Tuki, Daniel
A icle — Published Ve sion
Violen Con lic and Hos ili y Towa ds E hno eligious
Ou g oups in Nige ia
Te o ism and Poli ical Violence
P o ided in Coope a ion wi h:
WZB Be lin Social Science Cen e
Sugges ed Ci a ion: Tuki, Daniel (2025) : Violen Con lic and Hos ili y Towa ds E hno eligious
Ou g oups in Nige ia, Te o ism and Poli ical Violence, ISSN 1556-1836, Taylo & F ancis, London,
Vol. 37, Iss. 2, pp. 239-261,
h ps://doi.o g/10.1080/09546553.2023.2285939
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/307778.2
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Violen Con lic and Hos ili y Towa ds E hno eligious Ou g oups in
Nige ia
Daniel Tuki
Mig a ion, In eg a ion and T ansna ionaliza ion Resea ch Uni , WZB Be lin Social Science Cen e , Be lin, Ge many
ABSTRACT
This s udy examined he e ec o exposu e o iolen con lic on hos ili y
owa ds e hnic and eligious ou g oups among Nige ia’s popula ion and
among i s wo majo eligious g oups (i.e., Ch is ians and Muslims). Violen
con lic had a obus posi i e e ec on ou g oup hos ili y among he
Nige ian popula ion and among Ch is ians. A plausible mechanism behind
his inding is ha he h ea posed by iolen con lic s eng hens ing oup
cohesion, e odes us in ou g oup membe s, and makes in e g oup bound-
a ies salien . This is especially so when he opposi e pa y o he con lic
cons i u es a dis inc cul u al ou g oup. The main con lic a ec ing Ch is ians
in ol es nomadic pas o alis s o Fulani e hnici y, who a e Muslims. Among
Muslims, iolen con lic a he had a weak posi i e e ec on ou g oup
hos ili y ha was no obus o al e na i e ope a ionaliza ions o ou g oup
hos ili y. The null e ec migh be because he main con lic a ec ing Muslims
— he Boko Ha am insu gency—does no in ol e Ch is ians. A signi ican
numbe o Muslims a e also a ec ed by con lic s in ol ing nomadic Fulani
pas o alis s.
KEYWORDS
Violen con lic ; con lic
exposu e; ou g oup hos ili y;
e hnici y; eligion; Nige ia
In oduc ion
A cu so y look a Nige ia e eals ha i has a dyadic s uc u e comp ising a p edominan ly Ch is ian
Sou he n Region and a p edominan ly Muslim No he n Region. Al hough he e a e some o e laps
be ween he wo egions, he con as be ween hem is qui e s a k. The o e lap be ween eligion and
e hnici y makes he aul line be ween he wo egions e en mo e salien .
1
This No h-Sou h bi u ca-
ion is appa en when one looks a Nige ia h ough he lens o he nine ci iliza ions in o which Samuel
Hun ing on di ided he wo ld: Nige ia’s No he n Region was associa ed wi h Islamic ci iliza ion,
while he Sou he n Region was associa ed wi h A ican ci iliza ion.
2
This cul u al di ide has his o ical oo s. Islam i s came o No he n Nige ia be ween he ele en h
and ou een h cen u ies h ough he ans-Saha a ade be ween he Hausa people o No he n
Nige ia and me chan s om he Magh eb s a es. Besides he exchange o angible commodi ies,
he e was also a di usion o cul u al and eligious alues.
3
Islam gained a s onge oo hold in he
egion be ween 1804 o 1808, when a cle ic o Fulani e hnici y, Usman dan Fodio, launched a jihad
agains he ule s o he Hausa kingdoms. The jihad led o he es ablishmen o he Soko o Calipha e,
which consis ed o se e al emi a es. The calipha e was in exis ence o a cen u y un il i s conques by
B i ish o ces a he beginning o he wen ie h cen u y.
4
Al hough Ch is iani y in Nige ia can be
aced o he i een h cen u y when Po uguese sla e ade s isi ed Nige ia’s sou he nmos pa s, i
was no un il he 1840s ha he eligion s a ed o gain a oo hold, p opaga ed by eed sla es om
Sie a Leone and missiona ies om he Wes .
5
CONTACT Daniel Tuki [email p o ec ed] Mig a ion, In eg a ion and T ansna ionaliza ion Resea ch Uni , WZB Be lin Social
Science Cen e , Reichpie schu e 50, Be lin 10785, Ge many
TERRORISM AND POLITICAL VIOLENCE
2025, VOL. 37, NO. 2, 239–261
h ps://doi.o g/10.1080/09546553.2023.2285939
© 2024 The Au ho (s). Published wi h license by Taylo & F ancis G oup, LLC.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/4.0/),
which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. The e ms on which his
a icle has been published allow he pos ing o he Accep ed Manusc ip in a eposi o y by he au ho (s) o wi h hei consen .
Ch is ian missiona y e angeliza ion was concen a ed in Sou he n Nige ia because he Muslim
ule s in he No he n Region, in an e o o p ese e hei eligious way o li e, o bade Ch is ian
p osely iza ion in he egion.
6
The B i ish go e nmen did no change much in No he n Nige ia
a e cap u ing i . They app op ia ed he exis ing ins i u ions and e en used he local Hausa
language in adminis e ing he No he n p o ec o a e. Con e sely, he policies o Wes e niza ion
and Ch is ianiza ion we e pu sued e en ly in he Sou he n P o ec o a e because i s popula ion
was mo e open o Wes e n in luence.
7
A e Nige ia’s independence om B i ish colonial ule in
1960, i emained di ided along e hnic and eligious lines. Commen ing on he No he n-Sou he n
dicho omy, Coleman obse ed: “Ce ain basic unde lying di e ences in his o y, cul u e, empe a-
men , and le els o de elopmen and accul u a ion p o ided he classical se ing o in e g oup
ic ion.”
8
Nige ia’s his o ical imeline is punc ua ed by e hno eligious con lic s as e idenced by he Hausa-
Igbo io s o 1945,
9
he Kano io s o 1953 be ween No he ne s and Sou he ne s,
10
he pog oms o
1966 agains membe s o he Igbo e hnic g oup, which led o he Bia an Wa om 1967 o 1970,
11
he
Ka anchan io s in Kaduna be ween Ch is ians and Muslims in 1987,
12
he Kano io s o 1991 be ween
Ch is ians and Muslims,
13
he Sha iah C isis in Kaduna be ween Ch is ians and Muslims in 2000,
14
he 2011 pos -elec ion iolence, which had a eligious unde one,
15
and he ecu en clashes be ween
Ch is ians and Muslims in Jos,
16
amongs o he s. E hno eligious con lic s a e no peculia o Nige ia;
hey occu in se e al coun ies a ound he wo ld. Examples include he con lic be ween membe s o
he Sinhalese and Tamil e hnic g oups in S i Lanka,
17
he con lic be ween Muslims and Buddhis s in
Myanma ,
18
he iolen clashes be ween Hindus and Muslims
19
and hose be ween he Assamese and
Bengalis,
20
bo h in India, he Nago no-Ka abakh con lic be ween A menians and Aze baijanis,
21
he
Malay-Chinese con lic s
22
and he dispu es be ween Muslims and non-Muslims,
23
bo h in Malaysia,
amongs o he s.
E hno eligious con lic s can lead o seg ega ion, which in u n solidi ies in e g oup bounda ies.
24
Some s udies ha e ecommended seg ega ion as a s a egy o mi iga ing in e e hnic con lic s because
sepa a ing he con lic ing g oups elimina es he exis ing h ea .
25
Bollig has conduc ed a s udy among
nomadic ibes in Kenya whe e he inds ha in e e hnic ies ia ma iage and iendships does no lead
o con lic ing loyal ies, bu nei he does i a enua e he isk o in e e hnic con lic .
26
Using a game
heo e ic app oach, La son has shown ha in e e hnic coope a ion could inc ease he isk o con lic ,
especially when coope a ion is unde pinned by he h ea o e alia ion.
27
She poin s ou ha he speed
o e alia ion depends on he densi y o a g oup’s ne wo k: In o ma ion lows quickly wi hin dense
ne wo ks, and his makes i easy o e alia ion o be me ed ou o misbeha io , which in u n makes
he h ea s made by g oups wi h dense ne wo ks c edible. Con e sely, g oups cha ac e ized by spa se
ne wo ks whe e in o ma ion sp eads slowly, ind i ha d o make c edible h ea s because he slow
di usion o in o ma ion wi hin he ne wo k cons ains g oup mobiliza ion. This is p oblema ic
because i makes con lic s in ac able. Fo ins ance, when a peace ag eemen be ween con lic ing
g oups has been eached, i migh ake longe o his de elopmen o sp ead h ough a spa se
ne wo k; a acks a e he ag eemen could be in e p e ed as unwa an ed, which hen igge s
a new wa e o iolence. She cau ions ha “unca e ul e o s o p omo e peace by imposing c oss-
g oup ies can do mo e ha m han good, especially i hey a e aimed a he mos pe iphe al membe s o
bo h g oups.” (470).
Some s udies a gue ha in e g oup con ac is c ucial in educing e hnic con lic s. Fo ins ance, Eke
has conduc ed a quali a i e s udy in he ci y o Jos in Nige ia’s Middlebel Region whe e he inds ha
when mu ual dis us is p esen and e hnic g oups pe cei e each o he as h ea s, in e e hnic iolence
is likely o e up e en when bo h g oups a e comple ely seg ega ed.
28
This is because seg ega ion
“elimina e[s] oppo uni ies o pos -con lic econcilia ion” and hinde s he ebuilding o in e e hnic
us . Fu he mo e, he asse s ha e en hough pa ial seg ega ion does no en i ely elimina e he
pe cei ed h ea om he ou g oup, i none heless c ea es a enues o con ac be ween he i al g oups
240 D. TUKI
which se e as condui s o es ablishing us and a e ing u u e con lic . Ano he key inding o his
s udy is ha when he leade s o he di e en g oups engage in dialogue, his signals o he g oup
membe s ha he e hnic ou g oup is no so h ea ening and dispu es can be esol ed amicably wi hou
ecou se o iolence. Rydg en e al. ha e conduc ed a s udy in I aq whe e hey ind ha people who
spend ime in e hnic he e ogenous spaces a e mo e likely o de elop iendship ies ac oss e hnic
bounda ies, a e mo e ole an owa ds people o o he e hnic g oups, and epo highe le els o
in e e hnic us .
29
Simila ly, Kanas e al., elying upon la ge-N su ey da a collec ed om Muslim and
Ch is ian s uden s in he Philippines and Indonesia, ha e ound ha in e eligious iendships educe
hos ili y owa ds eligious ou g oups.
30
These indings a e cong uen wi h he a gumen o Allpo
who con ends ha “sepa a eness” heigh ens he isk o con lic because i leads o he exagge a ion o
in e g oup di e ences.
31
P esen -day Nige ia emains pola ized along e hnic and eligious lines.
32
Nige ians de ine hei
iden i y “by a ilia ion o eligious and e hnic g oups a he han he Nige ian s a e.”
33
The wo majo
con lic s ha ha e a aged Nige ia du ing he pas wo decades—i.e., he Boko Ha am insu gency and
he iolen clashes be ween nomadic pas o alis s and esiden communi ies—ha e aken a eligious
u n because o he dis inc e hno eligious iden i ies o he con lic ac o s. Despi e he pe sis ence o
iolen con lic s in Nige ia, no s udy, o he bes o my knowledge, has examined how hese con lic s
in luence hos ili y owa ds e hnic and eligious ou g oups using ep esen a i e su ey da a o
Nige ia’s popula ion and econome ic echniques. Mo eo e , no s udy has examined he he e ogenous
e ec s o iolen con lic on ou g oup hos ili y among Nige ia’s wo majo eligious g oups—i.e.,
Ch is ians and Muslims. This s udy does so.
To measu e ou g oup hos ili y, I de eloped an addi i e indica o by combining he esponses o
wo su ey i ems p obing he esponden s’ willingness o ha e people om a di e en eligion and
people om a di e en e hnic g oup as neighbo s. To measu e exposu e o iolen con lic , I d ew
bu e s wi h a adius o hi y kilome e s a ound he esponden s’ dwellings using QGIS so wa e and
coun ed he o al numbe o iolen con lic s wi hin hem. I was able o do ha because I elied on da a
ob ained om A oba ome e
34
and he A med Con lic Loca ion and E en s Da abase (ACLED),
35
bo h o which a e geo e e enced. Causal iden i ica ion s emmed om ins umen ing con lic exposu e
wi h o es co e . The eg ession esul s show ha among he Nige ian popula ion and among
Ch is ians, exposu e o iolen con lic has a posi i e e ec on ou g oup hos ili y. A plausible
mechanism behind his inding is ha he h ea o iolen con lic s eng hens ing oup cohesion,
e odes us in ou g oup membe s, and makes in e g oup bounda ies salien . This is especially so
when he opposi e pa y o he con lic cons i u es a dis inc cul u al ou g oup. The main con lic
a ec ing Ch is ians in ol es nomadic pas o alis s o Fulani e hnici y, who a e Muslims. Among
Muslims, iolen con lic had a weak posi i e e ec on ou g oup hos ili y ha was no obus o
al e na i e ope a ionaliza ions o ou g oup hos ili y. A possible eason o he null e ec among
Muslims is ha he main con lic a ec ing hem— he Boko Ha am insu gency—does no in ol e
Ch is ians. Many Muslims ha e also been a ec ed by con lic s in ol ing nomadic Fulani pas o alis s.
This s udy con ibu es o he b oade li e a u e on in e g oup ela ions in he shadow o iolen
con lic .
36
The subsequen sec ions a e o ganized as ollows: In he second sec ion, I discuss he end
o iolen con lic s in Nige ia, a e which I e iew he li e a u e on he nexus be ween con lic and
social cohesion. Nex , I ope a ionalize he a iables ha will be used o es ima e he eg ession models
and discuss he empi ical s a egy; I hen p esen he eg ession esul s and discuss hem, a e which I
summa ize he pape and conclude.
Violen con lic s in Nige ia
Nige ia has wi nessed a lo o iolen con lic s du ing he pas wo decades. Da a om ACLED
37
shows
ha Nige ia had a o al o 18,781 inciden s be ween 1997 o 2022, which makes i he coun y wi h he
hi d highes incidence o iolen con lic in A ica.
38
Only Somalia and he Democ a ic Republic o
Congo pe o med wo se. These inciden s caused 98,877 a ali ies. The dis ibu ion o iolen con lic
TERRORISM AND POLITICAL VIOLENCE 241
inciden s a ies ac oss Nige ia’s wo egions: 68 pe cen o hem occu ed in No he n Nige ia while
he emaining 32 pe cen occu ed in he Sou he n Region. The con lic s a e also sp ead une enly
ac oss he yea s, wi h 9 pe cen o hem occu ing be ween 1997 o 2008, and he emaining 91 pe cen
occu ing be ween 2009 o 2022.
The wo majo con lic s a ec ing Nige ia a e he Boko Ha am insu gency and he iolen clashes
be ween Muslim nomadic pas o alis s o Fulani e hnici y and esiden communi ies (especially hose
in ol ed in c op cul i a ion). A epo by he Ins i u e o Economics and Peace no ed: “In Nige ia,
e o is ac i i y is domina ed by Fulani ex emis s and Boko Ha am. Toge he , hey accoun o
78 pe cen o e o - ela ed inciden s and 86 pe cen o dea hs om e o ism.”
39
(p. 21) The incidence
o iolen con lic in Nige ia can oughly be b oken down in o wo epochs: p e- and pos -Boko Ha am
e as. The p e-Boko Ha am e a co e s he pe iod om 1997 o 2008 be o e he adical Islamis g oup,
Boko Ha am, s a ed i s insu gency. The pos -Boko Ha am e a co e s he yea s om 2009 onwa ds
a e Boko Ha am launched i s i s a ack. The Boko Ha am insu gency ushe ed Nige ia in o a phase
o iolence i had ne e wi nessed. The ACLED da a shows ha be ween 2009 o 2022, he e we e 4,776
inciden s whe e a leas one o he pa ies o he con lic was Boko Ha am. These inciden s caused
a o al o 43,019 a ali ies. Because Boko Ha am a acks a e concen a ed in No heas e n Nige ia
whe e he popula ion is p edominan ly Muslim (see Figu e 1), mos o he a ali ies om hese a acks
a e Muslims.
Figu e 1. Inciden s in ol ing Boko Ha am and nomadic Fulani pas o alis s (1997–2022). The igu e shows he adminis a i e
bounda ies o he s a es ha cons i u e Nige ia’s No he n and Sou he n Regions. The ed do s show he geoloca ions o con lic s
whe e a leas one o he ac o s is Boko Ha am. The blue do s show he geoloca ions o con lic s whe e a leas one o he ac o s is
a “Pas o alis ” o belongs o he “Fulani” e hnic g oup. Vi ually all he ac o s de ined as pas o alis s in he ACLED da ase a e
iden i ied as “Fulani E hnic mili ia,” which makes he wo e ms almos synonymous. Al hough No he n Nige ia has a p edominan ly
Muslim popula ion, he e a e a ew s a es he e like Benue and Pla eau, whe e he popula ion is p edominan ly Ch is ian and
Muslims cons i u e a mino i y. These wo s a es, which we e no cap u ed by he Muslim jihadis s in he ea ly nine een h cen u y,
ha e he highes incidence o con lic s in ol ing nomadic Fulani pas o alis s. The shape iles con aining Nige ia’s adminis a i e
bounda ies was de eloped by UNOCHA.
242 D. TUKI
Nige ians end o associa e Muslims wi h ex emism. The Round 7 A oba ome e su ey con-
duc ed in 2017, and which is ep esen a i e o Nige ia’s popula ion, had a ques ion whe e espon-
den s we e asked abou he deg ee o which hey hough Muslims suppo ed ex emis g oups.
40
26 pe cen o hem chose he “none” esponse ca ego y, 37 pe cen chose he “some o hem” esponse
ca ego y, 24 pe cen chose he “mos o hem” esponse ca ego y, 7 pe cen chose he “all o hem”
esponse ca ego y, while he emaining 6 pe cen e used o answe he ques ion. This sugges s ha
68 pe cen o Nige ians associa e Muslims wi h ex emism a leas o some deg ee. Disagg ega ing he
da a based on eligious a ilia ion e ealed ha compa ed o Muslims, Ch is ians a e mo e likely o
associa e Muslims wi h ex emism: 84 and 48 pe cen o Ch is ians and Muslims espec i ely asso-
cia ed Muslims wi h ex emism a leas o some deg ee.
The iolen clashes be ween nomadic Fulani pas o alis s and esiden communi ies a e he second
majo con lic a ec ing Nige ia. This con lic , which is p ima ily caused by inc eased compe i ion o e
land and wa e esou ces due o d ough s, has quickly aken a eligious u n because o he dis inc
e hnic and eligious iden i ies o he opposing pa ies. Some epo s ha e po ayed con lic s in ol ing
pas o alis s as a acks on Ch is ians by Muslims because he pas o alis s a e Muslims and mos o he
communi ies whe e hese con lic s a e concen a ed ha e p edominan ly Ch is ian popula ions.
41
Relying on la ge-N su ey da a collec ed om Kaduna, he s a e wi h he hi d highes incidence o
a me -pas o alis con lic s in Nige ia, Tuki ound ha Ch is ians and Muslims iew he con lic
di e en ly: 52 pe cen o Ch is ians ag ee ha a me -pas o alis con lic s a e caused by eligion; only
17 pe cen o Muslims hold his iew.
42
The ACLED da a shows ha be ween 1997 o 2022, he e we e
2,416 iolen con lic s whe e a leas one o he ac o s was a pas o alis o belonged o he Fulani e hnic
g oup. These inciden s caused a o al o 15,333 a ali ies. As shown in Figu e 2, inciden s in ol ing
nomadic Fulani pas o alis s, unlike Boko Ha am a acks, a e sp ead ac oss all o Nige ia’s hi y-six
s a es. This is due o he mig a o y na u e o pas o alis s in sea ch o pas u e o hei li es ock.
Theo e ical conside a ions
Some s udies ha e shown ha exposu e o iolen con lic could os e social cohesion among
ing oup membe s. In a s udy conduc ed in Nepal, Gilligan e al. ound ha communi ies exposed
o iolen con lic had highe le els o ing oup us and p osocial beha io han hose ha we e
no .
43
The mechanism behind his inding was ha communi y membe s who we e no socially
o ien ed led he con lic zone lea ing behind hose who we e mo e socially o ien ed. Mo eo e ,
he common h ea posed by con lic p omp ed communi y membe s o band oge he so hey
could be e cope. Cal o e al. conduc ed a s udy in Mali whe e hey ound ha con lic exposu e
had a posi i e e ec on p osocial beha io .
44
Al hough hey acknowledged ha social cohesion
could os e pos -con lic eco e y, hey poin ed ou ha in he case o Mali his was p oblema ic
because inc eased social pa icipa ion was obse ed only in amily and e hnically homogenous
associa ions—i.e., “inwa d-looking associa ions.” This ein o ced kinship ies, made e hnic aul
lines salien , and heigh ened he isk o u he con lic . Rohne e al. had a simila inding in
a s udy conduc ed in Uganda whe e hey ound ha con lic exposu e s eng hened cohesion
wi hin e hnic ing oups.
45
Con lic has also been ound o e ode social cohesion. Weidmann and Zü che (3) ound ha
iolen con lic os e ed di isions in A ghan communi ies because i “could in oduce shi ing loyal ies
o he igh ing pa ies and hus in oduce new in e nal clea ages.”
46
Relying on su ey da a collec ed
om membe s o he Tamil e hnic g oup in S i Lanka, G eine and Filsinge ound ha men who had
been ic ims o sexual iolence du ing he S i Lankan Ci il Wa we e dis us ul o bo h membe s o
hei e hnic g oup and he e hnic ou g oup—i.e., he Sinhalese.
47
Con e sely, women who had been
ic ims o sexual iolence we e dis us ul o hei e hnic ing oup and had highe le els o us in he
e hnic ou g oup. They explained he e osion o ing oup us on he g ounds ha “ he con lic was
cha ac e ized by a clima e o dis us due o denuncia ions and be ayal wi hin Tamil communi ies
wi h ha m ul consequences o in-g oup cohesion” (2). Using ep esen a i e su ey da a o Pakis an,
TERRORISM AND POLITICAL VIOLENCE 243
Ahmad and Rehman ound ha exposu e o e o is a acks nega i ely co ela ed wi h in e pe sonal
us .
48
Rohne e al. had a simila inding in Uganda whe e hey ound ha con lic exposu e educed
gene alized social us .
49
In a s udy conduc ed in Nige ia, Tuki showed ha exposu e o con lic s
in ol ing nomadic Fulani pas o alis s led o dis us in bo h membe s o he Fulani e hnic g oup and
Muslims.
50
This was because he Fulani pas o alis s we e Muslims and he popula ion con la ed Fulani
e hnici y wi h being Muslim. Simila ly, Kanas e al., in a s udy conduc ed among Muslim and
Ch is ian s uden s in Indonesia and he Philippines ound ha he expe ience o in e eligious
iolence leads o hos ili y owa ds eligious ou g oups.
51
When he pe pe a o s o iolence belong o a dis inc cul u al ou g oup (e.g., based on e hnici y o
eligion), ing oup membe s migh associa e he en i e ou g oup wi h iolence e en i only a ew o hem
we e in ol ed in he ac , a phenomenon ha Hall e al. e e ed o as he “be e sa e han so y
app oach.”
52
This is associa ed wi h he concep o p ejudice which Allpo (7) de ined as “an a e si e o
hos ile a i ude owa ds a pe son o g oup, simply because he belongs o ha g oup, and is he e o e
p esumed o ha e he objec ionable quali ies asc ibed o he g oup.”
53
In a simila ein, Lickel e al.
de eloped a heo y o explain he psychological mechanisms unde lying e ibu i e iolence.
54
Vica ious
e ibu ion, hey obse ed, “occu s when a membe o a g oup commi s an ac o agg ession owa d
membe s o an ou g oup o an assaul o p o oca ion ha had no pe sonal consequences o him o he ,
bu did ha m a ellow ing oup membe .”
55
They poin ed ou ha when an ac o agg ession occu ed,
people who we e no di ec ly in ol ed in he con lic ied o make sense o i by cons uing i in e ms o
he b oade ing oup-ou g oup dicho omy be ween he con lic ac o s. I an ing oup-ou g oup dis inc-
ion was salien , hey would hen in e p e he e en in a way ha was a o able owa d hei ing oup and
encou aged e alia ion agains membe s o he ou g oup. Howe e , when ing oup-ou g oup dis inc ions
could no be ex apola ed om he ini ial ac o agg ession, people we e likely o in e p e i as a pe sonal
dispu e be ween wo indi iduals. This educed he likelihood o e alia ion.
Ahmed has shown how he e o is a ack ha occu ed in he U.S. on Sep embe 11, 2001
al e ed pe cep ions owa ds B i ish Muslims in he U.K.
56
The ensuing “Wa on Te o ” policy
shi ed he B i ish go e nmen ’s ocus om he di e se Asian iden i y o B i ish Muslims o
hei eligious iden i y, which po ayed hem as a “suspec communi y” and associa ed hem
wi h e o ism. As she concisely pu i , “i is he Muslim in B i ish Muslim which now shapes
he conc e e policies which go e n B i ish Muslims.”
57
Fe we da e al. conduc ed an expe i-
men al s udy in he U.S. whe e hey ound ha he associa ion o Muslim e ugees wi h
e o ism educed suppo o e ugee ese lemen bo h wi hin he U.S. and wi hin he
communi ies whe e he pa icipan s esided.
58
Thei analysis also showed ha exposing
subjec s o coun e ames ha challenged he po ayal o e ugees as h ea s had no
s a is ically signi ican e ec on suppo o e ugee ese lemen . This indica es ha nega i e
a i udes owa ds cul u al ou g oups, once o med, end o pe sis .
In a s udy conduc ed in Kenya, Schu e e al. ound ha indisc imina e iolence caused ea o
eligious ou g oups, s eng hened ing oup cohesion, and led o inc eased calls o esiden ial
seg ega ion along eligious lines.
59
Mo eo e , hey ound ha a acks pe pe a ed by Islamis
insu gen s led o dis us in Muslims. In ano he s udy conduc ed in India, Schu e e al. ound
ha con lic no only caused p ejudice owa ds eligious ou g oups and s eng hened ing oup
cohesion, bu also inc eased suppo o ex emis ac i i ies pe pe a ed by ing oup membe s.
60
Using expe imen s, Obaidi e al. ha e shown ha he pe cei ed cul u al h ea posed by Muslims
leads o inc eased suppo o he pe secu ion o he Muslim ou g oup among he Swedish and
Danish popula ions.
61
They also ound a simila e ec among Muslims who iew Wes e n cul u e
as decaden and a h ea o Islam. Con e sely, Whi e al. conduc ed an expe imen al s udy in
Sy ia, Bosnia and Koso o whe e hey ound ha hos ile a i udes owa ds ou g oups end o
change ollowing p oduc i e in e ac ions be ween he wo g oups.
62
This is consis en wi h he
p emise o he con ac hypo hesis pu o h by Allpo , which asse s ha in e g oup con ac ,
condi ional upon coope a ion owa ds a common goal and equali y be ween he g oups, educes
p ejudice.
63
244 D. TUKI
Re u ning o he Nige ian case, I expec con lic exposu e o ha e a posi i e e ec on
ou g oup hos ili y, especially because o how pola ized he coun y’s popula ion is along e hnic
and eligious lines, coupled wi h he huge impo ance ha Nige ians a ach o hei e hno eli-
gious iden i ies. This acili a es he cons uc ion o ing oups and ou g oups. Howe e , he e
migh be he e ogenous e ec s among Ch is ians and Muslims: Among Muslims, i is likely ha
exposu e o iolen con lic would ha e no e ec on hos ili y owa ds e hno eligious ou g oups.
This is because he main con lic a ec ing Muslims—i.e., he Boko Ha am insu gency—does no
in ol e Ch is ians. Mo eo e , a signi ican numbe o Muslims a e a ec ed by he iolen clashes
in ol ing nomadic Fulani pas o alis s who a e also Muslims. The common eligion o Islam
be ween he con lic ac o s hus makes i di icul o Muslims o es ablish ing oup-ou g oup
dis inc ions. Among Ch is ians, howe e , con lic exposu e is likely o ha e a posi i e e ec on
ou g oup hos ili y because he majo con lic a ec ing hem in ol es nomadic Fulani pas o alis s
who a e Muslims. Because he con lic ac o s belong o di e en eligious g oups, i becomes
easy o es ablish ing oup-ou g oup dis inc ions. Mo eo e , nomadic Fulani pas o alis s end o
be pe cei ed as a “suspec communi y” wi h a high p edisposi ion owa d iolence.
64
I will es
he ollowing hypo heses:
Hypo hesis 1: Among Nige ians, con lic exposu e leads o hos ili y owa ds e hno eligious
ou g oups.
Hypo hesis 2: Among Ch is ians, con lic exposu e leads o hos ili y owa ds e hno eligious
ou g oups.
Hypo hesis 3: Among Muslims, con lic exposu e has no e ec on hos ili y owa ds e hno eligious
ou g oups.
Da a and me hodology
This s udy elies on he Round 7 A oba ome e su ey da a
65
collec ed in 2017.
66
The da ase consis s
o 1,600 obse a ions and is ep esen a i e o Nige ia’s popula ion. Responden s we e d awn om
each o Nige ia’s hi y-six s a es and he ede al capi al e i o y—Abuja. O Nige ia’s 774 local
go e nmen a eas (LGAs) (i.e., municipali ies), da a we e collec ed om 147 o hem. Responden s
we e a leas eigh een yea s old, wi h males and emales equally ep esen ed in he sample. Table A1 in
he appendix epo s he summa y s a is ics o he a iables ha we e used o es ima e he eg ession
models.
Ope a ionaliza ion o he a iables
Dependen a iable
Ou g oup hos ili y. This is an addi i e indica o ha measu es he esponden s’ willingness o
ha e people om o he eligions and o he e hnic g oups as neighbo s. I was de i ed by
combining he esponses o he ollowing wo ques ions: “Fo each o he ollowing ypes o
people, please ell me whe he you would like ha ing people om his g oup as neighbo s,
dislike i , o no ca e: (a) People o a di e en eligion? (b) People om o he e hnic g oups?”
The esponses we e measu ed on a i e-poin o dinal scale anging om “1 = s ongly like,” o
“5 = s ongly dislike.” The addi i e indica o anges om 2 o 10, wi h highe alues deno ing
a highe le el o ou g oup hos ili y and ice e sa.
67
I ea ed he “don’ know” and “ e used
o answe ” esponses as missing obse a ions. I applied his ule o all a iables de i ed om
he A oba ome e su ey.
TERRORISM AND POLITICAL VIOLENCE 245
The wo su ey i ems had a C onbach Alpha s a is ic o 0.84, which shows in e nal
eliabili y. The wo i ems also had a co ela ion o 0.72, which highligh s he close associa ion
be ween e hnici y and eligion in Nige ia. As shown in he i s wo ba cha s om he op o
Figu e 2, Nige ians ha e a sligh ly highe le el o hos ili y owa ds eligious ou g oups han
e hnic ou g oups. Ch is ians a e sligh ly mo e hos ile owa ds people o a di e en e hnic
g oup han Muslims. Muslims a e sligh ly mo e hos ile owa ds people o a di e en eligion
han Ch is ians. Bo h Ch is ians and Muslims a e mo e hos ile owa ds people o a di e en
eligion han people o a di e en e hnic g oup.
Explana o y a iable
Violen con lic . This measu es he o al numbe o iolen con lic inciden s wi hin he hi y-
kilome e bu e a ound he esponden s’ dwellings. I de eloped he bu e s using QGIS
so wa e. This was possible because I elied upon da a ob ained om A oba ome e
68
and
ACLED,
69
bo h o which a e geo e e enced. Based on he ACLED da ase , I de ine a iolen
con lic as any inciden ha alls unde any o he ollowing h ee ca ego ies: Ba les, Violence
agains ci ilians, and Explosions/Remo e iolence.
70
Al hough he ACELD da ase is a ailable
s a ing om 1997 and is upda ed in eal ime, I excluded con lic inciden s ha occu ed
a e 2016. This lags he explana o y a iable since he dependen a iable is measu ed in
2017. I conside ed all he con lic inciden s wi hin he bu e om 1997 o 2016 because I am
pa icula ly in e es ed in he cumula i e e ec o iolen con lic . Some s udies ha e shown
ha memo ies om pas con lic s end o pe sis and could shape ac ion in he p esen .
71
Bu e s a e a mo e e icien way o measu ing exposu e o iolen con lic han he LGA
adminis a i e bounda ies. This is because he spa ial a ea occupied by each bu e is unique o
each esponden and allows o mo e a ia ion in he con lic exposu e a iable. I I had measu ed
con lic exposu e a he LGA le el, I would ha e associa ed all he esponden s esiding wi hin
a pa icula LGA wi h he o al numbe o con lic inciden s he e, which p esumes ha all
esponden s esiding wi hin a pa icula LGA a e exposed o he same le el o iolen con lic .
This would ha e been ine icien because inciden s in a con iguous LGA migh be nea e o
a esponden ’s dwelling han hose in he pa icula LGA whe e he/she esides. As shown in
Figu e 3, he esponden esides in Asa LGA, ye con lic s in Mo o, Olo unsogo, and O i I e LGAs
a e close o his/he dwelling han some inciden s in Asa LGA. Ano he challenge ha comes
along wi h wo king wi h Nige ia’s adminis a i e bounda ies (especially hose a he lowe le els)
is ha hey a e no clea ly de ined. In ac , he e we e a ew obse a ions whe e esponden s
esiding close o Nige ia’s na ional bo de we e mo e exposed o con lic s in he con iguous
Figu e 2. Hos ili y owa ds e hnic and eligious ou g oups. The y-axis shows he o al numbe o esponden s in he ull sample, and
he numbe o Muslim/Ch is ian esponden s who had answe ed he ele an ques ions ega ding hei willingness o ha e people
om a di e en eligion and e hnic g oup as neighbo s. The x-axis shows he pe cen age o esponden s who chose a pa icula
esponse ca ego y.
246 D. TUKI
Wu-Hausman s a is ics we e bo h insigni ican , which sugges s ha endogenei y was no p esen and
he use o an ins umen al a iable app oach o es ima e he model was inapp op ia e. I hus e-
es ima ed he model using OLS eg ession. As shown in model 3, iolen con lic emained s a is ically
insigni ican . This sugges s ha among Muslims, con lic exposu e has no e ec on hos ili y owa ds
people o a di e en eligion. This is consis en wi h Hypo hesis 3.
In models 4, 5, and 6, he dependen a iable measu es hos ili y owa ds e hnic ou g oups only. In
model 4 which was es ima ed using he Ch is ian subsample o esponden s, iolen con lic ca ied
he expec ed posi i e sign and was signi ican a he one pe cen le el, which indica es ha among
Ch is ians, exposu e o iolen con lic has a posi i e e ec on hos ili y owa ds e hnic ou g oups. The
size o he coe icien does no di e much om ha in model 1, which u he highligh s he close
associa ion be ween e hnici y and eligion in Nige ia. In model 5, which was es ima ed using he
Muslim subsample o esponden s, iolen con lic was s a is ically insigni ican . Because he Du bin
and Wu-Hausman s a is ics we e bo h s a is ically insigni ican , I e-es ima ed he model using OLS
eg ession. As shown in model 6, iolen con lic emained s a is ically insigni ican , which indica es
ha among Muslims, con lic exposu e has no e ec on hos ili y owa ds e hnic ou g oups. This
suppo s Hypo hesis 3.
Conclusion
This s udy examined he e ec o exposu e o iolen con lic on hos ili y owa ds e hno -
eligious ou g oups among he Nige ian popula ion and among i s wo majo eligious g oups
(i.e., Ch is ians and Muslims). Causal iden i ica ion s emmed om ins umen ing con lic
Table 3. E ec o iolen con lic on ou g oup hos ili y II (Full sample)
Ou g oup hos ili y
ϕ
Religion E hnici y
(1) (2) (3) (4)
Violen con lic
#
0.002** 0.014*** 0.002** 0.011***
(0.001) (0.004) (0.001) (0.004)
Nigh ime ligh
#
−0.236*** −0.194***
(0.066) (0.059)
P e alence o s un ing
#
2.149*** 1.618***
(0.687) (0.62)
Household dep i a ion 0.007 0.015*
(0.009) (0.008)
Log Popula ion size
#
−0.006 −0.00
(0.076) (0.068)
Educa ional le el −0.098*** −0.092***
(0.022) (0.02)
Religious a ilia ion 0.176 0.22*
(0.132) (0.119)
Gende −0.217*** −0.124*
(0.073) (0.066)
Age −0.006** −0.003
(0.003) (0.003)
Cons an 2.552*** 1.98** 2.286*** 1.738**
(0.072) (0.977) (0.065) (0.882)
Es ima ion me hod 2SLS 2SLS 2SLS 2SLS
Obse a ions 1439 1408 1438 1407
R-squa ed 0.116 0.118
Du bin s a is ic 9.083*** 12.01*** 6.578** 9.91***
Wu-Hausman s a is ic 8.95*** 11.787*** 6.470** 9.711***
Sa gan s a is ic 0.317 0.05 0.018 0.163
Basmann s a is ic 0.311 0.049 0.018 0.158
ϕ is he dependen a iable, # deno es a iables measu ed using bu e s wi h a adius o 30 kilome e s,
s anda d e o s a e in pa en hesis, *** p < 0.01, ** p < 0.05, * p < 0.10. All models a e es ima ed using wo-
s age leas squa es (2SLS) eg ession. All models con ain ixed e ec s o he esponden s’ e hnic g oups. Only
he second-s age eg essions a e epo ed he e.
TERRORISM AND POLITICAL VIOLENCE 253
exposu e wi h o es co e . The eg ession esul s showed ha among he Nige ian popula ion
and Ch is ians, con lic exposu e had a obus posi i e e ec on ou g oup hos ili y.
A plausible explana ion o his inding is ha he h ea o iolen con lic os e s cohesion
wi hin ing oup membe s, e odes us in ou g oup membe s, and makes in e g oup bound-
a ies salien . This is especially so when he opposi e pa y o he con lic cons i u es a dis inc
cul u al ou g oup. Because he main con lic a ec ing Ch is ians in ol es nomadic Fulani
pas o alis s who a e Muslims, i becomes easy o ing oup-ou g oup dis inc ions o be es ab-
lished. Among Muslims, iolen con lic had a weak posi i e e ec on ou g oup hos ili y ha
was obus o di e en ope a ionaliza ions o ou g oup hos ili y. This null e ec among
Muslims is likely because he main con lic a ec ing hem—i.e., he Boko Ha am insu gency
—does no in ol e Ch is ians. Mo eo e , a signi ican numbe o Muslims a e a ec ed by
con lic s in ol ing nomadic Fulani pas o alis s. The common eligion o Islam sha ed by he
pa ies makes he es ablishmen o ing oup-ou g oup bounda ies a duous, and makes i
illogical o Muslims o be hos ile owa ds Ch is ians.
The eg ession esul s also showed ha eligion is closely associa ed wi h e hnici y in
Nige ia, and he popula ion ends o con la e he wo. This is p oblema ic because i makes
in e g oup bounda ies mo e salien , which in u n heigh ens he isk o con lic . I he
Nige ian go e nmen in ends o educe iolen con lic and ou g oup hos ili y, i would
ha e o adop a policy ha ackles hese wo ac o s simul aneously because each one
ein o ces he o he . Fo ins ance, he go e nmen could educe he incidence o iolen
con lic by equipping i s secu i y agencies wi h he equisi e skills and equipmen needed o
espond p omp ly and e ec i ely o con lic si ua ions, while simul aneously pu suing policies
ha os e social cohesion and ele a e a sha ed na ional iden i y o e e hnic and eligious
Table 4. E ec o iolen con lic on ou g oup hos ili y III (Religious subsamples)
Ou g oup hos ili y
ϕ
Religion E hnici y
(1) (2) (3) (4) (5) (6)
(X ian) (Muslim) (Muslim) (X ian) (Muslim) (Muslim)
Violen con lic
#
0.025*** 0.004 0.001 0.021*** 0.003 0.001
(0.007) (0.005) (0.001) (0.006) (0.005) (0.001)
Nigh ime ligh
#
−0.464*** −0.05 −0.02 −0.393*** −0.051 −0.018
(0.113) (0.067) (0.022) (0.105) (0.061) (0.02)
P e alence o s un ing
#
−2.17** 2.738** 2.163*** −0.836 1.807 1.182**
(1.044) (1.356) (0.623) (0.972) (1.235) (0.565)
Household dep i a ion −0.001 0.022* 0.023* 0.015 0.019* 0.019*
(0.012) (0.013) (0.013) (0.011) (0.011) (0.012)
Log Popula ion size
#
0.111 −0.039 0.018 0.134 −0.027 0.035
(0.115) (0.143) (0.079) (0.108) (0.13) (0.072)
Educa ional le el 0.019 −0.109*** −0.102*** −0.02 −0.089*** −0.081***
(0.031) (0.029) (0.025) (0.029) (0.026) (0.022)
Gende −0.019 −0.514*** −0.517*** 0.062 −0.39*** −0.393***
(0.097) (0.101) (0.102) (0.09) (0.092) (0.092)
Age −0.005 −0.005 −0.004 −0.003 −0.00 0.001
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
Cons an 1.262 2.269 1.777* 0.729 2.079 1.546
(1.866) (1.458) (1.041) (1.737) (1.326) (0.945)
Es ima ion me hod 2SLS 2SLS OLS 2SLS 2SLS OLS
Obse a ions 807 601 601 806 601 601
R-squa ed 0.187 0.198 0.008 0.124 0.142
Du bin s a is ic 14.998*** 0.23 13.32*** 0.329
Wu-Hausman s a is ic 14.639*** 0.221 12.973*** 0.316
Sa gan s a is ic 1.232 4.236** 1.958 1.142
Basmann s a is ic 1.182 4.103** 1.88 1.1
ϕ is he dependen a iable, # deno es a iables measu ed using bu e s wi h a adius o hi y kilome e s, s anda d e o s a e in
pa en hesis, *** p < 0.01, ** p < 0.05, * p < 0.10. All models a e es ima ed using wo-s age leas squa es (2SLS) eg ession, excep o
models 3 and 6 which a e es ima ed using o dina y leas squa es (OLS) eg ession. All models con ain ixed e ec s o he
esponden s’ e hnic g oups. Only he second-s age eg essions a e epo ed he e.
254 D. TUKI
iden i ies, e.g., by encou aging in e -e hnic and in e - eligious dialogue. Howe e , I mus also
acknowledge ha he la e ecommenda ion migh be di icul o achie e because i is no
uncommon o he Nige ian eli es o exploi he e hnic and eligious di isions among he
popula ion o poli ical gain.
Acknowledgmen s
An ea ly e sion o his pape was p esen ed a a colloquium o ganized by he Mig a ion In eg a ion and
T ansna ionaliza ion Resea ch Uni a he WZB Be lin Social Science Cen e , Ge many. I hank he pa icipan s o
hei eedback. Thanks o he handling edi o and wo anonymous e e ees o hei eedback. Thanks o Roisin C onin
o edi o ial assis ance.
Disclosu e s a emen
No po en ial con lic o in e es was epo ed by he au ho (s).
Da a a ailabili y s a emen
The da a unde lying his s udy a e a ailable in he Ha a d Da a e se. h ps://doi.o g/10.7910/DVN/0WOK6X
No es on con ibu o
Daniel Tuki is a Resea ch Fellow a he WZB Be lin Social Science Cen e , Ge many. His esea ch ocuses on con lic
s udies and economic de elopmen .
ORCID
Daniel Tuki h p://o cid.o g/0000-0003-1097-3845
No es
1. Al hough Nige ia has 250 e hnic g oups, i has h ee majo ones: The Hausa/Fulani who a e p edominan ly
Muslim and mainly eside in No he n Nige ia. The Igbo and he Yo uba cons i u e he majo e hnic g oups in
Sou he n Nige ia. The Igbos a e p edominan ly Ch is ian, while he Yo uba is e enly spli be ween Muslims and
Ch is ians (D. D. Lai in, Hegemony and Cul u e: Poli ics and Religious Change Among he Yo uba (Chicago:
Uni e si y o Chicago P ess, 1986).
2. P. S. Hun ing on, The Clash O Ci iliza ions and he Remaking o Wo ld O de (New Yo k: Simon & Schus e ,
1996).
3. T. Falola and M. M. Hea on, A His o y o Nige ia (Camb idge: Camb idge Uni e si y P ess, 2008).
4. A. H. M. Ki k-G eene, The P inciples o Na i e Adminis a ion in Nige ia (London: Ox o d Uni e si y P ess,
1965).
5. Falola and Hea on, A His o y o Nige ia; A. F. Ogunsola, Legisla ion and Educa ion in No he n Nige ia (Ibadan:
Ox o d Uni e si y P ess, 1974).
6. I. O. Albe , “E hnic Residen ial Seg ega ion in Kano, Nige ia and i s An eceden s,” A ican S udy Monog aphs
17, no. 2 (1996): 85–100.
7. J. Campbell and M. T. Page, Nige ia: Wha E e yone Needs o Know (New Yo k: Ox o d Uni e si y P ess, 2018);
L. Diamond, Class, E hnici y and Democ acy in Nige ia: The Failu e o he Fi s Republic (London: Macmillan,
1988); J.S. Coleman, Nige ia: Backg ound o Na ionalism (Be keley: Uni e si y o Cali o nia P ess, 1958).
8. Coleman, Nige ia: Backg ound o Na ionalism.
9. L. Plo nico , “An Ea ly Nige ian Ci il Dis u bance: The 1945 Hausa-Ibo Rio in Jos,” Jou nal o Mode n A ican
S udies 9, no. 2 (1971): 297–305.
10. I. O. Albe , “Violence in Me opoli an Kano: A His o ical Pe spec i e,” in IFRA, U ban Violence in A ica
(Ibadan: Ins i u F ançais de Reche che en A ique, 1994), 111–138, h ps://books.openedi ion.o g/i a/788?
lang=en (accessed Oc obe 1, 2023).
TERRORISM AND POLITICAL VIOLENCE 255
11. C. Achebe, The e was a Coun y: A Pe sonal His o y o Bia a (London: Penguin Books, 2012); H. Ekwe-Ekwe,
Con lic and In e en ion in A ica: Nige ia, Angola, Zai e (New Yo k: Palg a e Macmillan, 1990).
12. J. Ib ahim, “The Poli ics o Religion in Nige ia: The Pa ame e s o he 1987 C isis in Kaduna S a e,” Re iew o
A ican Poli ical Economy 1, no. 45–46 (1989): 65–82.
13. Albe , “E hnic Residen ial Seg ega ion in Kano, Nige ia and i s An eceden s.”
14. H. Ange b and , “Poli ical Decen alisa ion and Con lic : The Sha ia C isis in Kaduna, Nige ia,” Jou nal o
Con empo a y A ican S udies 29, no. 1 (2011): 15–31; Human Righ s Wa ch, “The Miss Wo ld Rio s: Con inued
Impuni y o Killing in Kaduna,” Human Righ s Wa ch, 2003, h ps://www.h w.o g/ epo s/2003/nige ia0703/2.
h ml (accessed Augus 4, 2023).
15. H. Ange b and , “Deadly Elec ions: Pos -Elec ion Violence in Nige ia,” The Jou nal o Mode n A ican S udies 56,
no. 1 (2018): 143–167.
16. S. Eke, “Why does Seg ega ion P e en Con lic in Some Regions bu no O he s? In e oga ing Social Dis ance
Amid E hnic Con lic s in Jos, Nige ia,” Thi d Wo ld Qua e ly 43, no. 9 (2022): 2209–2224; J. K ause, “Gende
Dimensions o (non) Violence in Communal Con lic : The Case o Jos, Nige ia,” Compa a i e Poli ical S udies 52,
no. 10 (2019): 1466–1499; A.R. Mus apha, A. Higazi, J. La , and K. Ch omy, “Jos: Fea and Violence in Cen al
Nige ia,” in C eed and G ie ance: Muslim-Ch is ian Rela ions and Con lic Resolu ion in No he n Nige ia, eds. A.
R. Mus apha and D. Eh ha d (New Yo k: James Cu ey, 2018), 225–268; Human Righ s Wa ch, They do no Own
his Place: Go e nmen Disc imina ion Agains “Non-Indigenes” in Nige ia (New Yo k: Human Righ s Wa ch,
2006), h ps://www.h w.o g/si es/de aul / iles/ epo s/nige ia0406webwco e .pd (accessed Oc obe 11, 2023);
Human Righ s Wa ch, Jos: A ci y To n Apa (New Yo k: Human Righ s Wa ch, 2001), h ps://www.h w.o g/
si es/de aul / iles/ epo s/nige ia1201_1.pd (accessed Oc obe 11, 2023).
17. N. DeVo a, “F om E hnic Ou bidding o E hnic Con lic : The Ins i u ional Bases o S i Lanka’s Sepa a is Wa ,”
Na ions and Na ionalism 11, no. 1 (2005): 141–159; R. Ganguly, “S i Lanka’s E hnic Con lic : A a C oss oad
Be ween Peace and Wa ,” Thi d Wo ld Qua e ly 25, no. 5 (2004): 903–918; G. P adhan, “Economic cos o S i
Lanka’s E hnic Con lic ,” Jou nal o Con empo a y Asia 31, no. 3 (2001): 375–384.
18. G. Van Klinken and S. M. T. Aung, “The Con en ious Poli ics o An i-Muslim Scapegoa ing in Myanma ,”
Jou nal o Con empo a y Asia 47, no. 3 (2017): 353–375; M. Schissle , M. J. Wal on, and P. P. Thi, “Reconciling
Con adic ions: Buddhis -Muslim Violence, Na a i e Making and Memo y in Myanma ,” Jou nal o
Con empo a y Asia 47, no. 3 (2017): 376–395; N. Kipgen, “Con lic in Rakhine S a e in Myanma : Rohingya
Muslims’ Conund um,” Jou nal o Muslim Mino i y A ai s 33, no. 2 (2013): 298–310.
19. A. Mi a and D. Ray, “Implica ions o an Economic Theo y o Con lic : Hindu-Muslim Violence in India,”
Jou nal o Poli ical Economy 122, no. 4 (2014): 719–765; Z. Kausa , “Communal Rio s in India: Hindu–Muslim
Con lic and Resolu ion,” Jou nal o Muslim Mino i y A ai s 26, no. 3 (2006): 353–370.
20. C. K. Sha ma, “The Immig a ion Issue in Assam and Con lic s A ound i ,” Asian E hnici y 13, no. 3 (2012): 287–
309; A. T. Da nell and S. Pa ikh, “Religion, E hnici y, and he Role o he S a e: Explaining Con lic in Assam,”
E hnic and Racial S udies 11, no. 3 (1988): 263–281; S. Goswami, “E hnic Con lic in Assam,” The Indian Jou nal
o Poli ical Science 62, no. 1 (2001): 123–137.
21. T. Ge man, “The Nago no-Ka abakh Con lic be ween Aze baijan and A menia: Secu i y Issues in he
Caucasus,” Jou nal o Muslim Mino i y A ai s 32, no. 2 (2012): 216–229; B. Özkan, “Who Gains om he ‘No
Wa No Peace’ Si ua ion? A C i ical Analysis o he Nago no-Ka abakh Con lic ,” Geopoli ics 13, no. 3 (2008):
572–599.
22. Y. L. Fong and M. S. A. Ishak, “F aming In e e hnic Con lic in Malaysia: A Compa a i e Analysis o Newspape s
Co e age on he Ke is Polemics,” E hnici ies 14, no. 2 (2014): 252–278; K. K. Soong, “Racial Con lic in Malaysia:
Agains he O icial His o y,” Race & Class 49, no. 3 (2008): 33–53; E. K. B. Tan, “F om Sojou ne s o Ci izens:
Managing he E hnic Chinese Mino i y in Indonesia and Malaysia,” E hnic and Racial S udies 24, no. 6 (2001):
949–978.
23. Y. L. Fong and M. S. A. Ishak, “F aming In e -Religious Dispu e: A Compa a i e Analysis o Chinese-, English-,
and Malay-Language Newspape s in Peninsula Malaysia,” Chinese Jou nal o Communica ion 9, no. 2 (2016):
103–123.
24. M. Schaub, “Solida i y wi h a Sha p Edge: Communal Con lic and Local Collec i e Ac ion in Ru al Nige ia,”
A oba ome e , Wo king Pape No. 149, 2014; A. Scacco and S. S. Wa en, “Ch is ian-Muslim Rela ions in he
Shadow o Con lic : Insigh s om Kaduna, Nige ia,” in The Ox o d Handbook o Poli ics in Muslim Socie ies, eds.
M. Camme and P. Jones (New Yo k: Ox o d Uni e si y P ess, 2021),539–560; Albe , “E hnic Residen ial
Seg ega ion in Kano, Nige ia and i s An eceden s.”
25. A. B. Downes, “Mo e Bo de s, Less Con lic ? Pa i ion as a Solu ion o E hnic Ci il Wa s,” SAIS Re iew 26, no. 1
(2006): 9–61; J. Ti , “Di iding Coun ies o P omo e Peace: P ospec s o Long-Te m Success o Pa i ions,”
Jou nal o Peace Resea ch 42, no. 5 (2005): 545–562.
26. M. Bollig, “In a-and In e e hnic Con lic in No hwes Kenya: A Mul icausal Analysis o Con lic Beha iou ,”
An h opos 88, no. 1/3 (1993): 176–184.
27. J. M. La son, “In e e hnic Con lic and he Po en ial Dange s o C oss-G oup Ties,” Jou nal o Peace Resea ch 53,
no. 3 (2016): 459–471.
256 D. TUKI
28. S. Eke, “Why does Seg ega ion P e en Con lic in Some Regions bu no O he s?
29. J. Rydg en, D. So i, and M. Hälls en, “In e e hnic F iendship, T us , and Tole ance: Findings om Two No h
I aqi ci ies,” Ame ican Jou nal o Sociology 118, no. 6 (2013): 1650–1694.
30. A. Kanas, P. Scheepe s, and C. S e kens, “Posi i e and Nega i e Con ac and A i udes Towa ds he Religious
Ou -G oup: Tes ing he Con ac Hypo hesis in Con lic and Non-Con lic Regions o Indonesia and he
Philippines,” Social Science Resea ch 63 (2017): 95–110.
31. G. Allpo , The Na u e o P ejudice (Bos on: The Beacon P ess, 1954).
32. D. Tuki, “Pas o al Con lic s and (dis)T us : E idence om Nige ia using an Ins umen al Va iable App oach,”
WZB Be lin Social Science Cen e , Discussion Pape SP VI 2023–101, 2023; D. E. Agbiboa, “E hno-Religious
Con lic s and he Elusi e Ques o Na ional Iden i y in Nige ia,” Jou nal o Black S udies 44, no, 1 (2013): 3–30.
33. D. E. Agbiboa and B. Maiangwa, “Boko Ha am, Religious Violence, and he C isis o Na ional Iden i y in Nige ia:
Towa ds a Non-Killing App oach,” Jou nal o De eloping Socie ies 29, no. 4 (2013): 379–403.
34. A. BenYishay, R. Ro be g, J. Wells, Z. L , S. Goodman, L. Ko ace ic, and D. Run ola, “Geocoding A oba ome e
Rounds,” (2017) 1–6: Me hodology & Da a Quali y, “AidDa a,” h ps://www.aidda a.o g/publica ions/geocod
ing-a oba ome e - ounds-1-6-me hodology-da a-quali y.
35. C. Raleigh, A. Linke, H. Heg e, and J. Ka lsen, “In oducing ACLED—A med Con lic Loca ion and E en Da a,”
Jou nal o Peace Resea ch 47, no. 5 (2010): 651–660.
36. S. Schu e, C. Ruhe, and N. Sahool, “How Fea o Violence D i es In e g oup Con lic : E idence om
a Panel Su ey in India,” Te o ism and Poli ical Violence 35, no. 2 (2023): 229–247; S. Schu e, C. Ruhe,
and A. M. Linke, “How Indisc imina e Violence Fuels Con lic s Be ween G oups: E idence om Kenya,”
Social Science Resea ch 103 (2022): 102,653; D. Tuki, “Pas o al Con lic s and (dis)T us : E idence om
Nige ia Using an Ins umen al Va iable App oach, WZB Be lin Social Science Cen e ,” Discussion Pape
SP VI 2023–101, 2023; S. Whi , R. K. Wilson, and V. Mi ono a, “In e -G oup Con ac and Ou -G oup
Al uism A e Violence,” Jou nal o Economic Psychology 86 (2021): 1–19; T. Cal o, E. La allée,
M. Raza ind ako o, and F. Roubaud, “Fea no o Man? A med Con lic and Social Capi al in Malim,”
Jou nal o Compa a i e Economics 48, no. 2 (2020): 251–276; J. Fe we da, D. J. Flynn, and Y. Ho iuchi,
“Explaining Opposi ion o Re ugee Rese lemen : The ole o NIMBYism and Pe cei ed Th ea s,” Science
Ad ances 3, no. 9 (2017): 1–7.
37. C. Raleigh, A. Linke, H. Heg e, and J. Ka lsen, “In oducing ACLED—A med Con lic Loca ion and E en Da a,”
Jou nal o Peace Resea ch 47, no. 5 (2010): 651–660.
38. Based on he ACLED da ase , I de ine iolen con lic s as inciden s ca ego ized unde any o he ollowing h ee
ca ego ies: Ba les, Violence agains ci ilians, and Explosions/Remo e iolence. This implies ha I ha e excluded
inciden s ca ego ized as Rio s, P o es s, and S a egic de elopmen s.
39. Ins i u e o Economics and Peace, Global Te o ism Index 2019: Measu ing Peace in a Complex Wo ld (Sydney:
IEP, 2019), h p:// isiono humani y.o g/ esou ces (accessed July 5, 2023).
40. BenYishay e al., “Geocoding A oba ome e Rounds.”
41. Ch is ian Associa ion o Nige ia, “Nige ian Chu ch P o es Killing o Ch is ians,” CAN Nige ia, 2018, h ps://canng.
o g/news-and-e en s/news/173-nige ian-chu ch-p o es -killing-o -ch is ians (accessed July 5, 2023); Ch is ian
Associa ion o Nige ia, “Poli ical Opposi ion Sponso ship o Killings in Nige ia: Ques ions Buha i Go e nmen
Mus Answe ,” CAN Nige ia, 2018a, July 11, h ps://canng.o g/news-ande en s/news/181-poli ical-opposi ion-
sponso ship-o -killings-in-nige ia-ques ions-buha igo e nmen -mus -answe (accessed Augus 4, 2023).
42. D. Tuki, “Pas o al Con lic s and (dis)T us .”
43. M. J. Gilligan, B. J. Pasquale, and C. Samii, “Ci il Wa and Social Cohesion: Lab‐in‐ he‐Field E idence om
Nepal,” Ame ican Jou nal o Poli ical Science 58, no. 3 (2014): 604–619.
44. T. Cal o, E. La allée, M. Raza ind ako o, and F. Roubaud, “Fea No o Man? A med Con lic and Social Capi al
in Mali,” Jou nal o Compa a i e Economics 48, no. 2 (2020): 251–276.
45. D. Rohne , M. Thoenig, and F. Zilibo i, “Seeds o Dis us : Con lic in Uganda,” Jou nal o Economic G ow h 18,
no. 3 (2013): 217–252.
46. N.B. Weidmann and C. Zü che , “How Wa ime Violence A ec s Social Cohesion: The Spa ial-Tempo al
G a i y Model,” Ci il Wa s 15, no. 1 (2013): 1–18.
47. A. G eine and M. Filsinge , “(Dis) T us in he A e ma h o Sexual Violence: E idence om S i Lanka,”
Households in Con lic Ne wo k, Wo king Pape No. 377, 2022.
48. N. Ahmad and F. U. Rehman, “Does Te o ism Reduce T us ? Empi ical E idence om Pakis an,” De ence and
Peace Economics 33, no. 8 (2022): 993–1009.
49. D. Rohne , M. Thoenig, and F. Zilibo i, “Seeds o Dis us : Con lic in Uganda,” Jou nal o Economic G ow h 18,
no. 3 (2013): 217–252.
50. D. Tuki, “Pas o al Con lic s and (dis)T us .”
51. Kanas, Scheepe s, and S e kens, “Posi i e and Nega i e Con ac and A i udes Towa ds he Religious Ou -G oup.”
52. J. Hall, D. T. Kahn, E. Skoog, and M. Öbe g, “Wa Exposu e, Al uism and he Recalib a ion o Wel a e T adeo s
Towa ds Th ea ening Social Ca ego ies,” Jou nal o Expe imen al Social Psychology 94 (2021): 1–9.
53. Allpo , The Na u e o P ejudice.
TERRORISM AND POLITICAL VIOLENCE 257
54. B. Lickel, N. Mille , D. M. S ens om, T. F. Denson, and T. Schmade , “Vica ious Re ibu ion: The Role o
Collec i e Blame in In e g oup Agg ession,” Pe sonali y and Social Psychology Re iew 10, no. 4 (2006): 372–390.
55. Lickel, Mille , S ens om, Denson, and Schmade , “Vica ious e ibu ion.”
56. S. Ahmed, “B i ish Muslims Pe cep ions o Social Cohesion: F om Mul icul u alism o Communi y Cohesion
and he ‘Wa on Te o ,’” C ime, Law and Social Change 71, no. 5 (2019): 581–595.
57. Ahmed, “B i ish Muslims Pe cep ions o Social Cohesion.”
58. J. Fe we da, D. J. Flynn, & Y. Ho iuchi, “Explaining Opposi ion o Re ugee Rese lemen : The Role o NIMBYism
and Pe cei ed Th ea s,” Science Ad ances 3, no. 9 (2017): 1–7.
59. S. Schu e, C. Ruhe, and A. M. Linke, “How Indisc imina e Violence Fuels Con lic s be ween G oups: E idence
om Kenya,” Social Science Resea ch 103 (2022): 102653.
60. Schu e, Ruhe, and Sahoo, “How Fea o Violence D i es In e g oup Con lic .”
61. M. Obaidi, L. Thomsen, and R. Be gh, “They Think We a e a Th ea o hei Cul u e”: Me a-Cul u al Th ea Fuels
Willingness and Endo semen o Ex emis Violence agains he cul u al ou g oup. In e na ional Jou nal o
Con lic and Violence 12 (2018): 1–13.
62. S. Whi , R. K. Wilson, and V. Mi ono a, “In e -G oup Con ac and Ou -G oup Al uism a e Violence,” Jou nal
o Economic Psychology 86 (2021): 1–19.
63. See no e 53 abo e.
64. D. Tuki, “You’ e no Welcome! Violence and Suppo o a G azing Ban Policy in Kaduna, Nige ia.” Households
in Con lic Ne wo k, Wo king Pape , No. 397, 2023; P. F. Ejio o , “‘Fulanis a e Fo eign Te o is s’: The Social
Cons uc ion o a Suspec Communi y in he Sahel,” C i ical S udies on Te o ism 15, no. 2 (2022): 333–355;
S. Eke, “‘Nomad Sa age’ and He de -Fa me Con lic s in Nige ia: The (Un)making o an Ancien My h,” Thi d
Wo ld Qua e ly 41, no 5 (2020): 745–763.
65. See no e 34 abo e.
66. To access he A oba ome e da ase and he su ey ques ionnai e isi : h ps://www.a oba ome e .o g/.
67. In he o iginal A oba ome e da ase , highe o dinal alues deno e a lowe le el o ou g oup hos ili y and ice
e sa. Fo easy in e p e a ion o he eg ession esul s, I in e ed he o dinal alues assigned o he esponse
ca ego ies by sub ac ing each o hem om 6; his allows highe (lowe ) alues o deno e a highe (lowe ) le el o
ou g oup hos ili y.
68. See no e 34 abo e.
69. C L. Raleigh, A. Inke, H. Heg e, and J. Ka lsen, “In oducing ACLED—A med Con lic Loca ion and E en
Da a,” Jou nal o Peace Resea ch 47, no. 5 (2010): 651–660.
70. To access he ACLED da ase isi : h ps://acledda a.com/.
71. B. Wagone , and I. B escó, “Con lic and Memo y: The Pas in he P esen ,” Jou nal o Peace Psychology 22, no. 1
(2016): 3–4; B. Tin , “His o y, Memo y, and In ac able Con lic ,” Con lic Resolu ion Qua e ly, 27, no. 3 (2010):
239–256.
72. J. He bs , S a es and Powe in A ica: Compa a i e Lessons in Au ho i y and Con ol (P ince on Uni e si y P ess,
2000); P. Collie and A. Hoe le , “G eed and G ie ance in Ci il Wa ,” Wo ld Bank, Policy Resea ch Wo king
Pape No. 2355, 2000.
73. C. Xu, “E ec s o U baniza ion on T us : E idence om an Expe imen in he Field,” Jou nal o Economic
Psychology 87 (2021): 1–8; J. Delhey, and K. New on, “P edic ing C oss-Na ional Le els o Social T us : Global
Pa e n o No dic Excep ionalism? Eu opean Sociological Re iew 21, no. 4 (2005): 311–327.
74. To access he popula ion da ase isi : h ps://www.wo ldpop.o g/.
75. T. Ghosh, K. E. Baugh, C. D. El idge, M. Zhizhin, A. Poyda, and F. C. Hsu, “Ex ending he DMSP Nigh ime
Ligh s Time Se ies Beyond 2013,” Remo e Sensing 13, no. 5004 (2021): 1–19.
76. R. Flo ida, C. Mellande , and T. Gulden, “Global Me opolis: Assessing Economic Ac i i y in U ban Cen e s
based on Nigh ime Sa elli e Images,” The P o essional Geog aphe 64, no. 2 (2012): 178–187; C. Mellande ,
J. Lobo, K. S ola ick, and Z. Ma heson, “Nigh - ime Ligh Da a: A Good P oxy Measu e o Economic Ac i i y?”
PloS One 10, no. 10 (2015): 1–18; N.B. Weidmann, and S. Schu e, “Using Nigh Ligh Emissions o he
P edic ion o Local Weal h,” Jou nal o Peace Resea ch 54, no. 2 (2017): 125–140.
77. P. Collie , The Bo om Billion: Why he Poo es Coun ies a e Failing and Wha Can be Done Abou I (New Yo k:
Ox o d Uni e si y P ess, 2008).
78. A. J. S ewa , N. McCa y, and J. J. B yson, “Pola iza ion unde Rising Inequali y and Economic Decline,” Science
Ad ances 6, no. 50 (2020): 1–9; J. Delhey, and K. New on, “P edic ing C oss-Na ional Le els o Social T us :
Global Pa e n o No dic Excep ionalism?” Eu opean Sociological Re iew 21, no. 4 (2005): 311–327.
79. To access he nigh ime ligh da ase isi : h ps://eogda a.mines.edu/p oduc s/dmsp/.
80. C. Bosco, V. Alegana, T. Bi d, C. Pezzulo, L. Beng sson, A. So iche a, and A. J. Ta em, “Explo ing he High-
esolu ion Mapping o Gende -disagg ega ed De elopmen Indica o s,” Jou nal o The Royal Socie y In e ace 14,
no. 129 (2017): 1–12.
81. N. Von Uexkull, “Sus ained D ough , Vulne abili y and Ci il Con lic in Sub-Saha an A ica,” Poli ical
Geog aphy 43 (2014): 16–26; Q. T. Do and L. Iye , “Geog aphy, Po e y and Con lic in Nepal,” Jou nal o
258 D. TUKI
Peace Resea ch 47, no. 6 (2010): 735–748; P. Collie , The Bo om billion: Why he Poo es Coun ies a e Failing and
Wha can be Done Abou I (New Yo k: Ox o d Uni e si y P ess, 2008).
82. J. Ge eke, M. Schaub, and D. Baldassa i, “E hnic Di e si y, Po e y and Social T us in Ge many: E idence om
a Beha io al Measu e o T us ,” PloS One 13, no. 7 (2018): 1–15; A. Alessina, and E. La Fe a a, “The
De e minan s o T us , Na ional Bu eau o Economic Resea ch,” Wo king Pape No. 7621, 2000.
83. S. Go don, “Xenophobia Ac oss he Class Di ide: Sou h A ican A i udes Towa ds Fo eigne s 2003–2012,”
Jou nal o Con empo a y A ican S udies 33, no. 4 (2015): 494–509; D. E e a , “Xenophobia, S a e and Socie y in
Sou h A ica, 2008–2010,” Poli ikon 38, no. 1 (2011): 7–36; M. Choane, L. S. Shulika, and M. M hombeni, “An
Analysis o he Causes, E ec s and Rami ica ions o Xenophobia in Sou h A ica,” Insigh on A ica 3, no. 2
(2011): 129–142.
84. To access he p e alence o s un ing da ase isi : h ps://hub.wo ldpop.o g/geoda a/summa y?id=1268.
85. Including “p e alence o s un ing” and “household dep i a ion” in he same model did no lead o mul i-
collinea i y because he co ela ion be ween was −0.06.
86. Fe we da e al., “Explaining Opposi ion o Re ugee Rese lemen ”; A. T. Jenssen and H. Engesbak, “The Many
Faces o Educa ion: Why a e People wi h Lowe Educa ion mo e Hos ile Towa ds Immig an s han People wi h
Highe Educa ion?” Scandina ian Jou nal o Educa ional Resea ch 38, no. 1 (1994): 33–50.
87. P. Collie , and A. Hoe le , “G eed and G ie ance in Ci il Wa ,” Wo ld Bank, Policy Resea ch Wo king Pape
No. 2355, 2000.
88. M. Schaub, and D. Aue , “Rebel Rec ui men and Mig a ion: Theo y and E idence om Sou he n Senegal,”
Jou nal o Con lic Resolu ion (2022): 1–28; Q. T. Do and L. Iye , “Geog aphy, Po e y and Con lic in Nepal,”
Jou nal o Peace Resea ch 47, no. 6 (2010): 735–748.
89. B. Kayode, “Inside Nige ia’s Sambisa Fo es , he Boko Ha am Hideou whe e Kidnapped School Gi ls a e
Belie ed o be Held,” The Gua dian, Ap il 29, 2014, h ps://www. hegua dian.com/wo ld/2014/ap /29/nige ia-
sambisa- o es -boko-ha am-hideou -kidnapped-school-gi ls-belie ed- o-be-held (accessed July 5, 2023); B. G ill,
& T. Selande , “The De il in Nige ia: Boko Ha am’s Reign o Te o ,” De Spiegel, May 30, 2014, h ps://www.
spiegel.de/in e na ional/wo ld/boko-ha am-con inues- o- e o ize-no he n-nige ia-a-972282.h ml (accessed
July 5, 2023).
90. E. Akinwo u, “Almos 300 Schoolgi ls Kidnapped in Nige ia a e F ee, says S a e Go e no ,” The Gua dian,
Ma ch 2, 2021, h ps://www. hegua dian.com/wo ld/2021/ma /02/almos -300-schoolgi ls-kidnapped-in-nige ia-
a e- ee-says-s a e-go e no (accessed July 5, 2023).
91. To access he GlobCo e da ase and he codebook/ alida ion epo isi : h p://due.es in.esa.in /page_glob
co e .php; S. Bon emps, P. De ou ny, E. Van Bogae , O. A ino, V. Kalogi ou, and J.R. Pe ez, GLOBCOVER 2009
P oduc Desc ip ion and Valida ion Repo (Pa is and Lou ain-la-Neu e: Eu opean Space Agency and Uni e si é
Ca holique de Lou ain, 2011).
92. W. Chow and E. Han, “Rugged Te ain, Fo es Co e age, and Insu gency in Myanma ,” Con lic Managemen
and Peace Science, (2023): 1–20.
93. I also es ima ed some 2SLS eg ession models whe e I used ei he o es co e o i s squa e as
ins umen al a iables. The second-s age eg ession esul s we e consis en wi h hose epo ed in his
pape .
94. Table A2 in he appendix shows he e hnic dis ibu ion o he esponden s in he A oba ome e da ase .
95. M. Pa sons, “Mili an Religion, no Jus Clima e Change, is Fuelling Violence in Nige ia,” The C i ic,
2023, h ps:// hec i ic.co.uk/nige ias-clima e-o - e o / (accessed July 8, 2023); Ch is ian Associa ion o
Nige ia, “Nige ian Chu ch P o es Killing o Ch is ians,” CAN Nige ia, 2018, h ps://canng.o g/news-and
-e en s/news/173-nige ian-chu ch-p o es -killing-o -ch is ians (accessed July 5, 2023); Ch is ian
Associa ion o Nige ia, “Poli ical Opposi ion Sponso ship o Killings in Nige ia: Ques ions Buha i
Go e nmen mus Answe ,” CAN Nige ia, 2018a, July 11, h ps://canng.o g/news-ande en s/news/181-
poli ical-opposi ion-sponso ship-o -killings-in-nige ia-ques ions-buha igo e nmen -mus -answe
(accessed Augus 4, 2023).
TERRORISM AND POLITICAL VIOLENCE 259
Appendix
Table A2. E hnic dis ibu ion o esponden s
E hnic g oup F equency (n) Pe cen
Hausa 322 22.24
Igbo 251 17.33
Yo uba 328 22.65
Fulani 49 3.38
Ibibio 35 2.42
Kanu i 35 2.42
Ijaw 33 2.28
Ti 26 1.80
Ikwe e 25 1.73
E ik 24 1.66
Ebi a 20 1.38
Idoma 19 1.31
Nupe 18 1.24
Igala 16 1.10
Isoko 10 0.69
Edo 10 0.69
Gwa i 9 0.62
Kalaba i 9 0.62
Jukun 7 0.48
U hobo 4 0.28
Bi om 3 0.21
Shuwa-A ab 1 0.07
O he s 194 13.41
To al 1,448 100.00
Based on he Round 7 A oba ome e su ey da a collec ed
in 2017.
Table A1. Desc ip i e s a is ics
Va iable Obs. Mean S d. De . Min Max
Ou g oup hos ili y
ϕ
1437 4.389 2.351 2 10
Ou g oup hos ili y ( eligion) 1439 2.261 1.331 1 5
Ou g oup hos ili y (e hnici y) 1438 2.13 1.202 1 5
Violen con lic
#
1592 67.886 106.807 0 475
Nigh ime ligh
#
1592 3.1 5.191 0 20.104
P e alence o s un ing
#
1592 0.331 0.147 0.136 0.634
Household dep i a ion 1440 4.935 4.11 0 20
Log Popula ion size
#
1592 14.032 1.116 11.536 16.39
Educa ional le el 1445 4.513 2.155 0 9
Religious a ilia ion 1428 0.569 0.495 0 1
Gende 1448 0.501 0.5 0 1
Age 1447 32.658 12.428 18 80
Fo es co e 1592 0.599 0.223 0.056 0.985
(Fo es co e )
2
1592 0.408 0.26 0.003 0.969
ϕ is he dependen a iable which is de i ed by adding “Ou g oup hos ili y ( eligion)” and “Ou g oup hos ili y (e hnici y),” #
deno es a iables measu ed using bu e s wi h a adius o hi y kilome e s. Al hough he A oba ome e da ase has 1,600
po en ial obse a ions, he a iables in he able con ain ewe obse a ions because no all esponden s we e asked he
ele an ques ions. Also, I ea ed “don’ know” and “ e used o answe ” esponses as missing obse a ions which may ha e
exace ba ed he p oblem o lis wise dele ion.
260 D. TUKI
Table A3. Co ela ion be ween he a iables
Va iables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
(1) Ou g oup hos ili y 1.000
(2) Ou g oup hos ili y ( eligion) 0.938 1.000
(3) Ou g oup hos ili y (e hnici y) 0.925 0.736 1.000
(4) Violen con lic −0.118 −0.126 −0.092 1.000
(5) Nigh ime ligh −0.168 −0.172 −0.140 0.837 1.000
(6) P e alence o s un ing 0.098 0.109 0.072 −0.374 −0.397 1.000
(7) Household dep i a ion 0.049 0.030 0.062 −0.048 −0.068 −0.066 1.000
(8) log popula ion size −0.078 −0.086 −0.059 0.643 0.768 −0.500 −0.094 1.000
(9) Educa ional le el −0.158 −0.150 −0.144 0.250 0.270 −0.378 −0.102 0.332 1.000
(10) Religious a ilia ion −0.038 −0.045 −0.024 0.141 0.207 −0.649 0.058 0.358 0.369 1.000
(11) Gende −0.074 −0.083 −0.052 −0.002 −0.011 0.011 0.028 −0.012 0.096 −0.042 1.000
(12) Age −0.007 −0.025 0.013 0.013 −0.017 0.009 0.040 0.007 −0.175 −0.028 0.124 1.000
(13) Fo es co e −0.100 −0.094 −0.093 −0.094 −0.084 −0.318 0.060 0.054 0.079 0.298 0.003 0.056 1.000
(14) (Fo es co e )
2
−0.113 −0.107 −0.103 −0.120 −0.096 −0.301 0.042 0.075 0.062 0.289 0.000 0.069 0.982 1.000
TERRORISM AND POLITICAL VIOLENCE 261
