact is small or large. To help with that, Figure 3.4 displays estimates of the probability of identifying immigration or the environment as global, national, community, or personal threats given that a respondent is at the lowest or highest level of mortality salience, authoritarianism, or education, setting all other variables at their mean or mode. It demonstrates that the impact of some of these variables is substantial. Authoritarianism is the most obvious example. High authoritarians are roughly twice as likely as low authoritarians to view immigration as a threat at each level. For threat from environmental degradation the substantive impact is similar but in the opposite direction, with probabilities roughly halved when moving from the lowest to the highest levels of authoritarianism. Figure 3.4 also shows that mortality salience has a more modest range of influence – none at all on perceptions of environmental degradation – but that where it does move perceptions, for immigration as a global threat, the maximum effect is also to almost double the probability of identifying immigration as a threat. The impact of education is clearly more modest. In terms of its proportional effect on the probabilities of seeing threats, being educated to degree level or above lowers the probability of identifying immigration as a threat by about 20 per cent and raises the probability of identifying environmental degradation as a threat by between about 10 and 15 per cent. Lastly, we suggested that we should see more systematic influences on perceptions of global and national threats than on community and personal threats. A way of gauging this is by looking at the number of statistically significant effects in Table 3.7: there are thirty-six statistically significant relationships with sociotropic threats compared to twenty-four with subnational threats (at p < 0.10; thirty to seventeen at p < 0.05), which supports the hypothesis. We also said that there should be overlap in the origins of perceptions of specific threats at different levels. Table 3.7 confirms that the origins of perceptions of specific sociotropic threats are not entirely different from personal threats. But this is more true of some variables than others. Authoritarianism and age are factors at all levels of threats, but exposure to current affairs television largely affects perceptions of threat at the personal level, while others such as the control variables of

Threat of immigration

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Mortality s alience 0.6 0.5 0.4 0.3 0.2 0.1 0

Authoritarianism 0.6 0.5 0.4 0.3 0.2 0.1 0

Education 0.6 0.5 0.4 0.3 0.2 0.1 0

Threat of environmental degradation Mortality salience 0.6 0.5 0.4 0.3 0.2 0.1 0

Authoritarianism 0.6 0.5 0.4 0.3 0.2 0.1 0

Lowest level

Education 0.6 0.5 0.4 0.3 0.2 0.1 0

Highest level

Figure 3.4  The effects of mortality salience, authoritarianism, and education on threats from immigration and environmental degradation (y-axes = probability of identifying as a threat, x-axes = level of threat)



The scope of security threats and their causes

91

sex and religiosity are influential at sociotropic and subnational levels for different variables.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Conclusion This chapter began by drawing on our mini-focus group research in order to establish how citizens perceive and understand the concept of security threats and also to gauge the specific issues that they consider to be threatening in the context of their everyday lives. We argued on the basis of our analysis that participants tended to prioritise personal safety as the referent object of security, but that a scale of four levels – personal, community, national, and global – was invoked during discussions. Furthermore, a range of issues were identified as being of pressing importance – some of which are contained in the NSS list of priority risks (e.g., terrorism, weak border security, immigration), while others are not (e.g., economic security, Islamophobia). From here, the analysis then focused on the breadth of threats perceived by individuals, and then on perceptions of the specific security threats of terrorism, immigration, the economy, and environmental degradation. We have looked at this panoply of threats at the global, national community, and personal levels and examined the influences on perceptions of each of these. The analysis in this chapter has validated our claim that a general understanding of what drives perceptions of security threats demands going beyond specific ‘threats of the moment’ at the national and personal levels, which limits our ability to assess how threatened individuals are, by what, and whether there is variation in the determinants and effects. If we had just focused on perceptions of national-level threat from terrorism or immigration, for example, we could draw conclusions about the influences of authoritarianism or mortality salience that would not be valid when considering perceptions of the breadth of threats, perceptions of other threats such as economic threat, or perceptions of threat at other levels such as the global level. With the survey data, we began by looking at the breadth of threats at different levels, where previous research provides us with little guidance about what to expect. As we suggested, one could hypothesise that individuals will identify more security threats that are ‘closer to home’, such as burglary and knife crime, rather than potential threats to the state such as from a military crisis abroad, or that individuals ‘compartmentalise’, identifying just one or two key threats at each level. But neither turned out to be correct: instead, respondents identified most threats at the global level, and progressively fewer on average from national to community to individual level. It was also the case that individuals who identified more threats at one level tended to identify more threats at others. There was, however, some difference in the kinds of threats perceived at the sociotropic (global and national) and subnational (community and personal) levels, with

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

92

Everyday security threats

terrorism and religious extremism more prominent in the former and burglary, knife crime, and online fraud in the latter. When we examined the influences on these perceptions of the breadth of threats at different levels, we uncovered strong effects of mortality salience, watching more current affairs television, and age, at all levels – people who are older, more conscious of their mortality and are exposed to more information about the state of the world also see more threats everywhere. The effects of mortality salience, in particular, suggest that its influence is more pervasive than previously considered; it may not be confined to mortal physical danger and threats to cultural worldview.24 Similarly, the effects of authoritarianism were intriguing, indicating an expected tendency for authoritarians to see more threats but also showing that this tendency only applies to national and subnational threats; authoritarianism has no influence on perceptions of the breadth of global threats. Thus authoritarians are indeed chronically threatened, but our analysis suggests that in terms of the sheer number of threats identified, this does not extend beyond national borders. This was true even when we confined our analysis to perceptions of the breadth of Tier One threats in the NSS, where we might expect authoritarianism to be a stronger influence because these are global threats as defined by those in power. Finally, we demonstrated that individual identity as a citizen of the world, the nation or a subnational area appears to have some influence on perceptions of the relative breadth of threats at each level but the effects were weak and varied in direction. In the second part of the chapter, we shed further light on the influences on threats by focusing on perceptions of the four specific threats of terrorism, immigration, the economy, and environmental degradation. Our interest here was in examining the interaction of the characteristics of different threats and the characteristics of individuals. We tested several hypotheses and found support for each of them. We showed that there are systematic influences on the likelihood that an issue is identified as a threat, that these span dispositions such as authoritarianism and mortality salience, education, and, to a lesser extent, media exposure, and that there is systematic variation in the direction and statistical significance of these influences. Thus mortality salience is associated with identifying threats from terrorism and immigration but not with threats from the economy or the environment. This is in keeping with previous research on the kinds of concerns that render mortality salience influential, so it is worth reiterating that individuals higher in mortality salience are likely to perceive more threats at all levels. In contrast, authoritarianism is associated with threats from terrorism and immigration but also with a significantly lower likelihood of identifying threats from the economy or the environment, explaining the absence of a relationship between authoritarianism and perceptions of the breadth of global threats – authoritarians are significantly less likely to see issues such as the economy and

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

93

environmental degradation as security threats. Education had the effects we expected based on the theory that its effects on perceptions derive from base rate information in that more education is associated with a lower likelihood to see immigration as a security threat but a higher likelihood to see environmental degradation as a threat. Thus, this chapter has shown that there are systematic influences on perceptions of threats. We have also demonstrated that the binary distinction between national (sociotropic) and individual (personal) level considerations is too narrow; global and community-level considerations are additional, independent concerns. While the origins of threats at these different levels have some common causes such as authoritarianism, in general community and individual-level threats have fewer systematic causes according to our analysis. We now turn in Chapter 4 to consider the consequences of perceptions of security threats. Notes 1 Thus, the x-axes from one to seven threats extend to roughly the average number of threats identified at the global level but to the outer edges of the distribution for community and personal threats. 2 There is somewhat more change in the rankings of community and personal threats, but the number of respondents identifying more than two threats at these levels becomes small, i.e., the greater instability appears to be due to small numbers of respondents. 3 With dichotomous variables – each issue was identified as a threat or not at each level – we factor analysed the tetrachoric correlation matrix. We rotated the factor loadings. 4 We also asked about threats in the future. These differed little in number or content from those that respondents viewed as current threats, although there was a slight tendency to identify more threats in the future. 5 We describe how these variables are operationalised in Chapter 2. 6 We consider community and personal threats to be ‘non-sociotropic’ or ‘subnational’ and sometimes refer to them both as ‘personal’ for shorthand. 7 All estimates assume Poisson distributions for the breadth of threats, using Stata 13’s gsem command, because they are counts that are bounded at zero with a long right-hand tail. gsem models do not provide estimates of model fit (you can compare models with different restrictions – see note 21). However, alternative models that do not assume Poisson distributions, but that transform the counts with natural logs, or that ignore their skewness altogether, provide similar results with identical substantive implications. More importantly, they also give estimates of model fit and they indicate excellent fit to the data. Results of these models are available from the authors on request. 8 Although the effect does not quite reach conventional levels of statistical significance at the community level (p = 0.07).Wald tests of the difference in coefficients also show that the size of effects is greater on global and national threats (at p < 0.05) than on community threats and, for global, greater than on personal threats. However, fewer threats are identified at the community and personal levels. 9 One way of looking at this is to separate respondents who named the ‘physical’ dangers of knife crime and burglary from those who did not. When we did this, the difference in

94

10

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

11

12 13

14

15 16 17

18

19

20

21

Everyday security threats the effects of mortality salience between the two groups of respondents was not statistically significant; indeed, the coefficient was larger for respondents who did not identify these dangers. None of these results or conclusions is changed by additional controls for left–right attitudes as represented by party identification. This analysis is available from the authors on request. More than two-thirds of respondents see the same number of threats at the community and personal levels. Thus the results we present in Table 3.3 for the community level are almost identical for perceptions of threats at the individual level. To repeat, the analysis is limited to community-level threats because the results for personal threats are essentially the same. The reason for including a variable for ‘don’t knows’ is not because we have substantive interests or expectations regarding these respondents but in order to have a clean comparison of the effects of seeing oneself as a citizen of the nation or a subnational community with seeing oneself as a citizen of Europe or the world. The results in Table 3.4 are based on analysis of all 1,903 respondents who answered the questions about security threats. There are extremes in the distributions of the dependent variables, however, e.g., respondents who perceived twenty more security threats at one level than another. We checked the sensitivity of our analysis to these outliers by re-estimating models in which we eliminated respondents for whom differences in comparative perceptions of threats were more than two standard deviations from the mean. They did not alter the results. Within the limits that we discussed in Chapter 2, e.g., that these data were gathered at a single point in time and represent a snapshot of opinion. The social disorder that followed the global financial crisis in countries like Greece and Spain was well covered by the British media. Perceptions of threats may themselves be linked. For example, economic threat could be connected to the perception that immigration is a threat. We discuss these relationships below. This could be seen as contrary to the findings of Fritsche et al. (2012). Shaffer and Duckitt (2013) also find a small, positive, bivariate relationship between authoritarianism and their factor of ‘environmental and economic threats’ but this disappears with the addition of control variables. We agree with Fritsche et al.’s argument that it is unlikely that causation runs from perceptions of environmental degradation to authoritarianism, in which views of issues such as climate change inform authoritarian attitudes – and our models reflect this. Rather, authoritarians are simply less likely to be troubled by environmental issues. But if one manipulates the threat, as Fritsche et al. (2012) did in their experiments, this relationship may change. Wildavksy and Dake’s (1990) distinction between cultural biases towards ‘hierarchy’ rather than ‘individualism’ is similar to the biases towards conformity to approved social groups of authoritarians. They argue that individualists more than hierarchists will have economic growth at the forefront of their concerns. With respect to immigration, education may also pick up effects of tolerance and of differences in labour market competition from the less highly educated but the predicted effect would be the same. The table does not include data on model fit because the estimates are from logit models (using gsem in Stata 13) as all the endogenous variables are dichotomous – respondents

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



The scope of security threats and their causes

95

did or did not identify an issue as a threat at each level. However, maximum likelihood models that ignored this fact: 1) indicated that all the models had an excellent fit to the data (the comparative fit index, root mean square error of approximation, etc.), and 2) provided pretty much identical estimates to the logit models in terms of statistical significance, substantive implications, and simulated maximum effects. 22 A brief note on the control variables. Religiosity has few strong relationships with threats, but where we see them they are associated with reduced threats from terrorism and immigration. This is similar to the pattern of results for Muslims, although those relationships with threat tend to be stronger statistically. The results confirm that, when sex matters, it is women who are more likely to see threats, and that older individuals are more threatened in general. Being white versus non-white appears to have little impact. 23 As a check, we estimated models for the other seventeen threats to see the other threats for which authoritarianism has positive or negative relationships at the global level (there are seventeen rather than eighteen because ‘environmental degradation’ combines two threats). In addition to terrorism and immigration, authoritarianism is associated with a greater probability to identify border control, online fraud, knife crime, and burglary as global threats. There are, however, five global threats with which authoritarianism has a negative relationship, as well as the economy and the environment: religious extremism, Islamophobia, the far right, an international military crisis, and severe disruption of critical infrastructure. It is also noteworthy that while the positive relationships are certainly connected to insecurity, uncertainty, and danger, by the same token the issues that authoritarians do not care about at the global level seem broader than Duckitt and Sibley’s (2010) competition, resource scarcity, and inequality, extending to religious tension and a military crisis beyond national borders. 24 We were curious about which threats accounted for the tendency, shown in Table 3.2, for those more aware of their mortality to identify more personal threats. We looked at which of our threats showed a statistically significant effect of mortality salience at the personal level. They were: knife crime, burglary, a health pandemic, an international military crisis, and Islamophobia. Clearly some, but not all, of these represent physical dangers or threats to cultural worldview.

4

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Security threats and their consequences

Introduction We have argued that despite the prominence of the public’s putative role in contemporary security apparatuses, little is actually known about public attitudes towards the full spectrum of security threats. In this chapter, we examine the connection between the kinds of security concerns on which we have focused thus far – the overall breadth of threats and the four specific threats of terrorism, immigration, the economy, and environmental degradation at the global, national, community, and personal levels – and political attitudes and behaviours. While there are other reasons to understand the origins of perceptions of security threats, the issue becomes of less political import if these perceptions do not lead to the kinds of compromises in democratic citizenship, such as the desire for reductions in the civil liberties of outgroups, that previous research has established typically occur under conditions of threat. As in Chapter 3, in order to be able to understand security threats in this chapter more fully than heretofore we look at a range of potential impacts. Unlike in Chapter 3, however, in this chapter we draw exclusively on the survey data – not just on stereotypes of outgroups, for example, but also on voting behaviour and preferences for public spending. The analysis demonstrates inter alia that perceptions of the breadth of threats at different levels have systematically different consequences. For example, global and national threats are strong influences on voting intention, while community threats are not and individual threats are actually a negative influence. On the other hand, perceptions of community threat are a strong influence on the desire for strict criteria for citizenship and on attitudes towards immigrants. Previous research has speculated on why there is variation in such influence but none has examined the range of threats and outcomes that we do. As will become evident, however, the overriding message of this chapter is not that by examining the full spectrum of threats we can identify ‘the consequences’ of threat perceptions; quite the opposite,



Security threats and their consequences

97

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

while we identify some patterns, we illustrate that the considerable variation across choice of dependent variable, type of threat, and the level at which it is perceived, renders suspect any general claims about the consequences of threat perceptions. Measurement The range of effects on which we focus is broad, both because this reflects previous research on specific threats and their effects (e.g., Canetti-Nisim et al., 2009; Duckitt and Fisher, 2003; Huddy et al., 2005; Kam and Kinder, 2007; Landau et al., 2004; Merolla and Zechmeister, 2009), and in order to go beyond the vague notion that perceptions of personal threat matter here and there (Lau and Heldman, 2009). No single study of which we are aware has looked at the range of potential consequences of security threats on which we focus. Based on previous research, we examine the relationship between threats and four kinds of attitudes and (intended) behaviour.

Aggression and intolerance

Perceptions of threats may make individuals more defensive and protective of their identity and worldviews, hence the finding that after 9/11 individuals became more willing to compromise civil liberties in favour of enhanced security measures (Sanquist et al., 2008). Thus, perceptions of greater threat may lead individuals to favour policies such as severe punishment of those who break laws or harbour attitudes such as intolerance of disliked groups. We examine four indicators: a measure of punitiveness based on extent of agreement with four statements about the treatment of criminals (see Appendix for details), such as, ‘Violent criminals deserve to be deprived of some of their human rights’; the treatment of terrorists and whether it is right or wrong for the UK government to authorise the killing of a terrorist in a foreign country (54 per cent said it was right) or to authorise the killing of a terrorist in the UK (55 per cent said it was right). Finally, our measure of tolerance is adapted from a World Values Survey question about groups a respondent would not like to have as neighbours. We changed this to refer to groups that ‘the people you know where you live’ would not like to have as neighbours, under the assumption that social desirability would be less likely to inhibit responses but that we would still be tapping into the respondent’s own intolerance of groups. The average number of groups identified was between two and three, with a majority of respondents naming ‘drug addicts’, and the next four most frequent answers being ‘asylum seekers’, Muslims, immigrants from Africa, and immigrants from other parts of the EU.1

98

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Views of outgroups and minorities

Another manifestation of defensiveness in the face of security threats is negative views of outgroups such as immigrants and cultural and religious minorities (Green, 2009; Kalkan et al., 2009). We focus on four categories of perceptions. The first is cognitive stereotypes of non-white minorities and immigrants, specifically Muslims and East Europeans, based on six seven-point semantic differential scales, such as lazy–hardworking, and violent–peaceful. The most negative stereotypes would be a group seen as lazy, untrustworthy, unintelligent, violent, racist, and uncaring. While we asked these questions about whites, blacks, and homosexuals too, we concentrate on Muslims and East Europeans because both have been the subject of adverse publicity in the UK over the last decade (Nickels et al., 2012), but whereas the negativity about Muslims has pertained to terrorism and the perceived failures of multiculturalism, for East Europeans it has been about ‘stealing’ jobs or business from ‘British people’. Thus, if perceptions of threats lead to a general antipathy towards minorities, we might expect both groups to be stereotyped regardless of whether the focus is the threat from terrorism, immigration, or the economy. But if it is more specific than that, we should see variation. The second measure looks at the same two groups in terms of affect rather than cognition: the extent to which they are associated with five emotions such as ‘fear’ and ‘hatred’ rather than ‘calm’ and ‘warmth’, drawing on the notion that affective information can be at least as powerful as cognitive (Marcus et al., 2000). The other two categories look at slightly different attitudes towards minorities and outgroups, in terms of: 1) what should be the key requirements for British citizenship; and 2) attitudes towards immigrants. First, we asked about the importance of an array of potential criteria for citizenship, such as ‘Being committed to the way of life in the UK’, ranging from education and work skills to race and religion. Exploratory factor analysis revealed two strong factors, one that we label ‘skills’ – having needed work skills, education, and being able to speak English – and another that we call ‘culture’, which includes being Christian and being white (it also includes being wealthy).2 We combined these into separate indexes of the importance of skills and culture for citizenship, with higher scores on the index representing views that these criteria are particularly important. Our approach to immigration was similar in that we factor analysed answers to a series of agree–disagree statements about immigration and immigrants, such as that ‘Letting immigrants into the UK significantly increases the threat of terrorism’,3 again finding two strong factors, one primarily focused on the effects of immigrants on the economy (but also on their possible impact on terrorism), and the other focused on culture and multiculturalism. We combined the statements with the highest loadings on these factors into indicators of concerns



Security threats and their consequences

99

about the impact of immigration on the economy and terrorism, and about the impact of immigration on the UK’s culture.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Policy preferences

Perceptions of threats have been shown to affect behaviour designed to mitigate them, such as precautions taken when travelling (Huddy et al., 2002). The same logic suggests that perceptions of threat might also affect spending preferences for policies, and/or the willingness to pay more in tax in areas that are designed to mitigate security threats. Spending and taxation are not identical: an individual may wish to see existing government spending shift towards the mitigation of security threats but also be entirely unwilling to pay any more tax towards it – their preference would be satisfied by the reduction or elimination of spending elsewhere. We examined these issues in two ways. First, we gauged budgetary preferences by giving respondents an imaginary budget of £100 to be apportioned between twelve policy areas: fighting terrorism, border control, education, defence, the National Health Service, the environment, transport, paying off the national debt, work and pensions, international aid and development, and culture, media and sport. In our analysis we look at their allocations for combined spending on terrorism and border control, which we label ‘borders/crime’, education, and international aid. We do not expect a positive relationship between perceptions of threat and all of these; our interest is precisely in whether threat is associated with preferences for more spending in some areas and with less spending in others, as opposed to indifference. To gauge preferences for taxation, we asked respondents the extent to which they would support raising taxes by £50 a year for more spending on border security, defence, and the police – small increases but ones that divided respondents.4 Voting

Previous research in the US has indicated that the perception of threat from terrorism in particular leads voters to look for leaders with attributes such as strength and charisma (Merolla and Zechmeister, 2009). We examine the extent to which this appears to apply also to potential voters in the UK for terrorism and other threats. Respondents ranked eight attributes of leaders in order of importance to them, such as ‘relates to ordinary people’, ‘tells the truth’, and ‘strong on security’. We look at attributes in terms of how often they were ranked among the two most important for a leader to possess: ‘strength’, which we base on a combination of ‘decisive’, ‘strong on security’, and ‘strong on preserving British values’ (49 per cent ranked these among the two most important), and ‘tells the truth’, which was ranked among the top two attributes by 57 per cent

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

100

Everyday security threats

of the sample. Clearly, the first of these attributes is designed to assess the relationship between perceptions of security threats and the desire for a strong leader. We analyse the second to see whether a positive relationship between perceptions of threat and the desire for a strong leader is also accompanied by de-emphasising other attributes such as truth telling. With intention to vote we are tapping into the notion that threats heighten the stakes in an election. For threatened individuals, the perceived consequences of not voting may be higher than for individuals who are not threatened. We base vote intention on a scale combining self-reported certainty of voting in the next general election and whether the respondent named the party they expected to vote for, with ‘most certain’ being a respondent both expressing certainty to vote and naming the party he or she expected to vote for, next most certain being a respondent expressing certainty to vote but being unable to name the party he or she expected to vote for, and least certain being a respondent expressing something less than certainty that they would vote at the next general election. All the dependent variables are coded from 0 to 1 with the exception of the numbers of groups respondents claimed that people in their area would not want as neighbours, where we just take the number, and spending preferences for the £100 budget, for which the coding is in pounds. Expectations As in Chapter 3, we examine both the impact of perceptions of the breadth of threats and also of the specific security threats of terrorism, immigration, the economic crisis and environmental degradation. We are interested in several issues: Variation in the consequences of perceptions of the breadth of security threats

Previous research suggests that we should see influences of perceptions of security threats in all four of the areas under examination, but one of the questions that arises is whether all threats are alike in their consequences. The fact that Chapter 3 has shown some variation in their origins suggests that they will not be. The question is how. With perceptions of the breadth of threats, our interest is in whether individuals who identify more threats are also more aggressive, more likely to stereotype and denigrate minorities and outgroups, more likely to favour spending and taxation addressed at securing borders, more likely to favour a leader’s strength over other potential attributes, and more likely to be motivated to vote at the next national election. Chapter 3 has suggested that perceptions of more global threats, for example, may reflect a somewhat different orientation to the world



Security threats and their consequences

101

than perceptions of more national threats and that we might therefore see less illiberal consequences. But only by looking at this empirically can we be more certain.

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Variation in the consequences of perceptions of specific security threats

For the specific threats of terrorism, immigration, the economy, and environmental degradation, we suggested in Chapter 3 that variation in their origins derives from the interaction of individual-level characteristics and the characteristics of the threat. We described the threat of terrorism as pertaining to physical danger and disruption. Immigration, on the other hand, raises issues of economic threat (Sniderman et al., 2004), identity, and culture, and perhaps – where there is tension between immigrants and long-term residents – of the physical safety that we showed in Chapter 3 to be the leading characteristic of individuals’ definition of ‘security’. Security threats from environmental degradation, which pertain to damage to the physical world and indirectly to livelihood, and from economic threat, which is more ambiguous in its implications because it could encompass both social order concerns and considerations such as inequality, do not seem to have the same implications for physical safety. There are two main possibilities with respect to the consequences of perceived threats from these issues. One is that the psychological experience of threats is similar regardless of the issue, i.e., terrorist threat evokes anxiety and stress just as threat from the environment does. Notions that ‘threat’ thus leads individuals to defend their cultural worldviews may imply that both would lead to the denigration of outgroups, even though the characteristics of terrorist threats differ from those from environmental degradation. This seems unlikely, however. We have already demonstrated that authoritarianism raises the probability of identifying threats from terrorism and immigration but reduces the probability of identifying threats from the economy and the environment. We might also expect that the characteristics of threats will interact with the considerations that the consequences raise, e.g., that perceptions of threat from terrorism will have consequences for punitiveness and aggression while immigration will be strongly associated with concerns about cultural conformity as a requirement of citizenship. In contrast, perceptions of threat from environmental degradation should not be associated with these kinds of concerns because there is no relationship to the threat itself but might, for example, affect other variables such as spending preferences or the desire for a strong leader. Sociotropic versus personal threats

As we outlined in Chapter 1, previous research has largely concluded that sociotropic – by which is usually meant national – threats are far more important

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

102

Everyday security threats

drivers of political attitudes and behaviours than subnational threats. If this conclusion is valid, it should apply across the range of threats we examine because the logic of the clarity of responsibility argument discussed in Chapter 1 should apply not only to the economy but also to other issues of concern to individuals. As Huddy et al. (2005) suggest, however, one of the reasons why sociotropic considerations may appear more influential could be the tendency for researchers to focus on national-level, non-specific, concerns such as elections. Where there are specific policies that can mitigate a physical threat, for example, personal considerations are more salient (Huddy et al., 2005; Maoz and McCauley, 2009; Schildkraut, 2009). Similarly, clarity of responsibility and other theorised mechanisms for sociotropic effects appear less central to consequences of threats such as stereotyping minorities or a desire for more spending on border control. Stereotypes are unlikely to be linked to government policy as strongly, if at all, as voting behaviour. Rather, stereotypes are likely to be influenced by direct experience, such as contact with members of a minority group, or by indirect experience such as media exposure suggesting demographic change at the national or local level (Hopkins, 2010), which may foster the defensiveness and hostility towards outgroups that structures negative stereotypes. Indeed, theories of ‘realistic group conflict’ (e.g., Bobo, 1983) refer to subnational conflict between groups and the threat of changes in group relations. In addition, individuals who are more threatened are disposed to defend their cultural worldviews and identity and it is likely that ethnic and racial considerations are central components in that defence via the desire to reduce outgroups’ power. This is not to say that sociotropic concerns will not matter also – physical danger to the nation or perceived threats to national identity from immigrant minorities are likely to be important – but that the relative influence of subnational considerations should be stronger where their implications are also easier to discern than they are for voting intention. Analysis Our analysis follows the same format as Chapter 3 in that we estimate structural equation models. But we now add an additional path from perceptions of global, national, community, and personal security threats to outcome variables such as vote intention. While our primary interest in the exogenous variables such as authoritarianism and mortality salience is their relationships with the perceptions of threats we examined in Chapter 3, we also include them as control variables in the second stage of analysis because, for example, authoritarianism affects stereotyping (Kalkan et al., 2009; Kam and Kinder, 2007). We leave open the question of whether its influence is entirely mediated by threat (see CanettiNisim et al., 2009), estimating the direct and mediated (through threats) effects

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document



Security threats and their consequences

103

of all the exogenous variables. In sum, we allow variables such as mortality salience and authoritarianism to have a direct impact on measures like the attributes necessary for citizenship and an indirect impact via their effects on perceptions of threats, a conservative approach to gauging the effects of perceptions of threats. Before we proceed, it is also necessary to explain our approach to statistical significance in the tables that follow. We conduct two kinds of tests. The first is the straightforward test of statistical significance of a coefficient from zero, indicated by a *(p < 0.05) or # (p < 0.10) in the tables. The second is a Wald test of the statistical significance of differences in coefficients, indicated by a g, n, c or p. These symbols indicate differences between the effects of perceptions of global (g), national (n), community (c), or personal threat (p) (at p < 0.10). These tests allow us to explore differences in the impacts of perceptions of threats accounting for the uncertainty surrounding the estimated relationships. Individual estimates of effects may differ in sign and/or statistical significance from each other but we cannot be confident that the differences are statistically significant without the Wald tests. By the same token, our Wald tests will sometimes show statistically significant differences in the influence of two threats at different levels, although the estimated effect of each threat is not significantly different from zero – there is uncertainty surrounding the estimated influence of each threat but greater statistical certainty that their influences are at least different from each other. We display the direct effects of variables such as education and age in the tables for breadth of threats, but not their indirect effects via their influences on perceptions of threats (this was our focus in Chapter 3 and the additional paths make little difference to those estimates). To briefly summarise their effects: authoritarianism influences all the dependent variables except voting intention, and, as would be expected, is associated with aggression, intolerance, stereotyping, and negative emotions towards minorities and immigrants, the desire for more spending on securing the nation’s borders and less spending on education and international aid, a willingness to pay more tax towards border security, defence and policing, and valuing a leader who is strong but not a leader who tells the truth. That these relationships echo conventional wisdom about the effects of authoritarianism is reassuring about the construct validity of our indicators of potential consequences of threats. The impacts of other variables are less consistent in sign and statistical significance. Education and being a woman tend to be associated with more liberal attitudes towards civil liberties, and age and watching more current affairs on television with the opposite. The effects of mortality salience vary but are associated most strongly with cultural concerns about immigration and preferences for stricter cultural conformity as a requirement for citizenship, which is also both confirmatory of previous research and suggests that our measure of mortality salience is valid.

104

Everyday security threats

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Consequences of perceptions of the breadth of threats

We begin in Table 4.1 with the relationships between perceptions of the breadth of threats and aggression and intolerance. The results confirm that perceiving more threats affects these attitudes – seven of the sixteen estimates are statistically significantly at p < 0.10 and all are positive, implying that identifying more threats is associated with a greater willingness to sacrifice the civil liberties of criminals and with more intolerance of disliked groups. However, the salient threats are as likely to be subnational as sociotropic. This suggests that Huddy et al. (2002) are right that subnational considerations are more influential on attitudes towards specific policies that may mitigate threat, but indicates both that this also applies to individuals who simply perceive more subnational threats of any kind and also that it extends to intolerance. There is additional variation in the influence of perceptions of threats at different levels. Most of these differences are between sociotropic – global or national – and subnational – community or personal – threats, particularly between global and community-level threats. They suggest that perceptions of more global threats do not lead to such high levels of aggression and intolerance as perceptions of more threats to one’s community. Thus, as we suggested in Chapter 3, perceptions of more global threats appear to reflect a somewhat different outlook than perceptions of national and subnational threats. Table 4.2 extends the analysis to views of immigrants and minorities. Of the thirty-two estimated relationships, twelve are statistically significant from zero (at p < 0.10), indicating that perceptions of the breadth of threats are an influence but not an overwhelming factor in these attitudes. We also see, however, that there is both a clearer influence of subnational threats on these attitudes – seven of the twelve statistically significant relationships derive from perceptions of community or personal threats – and of differences between sociotropic and subnational threats. Indeed, of the thirteen statistically significantly different relationships between levels of threats, all are between sociotropic and subnational threats. They indicate that perceptions of personal threats are particularly influential on stereotypes and negative affect towards Muslims and East Europeans – we do not see much variation between them, despite what we suggested are differences in the kinds of threat that Muslims and East Europeans appear to represent – while community threats are particularly influential in raising the bar for requirements for citizenship and in broader concerns about the impact of immigration. The differences with sociotropic threats indicate that perceptions of global threats are less strongly associated with stereotypes and negativity towards immigrants and minorities than perceptions of community and personal threats. Our analysis of the principal threats identified by survey respondents, displayed in Figure 3-2, showed differences between sociotropic and subnational

Downloaded from manchesterhive © Copyright protected It is illegal to copy or distribute this document

Table 4.1  Perceptions of breadth of threats and aggression and intolerance Punitiveness

Right for the UK to authorise killing of a terrorist in a foreign country

Right for the UK to authorise killing of a terrorist in the UK

Number of groups would not have as neighbours

Global threats National threats Community threats Personal/family threats Authoritarianism Mortality salience Time spent watching current affairs TV Educated to degree level or higher Woman Age Muslim Religiosity White Constant

−0.002 (0.001)np 0.005 (0.002)*gc −0.002 (0.003)n 0.005 (0.003)#g 0.213 (0.015)* −0.026 (0.026) −0.011 (0.018) −0.044 (0.009)* 0.013 (0.009) 0.062 (0.021)* 0.018 (0.021) −0.047 (0.017)* 0.050 (0.018)* 0.478 (0.027)*

−0.005 (0.014)c 0.023 (0.020) 0.077 (0.032)*g 0.019 (0.029) 1.088 (0.161)* −0.775 (0.277)* 0.522 (0.190)* −0.178 (0.101)# −0.414 (0.100)* −0.130 (0.225) −0.727 (0.218)* −0.453 (0.178)* 0.462 (0.188)* −0.370 (0.281)

0.008 (0.014)c 0.016 (0.020) 0.078 (0.032)*g 0.027 (0.029) 1.272 (0.162)* −0.961 (0.277)* 0.480 (0.189)* −0.236 (0.101)* −0.362 (0.100)* −0.431 (0.266)# −0.359 (0.215)# −0.463 (0.177)* 0.542 (0.188)* −0.411 (0.281)

0.007 (0.004)#nc 0.023 (0.006)*gcp 0.042 (0.008)*gnp 0.001 (0.008)nc 0.368 (0.049)* 0.159 (0.083)# 0.057 (0.057) −0.078 (0.031)* −0.089 (0.030)* 0.313 (0.068)* −0.007 (0.074) −0.216 (0.057)* 0.304 (0.064)* 0.084 (0.092)

Log likelihood

−18609.526

−20248.243

−20248.936

−22928.116

Notes:  n = 1903; * p < 0.05; # p < 0.10 (two-tailed test). Darker shading = sociotropic threats, lighter shading = personal threats. g different from global threat at