! THE!LONDON!RIOTS!AND!THEIR!IMPACT!ON!ATTITUDES! TOWARDS!SAFETY!AND!SECURITY! TILMAN!HISARLI!

advertisement
!
!
THE!LONDON!RIOTS!AND!THEIR!IMPACT!ON!ATTITUDES!
TOWARDS!SAFETY!AND!SECURITY!
!
!
!
TILMAN!HISARLI!
!
!
ECONOMICS!UNDERGRADUATE!DISSERTATION!
!
!
ABSTRACT!
!
Based& on& data& from& the& British& Social& Attitude& Survey& 2011,& the& Metropolitan& Police&
Attitude& Survey& 2011& and& the& Home& Office,& the& author& examines& whether& the& London&
riots& (2011)& affected& attitudes& towards& security& and& safety& within& the& UK.& The&
perception& of& personal& safety& was& temporarily& lowered,& reaching& a& trough& within& the&
month& following& the& riots,& but& returning& to& pre& riot& levels& thereafter.& Exposure& to& riot&
crime& per& se& did& not& appear& to& significantly& affect& security& and& safety& attitudes.& In&
conjunction&with&the&extraordinary&time&path&of&support&levels&for&stiffer&sentences,&this&
leads& us& to& believe& that& the& media& was& mainly& responsible& for& the& public’s& desire& for&
enhanced&safety.&Stiffer&sentencing&for&riot&perpetrators&was&heavily&lobbied&for&by&the&
media,& and& support& peaked& within& one& month& following& the& riots,& remaining& at&
significantly&elevated&levels&afterwards.&
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
&
&
!
!
TABLE!OF!CONTENTS!
CHAPTER!I! INTRODUCTION!
3&
CHAPTER!II! CHANGES!TO!SAFETY!&!SECURITY!ATTITUDES!!
5&
ATTITUDES!AND!ATTITUDE!ASSESSMENT!
THE!GENERAL!VARIABLES!OF!SAFETY!AND!SECURITY!ATTITUDES!
GENETIC&FACTORS&
DIRECT&PERSONAL&EXPERIENCE&
PARENTAL&INFLUENCE&
GROUP&DETERMINANTS&
WHAT&IS&MISSING&
THE!IMPORTANCE!OF!HIGH!IMPACT!EVENTS!
TERRORISM&
INTERNATIONAL&AFFAIRS&SHOCKS&
RIOTS&
THE!LONGEVITY!OF!SECURITY!ATTITUDE!CHANGES!AFTER!EXTRAORDINARY!EVENTS!
THE!SIGNIFICANCE!FOR!OUR!STUDY!
5&
6!
6&
6&
6&
6&
7&
7&
7&
9&
10&
11&
13&
CHAPTER!III! TESTING!THE!THEORY!THROUGH!TIME!DIFFERENCING!
14&
THE!DATA!AND!METHOD!
THE!RESULTS!
14&
17&
CHAPTER!IV! EXAMINING!THE!CAUSAL!EFFECT!OF!THE!LONDON!RIOTS!
19&
THE!DATA!AND!METHOD!
THE!RESULTS!
A!CASE!FOR!THE!MEDIA!
19&
20&
22&
CHAPTER!V! CONCLUSION!
25!
BIBLIOGRAPHY!
26!
APPENDIX! CHAPTER!IV!REGRESSION!OUTPUT!USING!LPM!
30!
&
!
!
!
!
!
!
!
!
!
2&
&
Chapter!I!
&
Introduction!
&
The& public& disorderly& events& commonly& known& as& the& London& riots& represented& a&
massive&shock&to&public&life&and&safety.&There&is&a&large&body&of&literature&that&suggests&
that&such&highly&upsetting&events&have&an&effect&on&people’s&attitudes.&In&this&study,&we&
look& at& whether& the& London& riots& affected& attitudes& towards& security& and& safety.& This&
will& help& us& to& understand& whether& government& spending& should& be& used& to& react& to&
future&disruptions&to&ensure&the&population’s&happiness&and&wellbeing.&
&
The&London&riots&shook&the&UK’s&capital&from&the&6th&until&the&11th&of&August&2011&and&
are&often&labelled&the&“worst&unrest&since&race&riots&set&the&capital&ablaze&in&the&1980s”&
(Huffington&Post,&2011).&This&may&come&as&a&surprise&to&the&reader,&as&the&latter&spanned&
a& time& period& of& four& months,& while& the& 2011& riots,& with& a& duration& of& less& than& week,&
were&hardly&perennial.&Even&The&Guardian,&however,&frequently&links&the&London&riots&
to& the& “disturbances& two& decades& earlier”,& so& looking& at& the& crime& statistics& might& help&
our& understanding:& while& approximately& 70%& of& all& riotbrelated& crimes& (Home& Office,&
2011)&were&recorded&in&the&Greater&London&area,&relevant&criminal&behaviour&could&also&
significantly& be& observed& in& cities& like& Manchester,& Birmingham& or& Liverpool.& As& such,&
rather&than&spreading&evenly&through&the&country,&the&2011&riots&flared&up&in&individual,&
spatially&confined&regions&that&were&often&geographically&disjointed.&Even&within&Greater&
London,& there& was& significant& heterogeneity& in& riot& intensity,& with& 430& recorded&
incidents&in&Croydon,&and&less&than&50&in&Kensington&and&Chelsea.&According&to&Reuters&
(2011),&the&Association&of&British&Insurers&estimated&overall&damages&at&approximately&
£200m.&The&Guardian&reported&of&roughly&4,000&arrests,&and&crimes&spanned&across&the&
entire& spectrum,& ranging& from& arson& to& murder.& While,& given& these& figures,& the& media&
quite& rightly& condemned& the& perpetrators,& one& cannot& help& but& notice& the& often& times&
flamboyant&rhetoric&used&by&some.&The&Telegraph&(2011),&for&instance,&demanded&that&
“the& thugs& [criminals& who& shame& our& nation]& must& be& taught& to& respect& the& law& of& the&
land&the&hard&way”.&The&New&York&Daily&News&(2011)&even&went&so&far&as&to&name&and&
shame&“young&British&people&(…)&among&the&most&unpleasant&and&potentially&violent&(…)&
in&the&world”.&This&should&make&us&wonder&whether&the&London&riots&were&actually&as&
atrocious&as&these&newspapers&claimed.&&
&
Amidst& the& most& captivating& statements& was& The& Sun’s& (2011)& depicting& the& riots& as&
“anarchy& (…)& and& a& serious& threat& to& life”.& Quite& rightly& this& identifies& the& threat& to&
existence& as& the& riots’& most& unsettling& feature.& Therefore,& in& order& to& truly& assess&
whether&the&London&riots&indeed&traumatized&British&society,&one&must&look&at&whether&
the&disorders&had&an&effect&on&how&people&feel&about&their&own&safety&and&security.&That&
is& what& we& will& do& in& this& study.& We& want& to& assess& the& impact& of& the& London& riots& on&
people’s& attitudes& towards& security& and& safety.& It& will& enable& us& to& judge& whether& the&
riots& were& truly& as& disquieting& as& the& news& made& them& out& to& be,& or& whether& media&
coverage& was& in& the& spirit& of& “you& never& want& a& serious& crisis& to& go& to& waste”& (Rahm&
Emanuel,&2008).&
&
For& this& reason,& it& may& be& apparent& to& see& why& one& should& more& closely& examine& the&
London& riots;& however,& before& delving& deeper& into& the& topic,& we& owe& the& reader& an&
explanation&for&our&deep&curiosity&about&attitudes.&Stressing&the&importance&of&attitudes&
3&
&
is&by&no&means&a&modern&day&development.&In&fact,&already&in&1830,&Thomas&Macaulay&
pondered:&&
&
“What$are$laws$but$the$expression$of$the$opinion$of$some$class$over$the$rest$of$the$
community?$By$what$was$the$world$ever$governed$but$by$the$opinion$of$some$person$or$
persons?$By$what$else$can$it$ever$be$governed?”&
!
Herein,& he& describes& attitudes& and& opinions& as& the& driving& forces& behind& governing&
peoples,& nations,& and& the& world,& and& by& extrapolation& assigns& these& concepts&
extraordinary& significance& in& shaping& the& world& into& what& it& is& today.& Attitudes& affect&
governing,&governing&is&politics,&and&politics&is&the&framework&through&which&economic&
thought& materializes.& The& close& examination& of& attitudes& should& thus& be& a& repertorial&
exercise&for&the&practical&economist.&&
&
However,&it&is&not&economics&alone,&where&the&study&of&attitudes&is&of&estimable&rank.&For&
Harvard& professor& Gordon& Allport,& who& pioneered& the& field& of& personality& psychology,&
“the&concept&of&attitude&is&probably&the&most&distinctive&and&indispensable&concept&(…)&
in&social&psychology”.&While&this&statement&dates&back&to&1935,&the&eminence&of&attitudes&
has& not& faded,& but,& as& Herbert& Kelman& (1974)& put& it& “if& anything,& become& even& more&
central&in&social&psychology”.&&
&
Yet& the& question& remains,& why& attitudes& are& so& universally& chief.& Oskamp& (1991)&
theorizes& it& to& be& a& function& of& their& wide& applicability,& for& a& single& attitude& can&
summarize&many&disparate&behaviours.&It&is&the&cause&of&an&individual’s&conduct&towards&
her&environment&and&helps&to&explain&consistency&in&actions,&which&in&turn,&according&to&
Allport,&elucidates&the&stability&of&society.&By&analyzing&their&attitudebrelated&effects,&we&
can& therefore& examine& whether& the& London& riots& shocked& the& societal& status& quo&
sufficiently& so& as& to& evoke& core& changes& to& community& behaviour.& Attitudes& are& an&
interdisciplinary& concept,& and& reveal& how& agents& perceive& their& surroundings.& As& such,&
they& steer& our& society,& and& if& we& improve& our& understanding& towards& them,& we& can&
improve&our&reaction&to&future&societal&shocks.&
&
The&following&chapter&(Chapter&II)&will&review&the&existing&literature&around&safety&and&
security&attitudes,&with&a&focus&on&attitude&change&after&extraordinary&events.&Chapter&III&
will& preliminarily& test& the& theory& established& in& the& previous& chapter& using& a& first&
difference&approach&in&a&smaller&data&set,&and&Chapter&IV&will&finally&aim&to&examine&the&
causal&impact&of&the&London&riots&on&safety&and&security&attitudes.&Here,&we&use&a&larger&
data&set&in&a&difference&in&difference&model.&Chapter&V&will&serve&as&a&précis.&&
&
&
&
&
&
&
&
&
&
&
&
4&
&
Chapter!II!
!
Changes!to!safety!&!security!attitudes!following!high!impact!
events!
&
A&key&aspect&in&determining&quality&of&life&is&the&perception&of&safety&and&security.&While&
this&in&itself&would&justify&our&study’s&focus&on&these&two&concepts,&there&are&reasons&that&
go&beyond&the&individual&level.&For&instance,&the&implementation&of&security&and&safety&
through& political& leaders& concerns& a& major& share& of& government& expenditure.& Funded&
through& tax& revenue,& questions& regarding& the& budget& are& ever& so& salient& in& light& of&
recently&amplified&attention&on&the&abuse&of&taxpayers’&money&in&the&Euro&crisis.&Their&
significance&on&both&the&individual&and&aggregate&level&deservedly&puts&the&concepts&of&
safety&and&security&in&the&spotlight&of&this&study.&Change&to&these&attitudes,&however,&is&a&
concept&scarcely&evaluated&directly&in&academic&research.&Due&to&its&novelty,&the&purpose&
of& this& literature& review& is& to& evaluate& whether& methods& employed& in& topically& related&
papers&(i.e.&papers&that&discuss&the&effects&of&high&impact&events)&are&applicable&to&this&
study.& In& our& case,& this& is& a& more& crucial& exercise& than& evaluating& internal& cogency& of&
existing& literature& that& does& not& precisely& address& changes& to& safety& and& security&
attitudes.&Before&doing&so,&however,&we&need&to&uncover&how&attitudes&are&assessed&in&
existing&literature.&
&
Attitudes!and!attitude!assessment!
!
At& this& point,& it& helps& the& reader’s& understanding& to& shed& light& on& the& properties& of&
attitudes&within&the&academic&framework.&As&is&often&the&case&with&what&Skinner&(1957)&
calls& “mentalistic& concepts”,& the& definition& of& attitudes& varies& across& researches& to&
certain&degree&and&does&not&seem&to&be&clearly&defined.&While&Oskamp&(1991)&describes&
them& as& “learned& predispositions& to& respond& favourably& or& unfavourably& towards& a&
given&object”,&Gagné&and&Briggs&(1974)&attach&an&alternate&nuance.&They&define&attitudes&
as&an&“internal&state,&which&affects&an&individual’s&choice&of&action&towards&some&object,&
person&or&event”.&This&latter&definition&excludes&the&blatant&observation&of&attitudes,&but&
promotes&inference&about&them&on&the&basis&of&behavioural&manifestations.&This&is&what&
Aiken& (2002)& defines& as& “overt& methods”,& where& people’s& actions& and& words& are&
assumed& to& be& representative& of& their& attitudes& towards& the& objects& or& people& in&
question.& In& our& study,& we& could& therefore& assess& safety& and& security& attitudes& either&
through&revealing&actions&(e.g.&increased&spending&on&security&technology)&or&interview&
surveys.& In& research,& it& is& common& to& interview& subjects& enquiring& directly& about& their&
attitudes.&This&“explicit&method”&relies,&however,&on&the&respondents’&awareness&of,&and&
willingness&to&reveal&their&attitudes.&Although&presumed&unobtrusive&(enquiry&does&not&
alter&attitudes),&there&is&one&flaw.&In&an&attempt&to&assess&people’s&alleged&attitudes&and&
actions,&La&Piere&(1934)&asked&hotel&managers&whether&they&would&serve&“members&of&
the&Chinese&race”.&While&80%&responded&in&the&negative,&in&fact&almost&all&of&them&did.&In&
short:&people&may&be&unwilling&to&report&their&truthful&attitudes.&As&a&counterbmeasure,&
in&this&study,&we&shall&not&only&steer&clear&from&evaluating&questions&whose&responses&
could&have&a&pejorative&effect,&but&also&take&a&more&concealed&approach.&We&will&not&only&
examine&questions&that&specifically&mention&safety&and&security,&but&also&questions&that&
allow&inference.&Prior&to&evaluating&the&effects&of&upsetting&exogenous&events&on&safety&
5&
&
and&security&attitudes,&however,&we&need&to&understand&their&determinants&in&times&of&
calm.&
&
The!general!variables!of!safety!and!security!attitudes!
!
Applying&a&key&condition&from&Gerard&and&Orive&(1987),&for&attitude&formation&towards&
security&to&occur,&the&individual&must&be&expecting&to&somehow&be&affected&by&issues&of&
safety,& and& feels& the& psychological& urge& to& prepare& herself& for& the& encounter.& So,& in& an&
unstirred& environment,& what& are& the& general& determinants& that& provide& for& the&
heterogeneity& in& safety& and& security& attitudes& that& we& observe& in& everyday& life?&
Understanding&these&factors&will&support&our&evaluation&of&high&impact&events&(we&shall&
look& at& terrorism,& international& affairs& shocks& and& riots),& and& we& will& include& these& as&
control& variables& in& our& regression& analysis& of& the& London& riots& in& Chapters& III& and& IV.&
Therefore&they&are&crucial&for&the&specification&of&our&estimation&strategy.&According&to&
Oskamp&(1991)&there&are&four&applicable&categories:&&
&
Genetic!factors!one&usually&assumes&attitudes&to&be&learned.&However,&Gallup&(2000)&
found&that&women&when&compared&to&their&male&counterparts&“express&greater&fear&of&
crime”,&a&discovery&subsequently&supported&by&Mulvey&(2002).&While&we&believe&it&more&
likely& that& genetic& factors& determine& personal& experiences& that& in& turn& affect& security&
attitudes,& it& is& often& argued& that& genetic& factors& per& se& establish& a& “natural& tendency”.&
Either&way,&a&gender&variable&should&capture&the&discrepancy&between&men&and&women.!
&
Direct! personal! experience!quoting& Fazio& (1988)& direct& personal& experience& is& the&
most& fundamental& way& of& attitude& formation,& and& evokes& particularly& strong& attitudes&
when& compared& to& opinions& based& on& indirect& experience.& Specifically& repeated&
exposure&acts&attitude&enhancing&(Zajonc,&1968).&Therefore,&demographic&factors,&which&
destine& people& to& continuously& experience& certain& safety& and& security& concerns& (e.g.&
crime)&should&be&controlled&for.&Derived&from&victimization&rates,&these&are&age,&marital&
status&and&income,&both&on&an&individual&and&regional&level&(Bureau&of&Justice&Statistics,&
2001).&The&importance&of&repeated&exposure&is&an&interesting&proposition:&would&we&not&
expect&individuals&to&adapt?&Chapters&III&and&IV&will&tell&us&about&this&significance.!
&
Parental! influence!”parents& have& almost& total& control& over& an& infant’s& informational&
input”,& which& Oskamp& (1991)& describes& to& be& similar& to& penal& institutions.& A& decisive&
difference,&however,&is&that&a&child&has&no&prebexisting&experiences&that&may&counter&the&
parental& dictum,& as& opposed& to& criminals& (their& attitudes& are& more& robust).& Racial&
attitudes,&for&instance,&are&almost&directly&adopted&from&the&parents&(Ashmore&and&Del&
Boca,& 1976).& Further,& Sampson& and& Bartusch& (1998)& found& that& for& different& crimes,&
tolerance& levels& varied& across& ethnicities.& We& shall& therefore& include& corresponding&
control&variables.!
&
Group!determinants!this&regards&a&type&of&conformity&pressure&to&reference&groups&
“whose& standards& and& beliefs& one& accepts& and& wants& to& be& associated& with& (Oskamp,&
1991)”.&Pettigrew&(1967)&describes&the&effects&of&group&determinants&to&be&particularly&
strong& in& political& and& authoritarian& attitudes,& and& combined& with& Taylor’s& (1982)&
maxim&of&pluralistic&ignorance&(“others&are&more&prejudiced&than&we&are”)&makes&group&
dynamic& induced& attitudes& particularly& resistant& to& change.& Safety& is& a& function& of&
political& agendas& and& authoritarianism,& and& differently& educated& groups& are& prone& to&
6&
&
have& different& views& on& security.& Where& appropriate,& we& should& thus& control& for&
education&level&and&political&party&association.&We&expect&their&effects&to&be&particularly&
significant&at&the&extremes&(e.g.&for&those&with&very&high&or&very&low&education&or&with&
extreme&ideologies).!
&
What!is!missing!while&the&effect&of&the&media&on&attitudes&in&general&has&been&widely&
accepted,&this&author&finds&it&peculiar&that&no&direct&link&was&made&between&safety&and&
security&attitudes&and&media&exposure&in&the&reviewed&literature.&Given&the&often&times&
lurid&nature&of&boulevard&magazines&should&we&not&expect&media&reports&to&actively&
influence&safety&attitudes?&We&will&tentatively&include&a&corresponding&control&variable&
in&Chapter&III,&and&proceed&from&there.!
&
The!importance!of!high!impact!events!
!
While&the&general&variables&determine&security&attitudes&in&times&of&peace&and&calm,&it&is&
reasonable& to& expect& disturbance& from& these& levels& following& highbpowered& incidents.&
We& define& a& high& impact& event& as& causing& significant& upset& in& the& population,& and& as& a&
result,& it& reduces& perceived& levels& of& safety.& In& fact,& Oskamp& (1991)& ascribes& particular&
importance&to&“traumatic&or&frightening&[events]&in&the&development&of&attitudes”.&How&
can& one& assess& their& effects& on& safety& and& security& attitudes?& As& discussed& earlier,& a&
direct& method& is& to& retrieve& information& via& opinion& and& attitude& surveys;& its& indirect&
counterpart& is& to& observe& behaviour& and& treat& it& as& individuals’& attitude& revelations.&
Gordon& and& Arian& (2001)& have& shown& that& events,& which& severely& diminish& one’s&
perceived&level&of&security,&bolster&the&influence&of&emotions&on&policy&decision&making&
processes.& With& personal& attitudes& at& least& partially& driven& by& emotions,& it& should& thus&
be& possible,& inter& alia,& to& assess& changes& to& safety& and& security& attitudes& via& policy&
choices& (e.g.& increased& military/police& spending).& In& what& follows,& we& will& look& at&
terrorism,& riots& and& international& affairs& shocks,& such& as& wars& or& invasions,& to& unveil&
their&effects&on&individuals’&perception&of&security&and&safety.&It&allows&for&an&evaluation&
of& what& we& should& treat& as& our& dependent& variable,& which& must& be& representative& of&
safety& and& security& attitudes.& This& is& a& crucial& exercise,& as& we& refrain& from& examining&
respondents’&direct&testimony&regarding&their&security&attitudes,&in&order&to&reduce&the&
danger&of&misrepresentation&as&(previously&discussed)&in&La&Piere&(1935).&&
&
Terrorism!
!
The&most&memorable&acts&of&terror&are&arguably&the&9/11&attacks&on&the&United&States,&
the&2004&Madrid&bombings&and&the&London&7/7&attacks&in&2005.&&
&
Analysing&the&effects&of&the&Al&Qaeda&strike&on&the&US&on&11th&September&2001,&Davis&and&
Silver& (2004)& show& that& an& “increased& sense& of& threat& leads& to& greater& willingness& to&
concede&some&civil&liberties&in&favour&of&security&and&order”,&although&there&exists&some&
heterogeneity& as& to& the& effect’s& magnitude& across& different& ethnicities.& Thus,& a& rising&
willingness& to& concede& civil& liberties& should& indicate& more& sensitive& attitudes& towards&
safety&(making&it&a&candidate&for&our&dependent&variable).&Findings&are&based&on&survey&
data& gathered& on& a& postbhoc& basis,& and& are& in& unison& with& opinion& poll& results& from&
Huddy,& Khatib& and& Capelos& (2002).& These& also& show& a& significant& increase& in& the&
percentage& of& Americans& willing& to& “give& up& some& civil& liberties& in& order& to& combat&
terrorism”,&in&order&to&live&safer&lives.&
&
7&
&
An& alternative& evaluation& tool& for& assessing& changes& in& safety& and& security& attitudes& is&
treating& election& outcomes& as& revealed& preferences.& Based& on& varying& strength& of&
counterbterrorism&agendas&across&political&parties,&the&causal&effect&of&a&terror&attack&on&
election& results& can& give& an& indication& of& individuals’& desire& for& increased& national&
security&and&safety.&In&their&analysis&of&voter&sensitivity&to&terrorism,&Berrebi&and&Klor&
(2008),&using&a&difference&in&difference&approach&with&variation&across&space&and&time,&
conclude& that& the& average& terror& attack& in& and& against& Israel& leads& to& an& increase& in&
voting& share& of& 1.35& percentage& points& for& right& bloc& parties.& They& provide& statistical&
evidence& that& their& findings& are& caused& by& the& right& bloc’s& relative& emphasis& on&
counterterrorist& measures,& suggesting& an& elevated& attitude& towards& security& by& the&
Israeli& electorate.& Their& explanation& is& in& support& of& what& Stigler& (1973),& and&
consecutively& Kinder& and& Kiewit& (1981),& regard& as& the& policyboriented& voting&
hypothesis:& it& suggests& that& political& parties& gain& from& the& urgency& of& issues,& to& which&
they&pledge&to&assign&highest&importance.&&
&
Electoral&reactions&to&terrorist&attacks&do&vary,&however,&as&evidenced&by&evaluations&on&
the& 2004& Al& Qaeda& attack& on& Madrid.& Using& a& multinomial& logit& specification& for& the&
voting& decision,& Bali& (2007)& finds& that& the& terror& attack& played& a& decisive& role& in& the&
incumbent& People’s& Party’s& (PP)& electoral& defeat& only& days& later.& Results& are& based& on&
data& from& individual& level& surveys& directly& enquiring& about& the& election& impact& of& the&
attacks,&and&seem&at&odds&with&the&conclusions&drawn&from&Berrebi&and&Klor&(2008)&due&
to& the& PP’s& comparatively& strong& counterterrorist& focus.& It& is& therefore& striking& to& find&
that& Montalvo& (2011),& by& employing& an& entirely& different& evaluation& method& and& data,&
arrives& at& the& same& outcome:& had& the& attacks& not& taken& place,& the& PP& would& have&
remained& in& office& with& a& 5b8& percentage& point& lead& over& their& socialist& pendant,& as&
predicted& by& Spanish& forecasting& agencies.& Montalvo,& using& actual& voting& results,&
capitalized&on&the&unusually&dense&succession&of&attack&and&election,&which&were&only&3&
days&apart.&It&meant&that&Spanish&nationals&living&abroad&and&whose&voting&deadline&was&
by&coincidence&set&prior&to&the&attack,&were&unaffected&by&the&terror&act&in&their&electoral&
decision,&enticing&a&difference&in&difference&estimation.&&
&
Do& these& results& reveal& a& reduced& desire& for& safety& and& security& amongst& the& Spanish&
population?& While& certainly& the& case& under& the& policy& voting& hypothesis,& this& author&
believes& that& statistics& by& the& Centro& de& Investigaciones& Sociológicas& (CIS)& suggest&
otherwise:&while&prior&to&the&attack&only&40%&of&respondents&classified&terrorism&as&one&
of&the&most&salient&issues&in&Spain&(CIS&2004),&the&Madrid&bombings&hoisted&this&number&
to&approximately&80%&in&March&2004&(CIS&2004).&Further,&the&PP&strongly&supported&US&
foreign&policy&in&the&Middle&East&through&direct&involvement&with&boots&on&the&ground&in&
Iraq,&a&factor&that&Bali&claims&was&critical&in&the&PP’s&electoral&defeat.&Voting&into&office&a&
leftist&government&with&diminished&transatlantic&relations&would&thus&be&consistent&with&
greater&desire&for&national&security&against&international&terrorism.&It&is&a&realization&of&
partisan&voting&behaviour&hypothesized&by&Powell&and&Whitten&(1993),&which&suggests&
that& incumbent& parties& are& assessed& on& their& performance& on& topics& to& which& they&
attribute&prominence.&In&essence,&this&means&for&our&study,&that&if&one&were&to&use&voting&
behaviour& as& the& dependent& variable,& an& acute& understanding& of& political& intricacies& is&
indispensable.&
&
&
&
&
8&
&
International!affairs!shocks!
!
Research& on& international& shocks,& such& as& wars,& invasions& or& foreign& policy& crises&
provides& another& source& for& determining& a& potential& dependent& variable& that& captures&
security& attitudes.& It& differs& from& terrorism& inasmuch& as& it& concerns& clashes& between&
entire&nations.&While&these&are&arguably&more&inert,&their&consequences&are&potentially&
more& severe.& Although& Oskamp& (1991)& describes& internationally& oriented& attitudes&
particularly&robust&to&change,&they&can&be&affected&through&personal&experience&or&what&
he& calls& spectacular& events.& In& fact& “when& rapid& attitude& shifts& occur,& they& are& usually&
related&to&major&events&in&international&affairs”.&The&channel,&through&which&wars&affect&
safety& attitudes,& is& fear;& fear& that& is& fuelled& by& the& dissemination& of& information& about&
potential& attacks& à& la& Chivian& et& al.& (1982),& who& hypothesize& that& even& under& a& small&
scale&nuclear&attack&on&the&United&States&the&full&stock&of&American&blood&supplies&would&
be&needed&within&the&first&24&hours.&&
&
There& is& an& assortment& of& pathways& through& which& one& can& capture& effects& of& high&
impact& international& events& on& security& attitudes.& One& is& to& look& at& public& approval&
ratings& regarding& a& country’s& internationalism& as& opposed& to& isolationism& (that& is&
restricting& offshore& intervention).& A& brief& example& may& clarify:& Gallup& (1972),& almost&
two& decades& after& World& War& I,& enquired& whether& the& public& would& support& American&
involvement& in& another& large& scale& European& war.& Only& 5%& affirmed.& In& 1939,& almost&
two&thirds&of&Americans&said&the&United&States&should&remain&impartial&if&Germany&and&
Italy& were& to& attack& the& United& Kingdom& and& France& (Free& && Cantril& 1967).& Following&
World&War&II,&however,&Page&and&Shapiro&(1982)&find&U.S.&internationalism&to&reach&its&
maximum& popularity.& Successful& wars& seem& to& boost& internationalism,& whilst&
contentious& interventions,& that& put& the& country& at& risk,& encourage& solitary& behaviour.&
The&reduction&in&internationalist&approval&after&the&Vietnam&War&is&a&case&in&point.&Here,&
the& effects& of& ‘bad& wars’& are& probably& stronger.& Tetlock& (1981)& has& found& isolationist&
speeches&to&be&emotionally&more&charged,&and&emotion&greatly&influences&attitudes.&As&
the& London& riots,& however,& are& an& internal& uprising,& this& approach& may& be& difficult& to&
emulate.&&
&
A&better&tool&may&be&enquiring&about&the&most&salient&concerns&of&citizens,&for&concerns&
affect& the& security& of& individuals’& lives.& Wars& and& international& crises& seem& to& have&
predominant& ranks& on& such& ‘concern& indices’.& In& 1953,& more& than& half& of& U.S& citizens&
declared& the& Korean& War& as& the& most& upsetting& concern,& and& in& 1963& “danger& from&
Cuba”& ranked& highest& for& a& quarter& of& the& population& (Gallup& 1972,& Smith& 1985).& A&
closely& related& and& highly& applicable& method& is& looking& at& expenditure& on& defence& and&
security& enhancing& equipment,& or& alternatively& at& its& popularity.& Wars& and& similar&
international& affairs& tend& to& lead& to& arms& races.& During& the& Cold& War,& in& the& World&
Federalist& Newsletter& of& 1980& it& was& claimed& “the& money& required& to& provide& food,&
water,& education,& health& and& housing& to& everyone& in& the& world& (…)& is& a& huge& sum& of&
money!about&as&much&as&the&world&spends&on&arms&every&two&weeks”.&Observing&the&
public&opinion&on&military&or&police&expenditure&is&the&pendant&in&the&case&of&the&London&
riots.&
&
However,& not& only& expenditure& ratings,& but& also& ‘trust& ratings’& about& the& institutions&
themselves& are& highly& sensitive& to& wars& and& illuminative& about& changing& views& on&
security.& For& instance,& in& 1974,& the& Institute& for& Social& Research& concluded& that& falling&
levels&of&‘trust&in&the&U.S.&government’&were&largely&due&to&the&distress&surrounding&the&
9&
&
Vietnam& War.& To& more& closely& capture& the& effects& on& security& attitudes,& one& should&
assess& trust& or& approval& ratings& of& government& arms& in& charge& of& maintaining& public&
safety.&Evaluating&attitudes&towards&the&police&may,&in&this&regard,&be&indicative.&
&
&
Riots!
!
As&evidenced&by&Campbell&et&al.&(1976)&internal&uprisings&such&as&disorders&and&public&
protests& are& amongst& the& most& potent& factors& that& negatively& affect& quality& of& life& and&
should&thus&lend&themselves&particularly&well&to&assessing&security&attitudes.&It&is&hence&
surprising& that& there& are& only& few& academic& analyses& on& riots& that& allow& for& inference&
about& perceived& safety.& One& of& these& is& Collins’& and& Margo’s& (2007)& examination& of&
property&values&following&the&1960s&riots&in&American&cities.&According&to&the&authors,&
property&values&“capture&a&broad&range&of&local&characteristics&and&amenities”,&amongst&
which& they& mention& security.& Using& OLS& and& IV& estimations& they& find& that& “both&
moderate&and&severe&riots&reduced&the&growth&rate&of&housing&values&between&1960&and&
1980&(...)&visible&in&overall&property&value&trends”.&Per&se,&their&results&suggest&riots&lead&
to&an&elevated&desire&for&safety,&however,&there&are&several&causes&for&concern&regarding&
their&methodology:&evaluating&property&prices&across&several&decades&bears&the&risk&of&
shocks& other& than& riots& distorting& the& picture.& Further,& as& the& authors& state& in& the&
abstract,&in&the&“1960s&many&American&cities&experienced&riots”.&This,&to&certain&extent,&
makes&rioting&behaviour&foreseeable&and&questions&the&exogeneity&of&the&riots&examined.&
Regardless,& in& their& twobfold& approach,& they& first& use& OLS& to& measure& the& effect& of& riot&
severity&on&property&value&trends,&and&then&continue&to&employ&“rainfall&in&April&1968”&
(the& month& of& the& Martin& Luther& King& assassination)& as& an& instrumented& variable& to&
isolate&“exogenous&variation&in&riot&severity”.&In&the&OLS&regression,&however,&looms&the&
presence&of&endogeneity&bias,&as&changes&in&residential&property&values&may&also&affect&
the& intensity& of& riots.& Rising& property& values& in& a& specific& region& are& likely& correlated&
with&an&upgrade&in&social&stratum,&therefore&reducing&the&probability&of&intense&protests.&
Using&rainfall&for&2SLS&may&further&fail&the&exclusion&restriction:&rain&in&April&1968&may&
be& positively& correlated& with& overall& rainfall,& and& overall& rainfall& is& likely& to& inhibit&
property& value& growth& in& an& increasingly& wealthy& society& (i.e.& the& rich& move& to& sunny&
places).& More& generally,& though,& with& heterogeneous& treatment& effects,& an& IV& approach&
measures& the& local& average& treatment& effect& (LATE).& An& unbiased& estimation& of& LATE&
requires&no&‘defiers’.&In&the&context&of&their&paper&this&means&that&rainfall&does&not&make&
people&more&likely&to&protest,&and&sunshine&does&not&make&them&more&likely&to&remain&in&
their& homes.& The& very& nature& of& riots& (as& a& symbol& of& protest& against& authority),&
however,& could& mobilize& people& specifically& under& rainfall& as& an& indication& of& their&
commitment&(in&the&spirit&of&“now&more&than&ever”).&With&defiers,&the&IV&inferences&may&
be&spurious.&&
&
The&vast&majority&of&existing&literature&on&riots&aims&to&explain&their&causes.&DiPasquale&
and& Glaeser& (1996)& for& instance& look& at& the& determinants& of& the& Los& Angeles& riots& in&
1990.&This&observation&also&seems&to&hold&true&for&the&London&riots&of&2011.&Tonkin&et&
al.& (2012)& aim& to& explain& them& with& heavy& focus& on& social& media,& Davies& et& al.& (2013)&
derive& a& mathematical& model& that& seeks& to& explain& the& riots’& geographical& spread,& and&
Millington&(2012)&assesses&the&influence&of&London’s&urban&structure.&In&fact,&Braakman&
(2012)&seems&to&be&one&of&the&very&few&to&analyze&the&effects&of&the&London&riots.&Using&a&
difference& in& difference& approach& he,& however,& only& looks& to& evaluate& effects& on& crime&
after&the&riots&via&the&channels&of&unemployment&and&policing.&Overlooking&the&diverging&
10&
&
prebtrends& of& his& treatment& and& control& groups& shortly& prior& to& the& London& riots,& he&
observes& falling& crime& levels& that& cannot& be& explained& by& either& of& the& above& channels,&
hinting& at& a& psychological& explanation& (i.e.& the& London& riots& serving& as& an& outlet& for&
“letting&off&steam”).&&
&
The!longevity!of!security!attitude!changes!after!extraordinary!events!
!
Rokeach& (1968)& defined& attitudes& as& a& “relatively& enduring& organization& of& beliefs&
around& an& object& or& situation”.& It& suggests& that& in& order& to& truly& declare& attitude&
alteration,&one&must&find&relatively&prolonged&changes&in&the&dependent&variable.&In&his&
précis& on& existing& literature,& Oskamp& (1991)& concludes& that& “often& even& spectacular&
events& (...)& may& cause& only& a& brief& fluctuation& followed& by& a& return& to& the& prebexisting&
attitude”.& Bali& (2007)& confirms& this& in& the& case& of& terrorism& when& she& finds& levels& of&
concern& “to& approach& prebattack& levels”& several& months& after& the& Madrid& bombings.&
Huddy& et& al.& (2002),& too,& found& support& levels& for& federal& surveillance& of& ebmail& and&
telephone& communication& to& decline& over& time& following& 9/11.& Within& 2& months,& polls&
from& both& FOX& and& CBS& showed& public& support& to& fall& by& more& than& a& quarter,& and&
perceived&threat&fell&to&pre&9/11&levels.&
&
Collins& and& Margo& (2007)& render& it& “entirely& plausible& that& a& riot’s& effect& on& perceived&
amenities”,& such& as& safety& and& security,& is& ephemeral,& too.& Yet& Matheson& and& Baade&
(2004)&find&the&Rodney&King&riots&in&Los&Angeles&to&clearly&have&had&persistent&effects,&
and&Riley&and&Pettigrew&(1976)&show&prolonged&changes&in&public&opinion&following&the&
Martin&Luther&King&assassination.&Given&this&heterogeneity&in&results,&it&is&interesting&to&
examine,&which&category&the&London&riots&fall&into.&
11&
&
#
12#
#
The$significance$for$our$study$
$
From#our#review#of#the#general#variables#of#security#and#safety#attitudes#we#have#found#
canonical#control#variables#that#should#allow#us#to#isolate#the#riot#impacts.#Our#critical#
standpoint#allowed#us#to#further#identify#unorthodox#independent#variables#that#we#will#
tentatively#include#in#our#regressions#in#Chapters#III#and#IV.#Reviewing#the#literature#on#
high# impact# events# (more# specifically# terrorism# and# international# affairs# shocks)# we#
were# able# to# discover# multiple# candidates# for# our# dependent# variable# that# must#
accurately# reflect# safety# and# security# attitudes.# A# critical# analysis# on# riot# literature#
revealed# not# only# certain# methodological# flaws,# but# also# that# existing# papers# on# the#
London#riots#lack#thorough#evaluation#of#their#effects!especially#regarding#the#human#
and#psychological#consequences.#In#addition,#the#question#of#whether#the#London#riots#
had#longHlasting#effects#is#still#pristine.#We#aim#to#fill#this#void#with#our#study.#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
$
13#
#
Chapter$III$
$
Testing(the(theory(through(time(differencing! a(preliminary(
approach(
(
The$data$and$method$
$
The#purpose#of#this#chapter,#to#large#extent,#is#verification#and#justification#of#the#theoryH
based# control# variables# found# in# Chapter# II,# as# well# as# a# preliminary# exploration# into#
whether#the#2011#riots#affected#safety#and#security#attitudes.#
#
The#dataset#in#use#is#the#2011#British#Social#Attitudes#survey,#which,#with#a#sample#size#
of# 3,308,# aims# to# capture# “people’s# changing# social,# political# and# moral# attitudes”#
(www.britsocat.co.uk,# 2011).# Fieldwork# took# place# between# the# 21st# of# June# 2011# and#
the#6th#of#November#2011#and#therefore,#since#the#riots#occurred#in#August,#allows#a#first#
difference#estimation#across#time#(i.e.#pre#riots#versus#post#riots)1,#after#merging#it#with#
exclusively# released# date# identifiers.# While# the# dataset# does# include# a# geographical#
identifier,# variation# in# riot# intensity# across# different# localities# cannot# be# used# in# this#
instance.# This# is# due# to# the# high# level# identification# of# geographical# units# into#
Government#Office#Regions#(GOR).#One#such#GOR,#for#instance,#is#Wales.#While#no#rioting#
occurred#in#Aberystwyth#(western#Wales),#criminal#activity#was#verified#in#Cardiff.#It#is#
thus#impossible#to#use#GOR#as#an#accurate#identifier#for#the#separation#of#individual#data#
points#into#treatment#or#control#groups.#Further,#due#to#the#form#of#the#time#identifiers,#
the#crucial#exercise#of#checking#parallel#preHtrends#is#impossible,#making#a#difference#in#
difference#estimation#unattainable.##
#
To# see# which# dependent# variable# to# choose# as# a# representation# of# security# and# safety#
attitudes,# we# look# at# what# existing# literature# from# riots,# terrorism# and# international#
affairs# shocks# suggests# (see# Chapter# II).# From# the# latter,# we# know# that# salience# of#
concerns#is#a#good#indicator#for#stronger#attitudes#(i),#and#from#terrorism#literature#we#
found#people#are#willing#to#give#up#some#degrees#of#freedom#for#enhanced#security.#This#
need,#for#instance,#may#manifest#itself#through#the#desire#for#a#stronger#police#force#(ii).#
We# therefore# choose# to# look# at# ‘the# probability# of# listing# increased# police# and# defence#
expenditure#as#one’s#priority#as#our#dependent#variable,#as#it#is#a#combination#of#both#(i)#
and# (ii).# The# following# figure# is# a# first# hint# at# the# behaviour# of# our# chosen# dependent#
variable,#merely#using#mean#levels:#
#
#
#
#
#
#
#
#
#
########################################################
1#The#original#dataset#did#not#include#a#time#identifier,#which#had#to#be#requested#from#the#data#depositor.#
I#am#indebted#to#Roger#Stafford#from#NatCen#for#his#kind#cooperation.#
14#
#
Figure#1#
#
#
#
From# Chapter# II# we# further# know# which# control# variables# theory# suggests# to# include,#
next# to# our# independent# variable# of# interest# (“post”).# “Post”# indicates# whether# a# data#
point#was#observed#prior#to#the#riots,#or#afterwards.#The#control#variables#included#are:#
#
a. male!a#binary#variable#at#unity#if#the#respondent#is#male#
b. RAge!stating#the#respondent’s#age#
c. Rmarried!a#binary#variable#at#unity#if#the#respondent#is#married#
d. White!a#binary#variable#at#unity#if#the#respondent#is#white#
e. eduhigh!a#binary#variable#at#unity#if#the#respondent#has#a#university#degree#
f. edulow!a#binary#variable#at#unity#if#the#respondent#has#no#qualification#
g. extremeideo!a# binary# variable# at# unity# if# the# respondent# has# a# deeply#
nationalist#or#socialist#party#ideology#
h. ReadNews!a#binary#variable#at#unity#if#the#respondent#reads#the#morning#news#
more#than#three#times#per#week#
i. post!a#binary#variable#at#unity#if#the#respondent#was#interviewed#after#the#riots#
#
At#this#point,#the#reader#might#have#noticed#the#lack#of#an#income#control#variable.#While#
we# do# believe# income# to# be# an# important# explanatory# variable,# the# BSA# data# does# not#
include# income# categories.# The# only# variable# in# the# vicinity# is# ‘income# comfort’,# which#
per# se,# however,# is# dependent# on# a# variety# of# explanatory# variables.# Inclusion# in# our#
model#would#therefore#likely#pose#a#multicollinearity#problem.##
#
#
15#
#
#
The# “post”# variable# represents# our# first# difference# approach# across# time,# and# its#
coefficient#is#of#interest#when#evaluating#whether#the#riots#affected#safety#and#security#
attitudes.# In# order# to# give# some# credibility# to# the# notion,# however,# that# differencing#
across#time#is#meaningful,#one#must#assess,#whether#the#subsequently#derived#preHriot#
and# postHriot# groups# are# of# similar# makeup.# Otherwise,# differences# in# levels# of# the#
dependent# variable# across# these# two# groups# may# be# due# to# large# discrepancies# in# the#
levels#of#the#control#variables:##
#
Figure#2#
#
#
As#shown#in#Figure#2,#the#setups#of#both#groups#in#terms#of#control#variables#are#highly#
similar,# and# while# there# are# significantly# more# observations# in# the# postHriot# group,#
means#and#standard#deviations#of#all#variables#are#almost#identical.#This#generates#some#
confidence# when# interpreting# our# first# difference# estimation# as# an# indication# for#
whether#the#riots#affected#safety#and#security#attitudes.#
#
Since# our# dependent# variable# is# measured# as# a# probability# (listing# increased# police#
expenditure# as# one’s# highest# priority# is# binary),# it# is# bounded# between# 0# and# 1.# We#
therefore# use# a# Probit# specification.# Unfortunately,# this# makes# the# interpretation# of#
coefficients# less# straightforward.# However,# we# are# more# interested# in# whether# the#
included# independent# variables# have# explanatory# statistical# power,# and# will# therefore#
only#comment#on#significance#and#direction.#If#one#were#to#discuss#the#effect#sizes,#one#
should# focus# on# the# average# partial# effect,# given# the# multiple# binary# controls# and# the#
probit#model.#
#
#
#
16#
#
The$results$
$
The#results#from#the#regression#are#presented#in#the#figure#below:#
#
#
Figure#3#
#
#
The# coefficients# suggest# that# being# male,# being# white# and# increased# age# all# make#
respondents# more# likely# to# list# higher# police# expenditure# as# their# first# priority# (these#
estimates# are# significant# at# the# 5%# level),# suggesting# a# greater# desire# for# security# and#
safety.# The# positive# coefficient# on# “male”# is# of# considerable# size,# and# may# come# as# a#
surprise#to#the#reader,#given#that#Mulvey#(2002)#and#Gallup#(2000)#both#found#women#
to# “express# greater# fear# of# crime”.# This# suggests,# that# indeed# there# do# exist# genetic#
factors#that#“establish#a#natural#tendency”#(see#Chapter#II).#One#hypothesis#may#be#that#
men# are# more# comfortable# with# the# idea# of# countering# crime# through# a# stronger#
executive# force# (‘iron# fist’)# than# women,# however,# further# research# is# needed# for# a#
confident# justification.# The# positive# coefficient# on# white,# on# the# other# hand,# may# be#
explained#by#the#overrepresentation#of#ethnic#minorities#in#the#criminal#justice#system#
(Ministry# of# Justice,# 2010),# both# when# victimized# and# being# a# suspect.# For# instance,#
based# on# crime# data# between# 2006# and# 2010,# “higher# rates# of# (…)# Stop# and# Searches#
were#recorded#for#all#BME#(ethnic#minority)#groups#(...)#than#for#the#White#group”.#This#
is# likely# to# cause# prejudice# and# weariness# against# the# police# amongst# minorities,# and#
explains# their# reduced# likelihood# of# listing# higher# police# expenditure# as# their# first#
priority.#Finally,#the#elderly#are#more#susceptible#to#crime#victimization,#and#physically#
more#reliant#on#police#for#protection.#
#
Further# in# accordance# with# the# existing# literature,# the# coefficients# on# education# levels#
suggest# that# both# very# high# and# very# low# education# have# respondents# less# likely# to#
desire# increased# police# spending.# This# observation# may# be# explained# by# the# following#
intuitive# argument:# While# the# very# highly# educated# are# more# likely# to# prioritize# more#
academically# and# philosophically# revered# policy# issues# (e.g.# immigration# reform,#
increasing# domestic# competitiveness),# the# very# poorly# educated# may# perhaps# prefer#
higher#benefit#payments#rather#than#stricter#lawHenforcement.##
#
Belonging# to# a# party# of# extreme# ideology,# on# the# other# hand,# is# likely# to# increase# the#
desire#for#higher#police#spending#(this#is#almost#significant#at#the#5%#level).#An#intuitive,#
rational# explanation# is# hard# to# find,# however,# as# ideological# extremists# do# often# times#
17#
#
not#have#cogent#objectives#and#realistic#expectations.#An#ultra#right#wing#individual#may,#
for# instance,# desire# a# stronger# police# out# of# the# belief# that# more# powerful# law#
enforcement# is# needed# to# materialize# his# nationalistic# agenda.# In# reality,# at# least# in#
moderate#countries,#a#stronger#police#is#more#likely#to#clamp#down#on#extremism.##
#
A#departure#from#literature#is#the#estimate#on#being#married#(Rmarried).#Its#coefficient#
is# insignificant,# and# shall# be# reHexamined# in# Chapter# IV.# In# addition,# our# hypothesized#
control# variable# “ReadNews”# proves# to# be# highly# insignificant;# this# is# a# particularly#
interesting#finding,#due#to#the#riots’#heavy#media#exposure.#It#must#be#noted,#however,#
that# media# exposure# might# still# have# had# a# significant# effect,# if# respondents# watched#
rather# than# read# the# news.# Most# importantly,# however,# is# the# positive# coefficient# on#
“post”,# which# means# respondents# interviewed# after# the# riots# were# more# likely# to#
prioritize#higher#police#expenditure,#when#controlling#for#all#other#explanatory#variables#
that#theory#mentions.#While#not#significant#at#the#5%#level,#its#PHvalue#at#0.076#suggests#
there#is#some#evidence#that#safety#and#security#attitudes#were#more#alert#after#the#riots.#
As# a# mere# first# difference# calculation,# however,# the# estimated# correlation# is# hard# to#
interpret# as# a# causal# effect# of# the# riots,# due# to# the# possibility# of# trending# behaviour# by#
the#dependent#variable.#We#make#a#more#thorough#analysis#in#the#next#chapter.#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
18#
#
Chapter$IV$
$
Examining(the(causal(effect(of(the(London(riots(
(
The$data$and$method$
$
In# order# to# directly# examine# the# consequences# of# the# London# riots,# a# more# rigorous#
method# than# in# Chapter# III# (time# differencing)# is# needed.# Making# use# of# the#
heterogeneity#in#riot#intensity#across#different#geographical#regions,#we#now#employ#a#
difference#in#difference#specification.#After#communication#with#the#Metropolitan#Police,#
we#were#able#to#get#access#to#their#dataset#enquiring#attitudes#towards#the#police#during#
2011.#With#the#study#solely#focussing#on#London,#it#featured#a#geographical#identifier#at#
local#authority#level,#and#a#daily#time#identifier#for#each#data#point.#We#matched#it#with#
data#from#the#Home#Office,#recording#all#London#riots#related#crimes#by#local#authority,#
and#found#considerable#variation#in#riot#intensity#across#London’s#local#authorities#(LA)#
(e.g.#430#incidents#in#Croydon,#but#only#23#in#Sutton).#The#mean#riot#related#crimes#per#
LA#were#approximately#107,#with#a#sample#standard#deviation#of#110.##
#
While# the# use# of# different# datasets# in# this# chapter# is# mainly# due# to# the# alteration# in#
method# and# specification,# it# also# allows# us# to# crossHcheck# whether# our# findings# in#
Chapter#III,#especially#regarding#the#coefficient#on#post#and#our#control#variables,#can#be#
confirmed# on# the# basis# of# different# data.# In# order# to# find# an# appropriate# dependent#
variable#in#the#new#dataset,#we#again#look#at#the#terrorism#literature.#Here,#Huddy#et#al.#
(2002)# measured# safety# attitudes# by# examining# the# likelihood# assigned# by# the# US#
population# to# another# terror# attack,# following# 9/11.# By# inference,# as# the# London# riots#
are#mainly#associated#with#criminal#activity,#we#should#therefore#be#able#to#assess#safety#
attitudes#through#reported#fear#levels#of#becoming#a#victim#of#crime.#We#call#this#binary#
variable# victimization_fear# (1# indicates# concern# with# being# victimized,# 0# indicates# no#
such#concern).#
#
Our# control# variables# were# to# large# extent# chosen# equivalently# to# Chapter# III,# with# the#
exception#of#education#levels,#party#association#and#newspaper#readership#(these#were#
not#available#in#the#data).#However,#as#a#representative#for#income,#we#included#socioH
economic#class.#Given#the#nature#of#our#dependent#variable,#it#also#seemed#appropriate#
to#control#for#levels#of#police#exposure#(e.g.#highpol#means#high#police#visibility),#as#one#
expects#higher#police#visibility#to#reduce#fear#of#crime#victimization.#Police#exposure#was#
further# interacted# with# the# pre/post# time# identifier,# to# control# for# the# increased#
deployment#of#police#following#the#riots.##
#
The# pre/post# time# identifier# is# called# post_1month,# where# the# post# group# includes# all#
observations#from#the#day#after#the#last#riot#incident,#up#until#1#month#after#the#riots.#In#
order# to# assess# the# longevity# of# the# effects,# the# time# identifier# is# then# redefined# to#
post_2month,#where#the#post#group#now#includes#the#observations#within#the#time#span#
of#2#months#after#the#riots.#
#
At#the#core#of#a#difference#in#difference#estimation#is#the#separation#of#data#points#into#
treatment#and#control#groups.#In#our#estimation,#all#observations#from#local#authorities#
with# very# low# riot# related# crime# were# assigned# into# the# control# group,# with# the#
19#
#
remainder#classified#as#treatment.#We#argue,#that#the#channel,#through#which#riot#crime#
affects#victimization#fear#further#depends#on#the#number#of#people#living#inside#the#local#
authority.# One# incident# in# a# densely# populated# area# will# hardly# be# noticed,# when# it# is#
likely# widely# perceived# in# a# small# community.# We# therefore# defined# treatment#
(treatment_b)#to#be#at#unity#if#a#local#authority#experienced#more#than#1.5#riot#related#
crime#incidents#per#10,000#inhabitants.##
#
The$results$
$
Using#a#probit#specification2,#the#results#for#the#short#term#are#as#follows:#
#
Figure#4#
#
#
########################################################
2#In#literature,#there#exists#a#debate#revolving#around#the#interpretation#and#validity#of#interaction#effects#
in#nonHlinear#models#(see#Ai#and#Norton#(2003),#as#opposed#to#Puhani#(2012).#Therefore,#all#results#
regarding#the#direction#and#statistical#significance#of#our#coefficients#in#our#probit#specification#have#also#
been#tested#and#confirmed#in#a#linear#probability#model#specification#(see#Appendix#for#regression#
output).#
20#
#
For#the#reader’s#information,#the#base#age#group#is#19#to#24#year#old#individuals,#and#the#
base# socioHeconomic# class# is# C2# (skilled# manual# workers).# # The# close# interpretation# of#
these#control#variables#is#left#to#the#reader.#
#
More# importantly,# we# find# the# coefficient# on# post_1month# to# be# positive# and# highly#
significant.#This#means#fear#of#victimization#was#higher#in#the#month#following#the#riots,#
than#it#was#before,#and#confirms#our#results#from#Chapter#III.#However,#the#coefficient#on#
treatment# is# insignificant,# so# there# is# not# sufficient# evidence# to# conclude# that# higher#
exposure# to# riot# crime# per# se# increased# fear# levels.# The# insignificant# difference# in#
difference#estimate#underpins#this.#Therefore,#we#can#say#with#certainty,#that#something#
must#have#increased#fear#levels#in#the#one#month#following#the#riots,#but#not#riot#crime#
itself.# Was# this# “something”# related# to# the# riots?# This# we# cannot# know# for# certain.#
However,#if#we#were#to#include#observations#up#to#two#months#following#the#riots,#and#
then#observed#an#insignificant#coefficient#on#our#post#variable#(post_2month),#it#would#
support# the# notion# that# elevated# levels# of# fear# were# indeed# riot,# but# not# riot# crime,#
related.##
#
Including#these#observations,#we#find:#
#
Figure#5#
#
21#
#
#
In# fact,# we# see# that# the# coefficient# on# post# (post_2month)# is# now# closer# to# zero# and#
insignificant.# In# combination# with# our# previous# results,# this# means# victimization# fear#
levels#were#at#base#prior#to#the#riots,#then#increased#directly#after#the#riots#took#place,#
and#returned#to#base#in#the#second#month#post#the#incidents.#It#is#unlikely#that#any#other#
event#could#have#caused#such#a#timeline#that#exactly#fits#the#London#riots,#and#affected#
the#population’s#fear#of#becoming#a#victim#of#crime#in#London.#
#
So#if#increased#fear#levels#were#indeed#related#to#the#London#riots,#but#not#caused#by#riot#
crime#itself#(insignificant#treatment#and#difference#in#difference#estimates),#what#aspect#
of#the#unrests#could#have#caused#these#more#cautious#safety#and#security#attitudes?#
$
A$case$for$the$media$
$
In# Chapter# I,# the# heavy# coverage# of# the# London# riots# by# the# media# became# apparent.#
Coupled# with# often# dramatic# rhetoric# about# riot# incidents,# could# the# media# have# been#
responsible#for#Londoners’#reduced#perception#of#safety?#With#radio,#newspaper#and#TV#
reports#permeating#all#local#authorities#in#London,#the#media#hypothesis#would#explain#
why#we#were#unable#to#detect#a#difference#in#security#attitudes#between#the#control#and#
treatment#groups,#yet#see#a#significant#increase#in#fear#levels#one#month#after#the#riots.#
#
In# order# to# give# more# credibility# to# the# media# hypothesis,# note# that# the# media# very#
heavily# lobbied# for# stiffer# sentences# for# riot# perpetrators.# Demanding# stiffer# sentences#
for#past#crimes,#however,#is#rather#uncommon,#and#appears#idiosyncratic#to#the#London#
riots#coverage.#Any#increase#in#support#for#stiffer#sentences#would#therefore#very#likely#
be#media#induced.##
#
Using# such# support# as# our# regressand,# we# would# expect# a# highly# significant# and#
persistent#post#coefficient,#and#insignificant#estimates#on#our#treatment#and#difference#
in#difference#variables#(these#two#only#measure#safety#attitude#effects#directly#linked#to#
actual#riot#crime,#which#we#hypothesize#to#be#insignificant).#
$
Figure#6#
$
$
$
$
$
$
$
$
$
$
$
$
$
$
22#
#
$
Given#satisfactory#preHtrends#in#the#support#levels#for#stiffer#sentences,#we#find#the#
following:#
#
Figure#7#
The# coefficients# on# our# treatment# and# difference# in# difference# variables# are# in# fact#
statistically# insignificant.# This# again# supports# the# hypothesis# that# more# alert# safety#
attitudes# are# unrelated# to# actual# riot# crime.# The# positive# and# statistically# significant#
coefficient# on# post# means# in# the# month# following# the# riots# there# was# a# significant#
increase# in# Londoners# demanding# stiffer# sentences,# as# was# heavily# lobbied# for# by# the#
media.##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
23#
#
#
Figure#8#
#
In#fact,#when#looking#at#Figure#8,#we#find#that#the#effect#on#post#after#2#months#is#still#
positive# (though# slightly# smaller)# and# statistically# significant.# It# shows,# that# while# the#
time# effect# on# victimization# fear# was# ephemeral,# the# time# effect# on# the# heavily# media#
lobbied#support#for#stiffer#sentences#is#more#persistent.#Further,#the#negative#treatment#
coefficient# is# now# significant.# This# indicates# that,# after# some# time,# the# influence# of# the#
media# is# stronger# in# local# authorities# where# people# did# not# experience# the# riots# first#
hand.#
#
$
$
$
$
$
$
$
$
$
$
$
$
$
24#
#
#
Chapter$V$
$
Conclusion(
$
From#our#analysis#in#Chapter#III#we#have#found#that,#from#a#time#perspective,#attitudes#
towards# security# and# safety# were# more# alert# after# the# London# riots.# This# was#
consecutively#confirmed#in#Chapter#IV,#using#a#different#data#set.#We#also#found#that#this#
effect# diminishes# over# time# (peaking# within# the# first# month# after# the# incidents),# albeit#
being#more#persistent#when#looking#at#support#for#stiffer#sentences#for#riot#perpetrators#
as#opposed#to#fear#of#victimization.#Further,#we#determined#that#riot#crime#per#se#did#not#
significantly# affect# attitudes# towards# security# and# safety.# These# results# are# robust,# and#
directly#deduced#from#our#data#analysis.#
#
Did#the#London#riots#affect#security#and#safety#attitudes?#We#believe#they#did,#due#to#the#
extraordinary#path#of#our#representative#dependent#variables#over#time.#However,#given#
the# insignificant# estimates# on# our# treatment# and# difference# in# difference# variables,# we#
deduce# that# the# channel# through# which# the# riots# affected# safety# attitudes# was# not# riot#
crime# itself.# Given# our# analysis# in# the# second# part# of# Chapter# IV,# we# induce# that# the#
media# is# likely# to# have# been# the# driving# force# behind# the# increased,# and# relatively#
persistent#desire#for#safety#and#security.##
#
It# is# important# to# note,# however,# that# the# latter# conclusion# cannot# be# directly,#
statistically#inferred#from#our#data,#but#is#an#induction#based#on#our#data#analysis#and#
media# behaviour# during# the# riots.# It# critically# hinges# upon# the# assumption# that# the#
desire#for#stiffer#sentences#was#incited#by#its#strong#media#lobbying.##
#
As#such,#an#optimal#estimation#strategy#would#in#the#base#case#include#a#control#variable#
for# media# exposure,# and# in# the# best# case# an# interaction# between# media# exposure# and#
time# in# our# difference# in# difference# estimation.# We# encourage# other# researchers#
interested# in# this# topic# to# include# these# variables.# Further,# a# more# clearHcut# division#
between#treatment#and#control#groups#would#be#favourable.#Ideally,#riot#intensity#in#our#
control#group#should#be#zero.#Due#to#the#specificities#of#our#dataset,#however,#we#had#to#
adopt#a#strategy#of#determining#a#riot#intensity#cutHoff#point.#Anything#below#this#point#
would#belong#to#the#control#group,#any#data#point#above#to#our#treatment#group.#This,#
for#instance,#could#potentially#contaminate#our#control#group#through#individuals,#which#
were#directly#affected#by#one#of#the#scarce#riot#incidents#in#control#local#authorities.##
#
Inter# alia,# this# study# provides# an# indication# of# the# importance# of# the# media# on# the#
formation#of#attitudes#towards#security#and#safety.#However,#establishing#such#proof#is#
beyond# the# scope# of# this# study.# We# would# welcome# increased# attention# on# this# issue#
from#future#research,#so#it#can#be#determined#whether#media#exposure#should#be#added#
to#the#body#of#literature#regarding#safety#and#security#attitudes.##
#
#
#
#
#
#
25#
#
Bibliography$
$
Ai,# C.,# &# Norton,# E.# (2003).# Interaction# Terms# in# Logit# and# Probit# Models.# Economics(
Letters,(80,(p123–129.#
#
Aiken,# L# (2002).# Attitudes( and( Related( Psychosocial( Constructs.# London:# Sage#
Publications.#p25.#
#
Allport,# G.# (1935).# Attitudes.# In# C.# Murchison# (Ed.),# A( handbook( of( social( psychology(
((pp.798H844).#Worcester,#MA:#Clark#University#Press.#
#
Ashmore,# R.,# &# Del# Boca,# F.# (1976).# Psychological# approaches# to# understanding#
intergroup#conflicts.#In#P.#A.#Katz#(Ed.),#Towards(the(elimination(of(racism.(Elmsford,#NY:#
Pergamon.#
#
Bali,# V.# (2007).# Terror# and# Elections:# Lessons# from# Spain.# Electoral( Studies,( 26,# p669H#
687.##
#
Berrebi,# C.,# &# Klor,# E.# (2008).# Are# Voters# Sensitive# to# Terrorism?# Direct# Evidence# from#
the#Israeli#Electorate,#American(Political(Science(Review,(102(3),#p279H301.#
#
Braakmann,# N.# (2012).# The# effect# of# the# 2011# London# riots# on# crime,# policing# and#
unemployment.#Munich(Personal(RePEc(Archive.(University#Library#of#Munich.#
#
Bureau#of#Justice#Statistics.#(2001,#January).# Criminal(victimization(in(the(United(States,(
1999.#Washington,#DC:#Author.#
#
Campbell,# A.,# Converse,# P.,# &# Rodgers,# W.# (1976).# The( quality( of( American( life:(
Perceptions,(evaluations(and(satisfactions.(New#York:#Russell#Sage#Foundation.#
#
Centro# de# Investigaciones# Sociológicas.# (2004).# Estudio( 2559( postelectoral( de( las(
elecciones( generales( y( autonómicas( de( Andalucía( 2004.# Available:#
http://www.cis.es/cis/opencms/ES/index.html.#Last#accessed#8th#Dec#2013.#
#
Chivian,# E.,# Chivian,# S.,# # Lifton,# R.,# &# Mack,# J.# (Eds.).# (1982).# Last( aid:( The( medical(
dimensions(of(nuclear(war.(San#Francisco:#Freeman.#
#
Collins,#W.,#&#Margo,#R.#(2007).#The#Economic#Aftermath#of#the#1960s#Riots#in#American#
Cities:#Evidence#from#Property#Values.#The(Journal(of(Economic(History,(67,#p849H883.#
#
Daniels,# A.# (2011).# Behind( England's( riots,( a( violent( and( entitled( generation( of( British(
young( people.# Available:# http://www.nydailynews.com/opinion/behindHenglandHriotsH
violentHentitledHgenerationHbritishHyoungHpeopleHarticleH1.945368.# Last# accessed# 6th#
Dec#2013.#
#
Davies,#T.,#Fry,#H.,#Wilson,#A.,#&#Bishop,#S.#(2013).#A#mathematical#model#of#the#London#
riots#and#their#policing.#Scientific(Reports,#3,#p1303.#
#
#
26#
#
Davis,#D.,#&#Silver,#B.#(2004).#Civil#liberties#vs.#security:#Public#opinion#in#the#context#of#
the#terrorist#attacks#on#America.#American(Journal(of(Political(Science,#48,#p28–46.#
#
DiPasquale,#D.,#&#Glaeser,#E.#(1996).#The#Los#Angeles#Riot#and#the#Economics#of#Urban#
Unrest.#Journal(of(Urban(Economics,(XLIII,#p52–78.##
#
#
Emanuel,# R.# (2009).# You( never( want( a( serious( crisis( to( go( to( waste.# Available# from:#
http://www.youtube.com/watch?v=1yeA_kHHLow.#Last#accessed#20th#Dec#2013.#
#
Fazio,# R.# (1988).# On# the# power# and# functionality# of# attitudes:# The# role# of# attitude#
accessability.#In#A.R.#Pratkanis,#S.J.#Breckler,#&#A.#G.#Greenwald#(Eds.),#Attitude(structure(
and(function.#Hilsdale,#NJ:Erlbaum.#
#
Free,#L.,#&#Cantril,#H.#(1967).#The(political(beliefs(of(Americans:(A(study(of(public(opinion,#
New#Brunswick,#NJ:#Rutgers#University#Press.#
#
Gallup,#G.#(1972).#The(Gallup(Poll:(Public(opinion(1935]1971.#New#York:#Random#House.#
#
Gallup,# G.,# Jr.# (2000).# The( Gallup( poll:( Public( opinion( 1999.( Wilmington,# DE:# Scholarly#
Resources.#
#
Gagné,#R.,#&#Briggs,#L.#(1974).#Principles(of(instructional(design.#New#York:#Holt,#Rinehart#
&#Winston.#
#
Gerard,#H.,#&#Orive,#R.#(1987).#The#dynamics#of#opinion#formation.#In#L.#Berkowitz#(ed.),#
Advances(in(experimental(social(psychology((Vol.#20).#San#Diego:#Academic#Press.#
#
Gordon,#C.,#&#Arian,#A.#(2001).#Threat#and#decision#making.#Journal(of(Conflict(Resolution,(
45,(p196H215.##
#
Home#Office.#(2011).#An(overview(of(recorded(crimes(and(arrests(resulting(from(disorder(
events( in( August( 2011.# Available:# https://www.gov.uk/government/publications/anH
overviewHofHrecordedHcrimesHandHarrestsHresultingHfromHdisorderHeventsHinHaugustH
2011.#Last#accessed#28th#Nov#2013.#
#
Huddy,#L.,#&#Khatib,#N.,#&#Capelos,#T.#(2002).#The#PollsHTrends:#Reactions#to#the#terrorist#
attacks#of#September#11,#2001.#Public(Opinion(Quarterly,(66,#p418H450.#
#
Institute# for# Social# Research.# (1974,# Winter).# Public# asked# to# rank# country’s# major#
institutions.#ISR(Newsletter,(1(20),#p8.#
#
Kelman,#H.#(1974).#Attitudes#are#alive#and#well#and#gainfully#employed#in#the#sphere#of#
action.#American(Psychologist,(29,(310H324.#
#
Kinder,D.,#&#Kiewit,#D.#(1981).#Sociotropic#politics:#The#American#case.#British(Journal(of(
Political(Science,(11,#p129H161.##
#
LaPiere,#R.#(1934).#Attitudes#versus#actions.#Social(Forces,(13,(p230H237.#
#
27#
#
Macaulay,# T.# (1830).# Southey’s( colloquies.# (Cited# in# G.# Seldes# (Compiler),# The( great(
quotations.#New#York:#Pocket#Books,#1967,#p.#706.).#
#
Matheson,#V.,#&#Baade,#R.#(2004).#Race#and#riots:#a#note#on#the#economic#impact#of#the#
Rodney#King#Riots.#Urban(Studies,(41(13),#p2691–2696.##
#
Metropolitan# Police.# (2013).# Metropolitan( Police( Public( Attitudes( Surveys,( 2000]2013.#
Available:#
http://www.met.police.uk/about/performance/confidence.htm.#
Last#
accessed#4th#Jan#2014.#
#
Millington,# G.# (2012).# Man# Dem# Link# Up:# London's# AntiHRiots# and# Urban# Modernism.#
Sociological(Research(Online,#17(4),#p12.#
#
Ministry#of#Justice.#(2010).#Statistics(on(Race(and(the(Criminal(Justice(System.#Available:#
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/219
967/statsHraceHcjsH2010.pdf.#Last#accessed#2nd#Feb#2014.#
#
Montalvo,# J.# (2011).# ReHexamining# the# evidence# on# the# electoral# impact# of# terrorist#
attacks:#The#Spanish#election#of#2004.#Electoral(Studies,#31,#p96H106.#
#
Mulvey,#A.#(2002).#Gender,#Economic#Context,#Perceptions#of#Safety,#and#Quality#of#Life:#
A#Case#Study#of#Lowell,#Massachusetts#(U.S.A.),#1982–96.#American(Journal(of(Community(
Psychology.#30#(5),#p655H679.#
#
NatCen.# (2011).# British( Social( Attitudes( 29.# Available:# http://www.bsaH29.natcen.ac.uk.#
Last#accessed#12th#Dec#2013.#
#
Neligan,# M.# (2011).# Riots( to( cost( over( £200( million( ]( ABI.# Available:#
http://uk.reuters.com/article/2011/08/11/ukHbritainHriotsHcostH
idUKTRE77A22H20110811.#Last#accessed#5th#Dec#2013.#
#
Office#
for#
National#
Statistics.#
(2011).#
Census(
data.#
Available:#
http://www.ons.gov.uk/ons/guideHmethod/census/2011/censusHdata/index.html.# Last#
accessed#10th#Feb#2014.#
#
Oskamp,#S#(1991).#Attitudes(and(Opinions.#2nd#ed.#New#Jersey:#PrenticeHHall.#p1H371.#
#
Page,# B.,# &# Shapiro,# R.,# (1982),# Changes# in# Americans’# policy# preferences,# 1935H1979.#
Public(Opinion(Quarterly,(46,(p24H42.#
#
Pettigrew,# T.# (1967).# Social# evaluation# theory:# Convergences# and# applications.# In# D.#
Levine#(Ed.),#Nebraska(symposium(on(motivation((Vol.#15,#p241H311).#Lincoln:#University#
of#Nebraska#Press.#
#
Powell,#G.,#&#Whitten,#G.#(1993).#A#CrossHNational#Analysis#of#Economic#Voting:#Taking#
Account#of#the#Political#Context.#American(Journal(of(Political(Science,(37(2),#p391–#414.##
#
Puhani,#P.#(2012).#The#Treatment#Effect,#the#Cross#Difference,#and#the#Interaction#Term#
in#Nonlinear#“DifferenceHinHDifferences”#Models.#Economics(Letters,(115,#p85H87.#
#
28#
#
Riley,# R.,# &# Pettigrew,# T.# (1976).# Dramatic# events# and# attitude# change.# Journal( of(
Personality(and(Social(Psychology,(34,#p1004–1015.#
#
Rogers,#S.#(2011).#Data(journalism(reading(the(riots:(what(we(know.(And(what(we(don't.#
Available:#
http://www.theguardian.com/news/datablog/2011/dec/09/dataH
journalismHreadingHriots.#Last#accessed#7th#Feb#2014.#
#
Rokeach,# M.# (1968).# Beliefs,( attitudes( and( values:( A( theory( of( organization( and( change.(#
San#Francisco:#JosseyHBass.#
#
Sampson,# R.,# &Bartusch,# D.# (1998).# Legal# cynicism# and# (subcultural?)# tolerance# of#
deviance:# The# neighbourhood# context# of# racial# differences.# Law( &( Society( Review,# 32,#
p777H804.#
#
Skinner,#B.#(1957).#Verbal(behavior.(New#York:#AppletonHCenturyHCrofts.#
#
Smith,# T.# (1985).# The# polls:# America’s# most# important# problems.# Part# I:# National# and#
international.#Public(Opinion(Quarterly,#49,(p264H274.#
#
Stigler,G.# (1973).# General# economic# conditions# and# national# elections.# American(
Economic(Review.(Papers(and(Proceedings(63,#p160H164.#
#
Stringer,#D.#(2011).#Tottenham(Riots:(David(Cameron(Recalls(Parliament(During(Ongoing(
Crisis.# Available:# http://www.huffingtonpost.com/2011/08/09/tottenhamHriotsH2011H
londonHparliamentH_n_921885.html.#Last#accessed#4th#Apr#2014.#
#
Taylor,#D.#(1982).#Pluralistic#ignorance#and#the#spiral#of#silence:#A#formal#analysis.#Public(
Opinion(Quarterly,#46,#p311H335.##
#
Telegraph# View.# (2011).# The( criminals( who( shame( our( nation.# Available:#
http://www.telegraph.co.uk/comment/telegraphHview/8691352/TheHcriminalsHwhoH
shameHourHnation.html.#Last#accessed#5th#Dec#2013.#
#
Tetlock,#P.#(1981).#Personality#and#isolationism:#Content#analysis#of#senatorial#speeches.#
Journal(of(Personality(and(Social(Psychology,(41,#p737H743.#
#
The#Sun.#(2011).#Be(decisive.#Dead#Link.#Last#accessed#1st#Dec#2013.#
#
Tonkin,#E.,#Pfeiffer,#H.,#&#Tourte,#G.#(2012).#Twitter,#information#sharing#and#the#London#
riots?.# Bulletin( of( the( American( Society( for( Information( Science( and( Technology,(
38(2),(p49–57.#
#
World#Federalist#Newsletter.#(1980,#Winter).#5(4),#p8.#
#
Zajonc,#R.#(1968).#Attitudinal#effects#of#mere#exposure.#Journal(of(Personality(and(Social(
Psychology,(9(2,(Pt.(2),#p1H27.#
#
$
$
29#
#
Appendix$
#
Chapter(IV(regression(output(using(LPM(
(
Figure#9#
#
DiD#on#victimization#fear#with#post_1month#(LPM#correspondent#to#Figure#4)#
#
(
(
(
(
(
(
(
(
30#
#
Figure#10#
#
DiD#on#victimization#fear#with#post_2month#(LPM#correspondent#to#Figure#5)#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
31#
#
Figure#11#
#
DiD#on#stiffer#sentences#with#post_1month#(LPM#correspondent#to#Figure#7)#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
32#
#
Figure#12#
#
DiD#on#stiffer#sentences#with#post_2month#(LPM#correspondent#to#Figure#8)#
#
#
#
33#
#
Download