Hunge , Sophia
A icle — Published Ve sion
Vi uous people and e il eli es? The ole o mo alizing
ames and no ma i e dis inc ions in iden i ying populis
discou se
Poli ical Resea ch Exchange
P o ided in Coope a ion wi h:
WZB Be lin Social Science Cen e
Sugges ed Ci a ion: Hunge , Sophia (2024) : Vi uous people and e il eli es? The ole o mo alizing
ames and no ma i e dis inc ions in iden i ying populis discou se, Poli ical Resea ch Exchange,
ISSN 2474-736X, Taylo & F ancis, London, Vol. 6, Iss. 1, pp. 1-26,
h ps://doi.o g/10.1080/2474736X.2024.2370308
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/308502
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/
RESEARCH ARTICLE
Vi uous people and e il eli es? The ole o mo alizing ames
and no ma i e dis inc ions in iden i ying populis discou se
Sophia Hunge
a,b
a
Uni e si y o B emen/SOCIUM, B emen, Ge many;
b
WZB Be lin Social Science Cen e , Be lin, Ge many
ABSTRACT
Populism has been a much-s udied concep in Social Science esea ch.
A g ea sha e o esea ch on he concep has been dedica ed o
measu ing he occu ence o populism in poli ical ex . In his
a icle, I p opose a no el measu e o populism ha conside s
es ablished heo e ical assump ions, i.e. p esen ing he people as
mo ally supe io and he eli e as e il. This mo al aming o he
an agonis ic g oups is necessa y o iden i y populis discou se while
keeping i sepa a e om empi ically ela ed concep s, e.g. adical
igh -wing ideology. The no el wo-s ep dic iona y de ec s mo ally-
amed e e ences o bo h g oups. I apply his app oach o a ex
co pus o all speeches gi en in he Eu opean Pa liamen om 1999
o 2014 and ca y ou ex ensi e alidi y checks. Taking he
mo alizing no ion o populism mo e se iously no only con ibu es
o ou heo e ical unde s anding o populis discou se and i s
impac on he poli ical sphe e, bu he new measu e also imp o es
p e ious app oaches o measu ing populism.
ARTICLE HISTORY
Recei ed 13 No embe 2023
Accep ed 13 June 2024
KEYWORDS
Populism, pa ies, Eu opean
pa liamen , ex -as-da a,
mo aliza ion
In oduc ion
Populism has become a buzzwo d (Hunge and Pax on 2022): I is used a ibu ed o poli-
icians o all kinds; is said o be con agious o mains eam pa ies; and supposedly a ec s
ci izens in hei poli ical beha iou . Pa icula ly i we a e in e es ed in how populis dis-
cou se a ec s o he ac o s, e.g. pa ies and o e s, we need a heo e ically in o med
way o measu e populism on a la ge scale. Many in luen ial s udies (e.g. Jage s and
Walg a e 2007; Rooduijn and Pauwels 2011) ha e laid impo an g oundwo k o he
endea ou o measu ing populis discou se ac oss la ge quan i ies o ex . D awing on
hese s udies, his a icle uses ex -as-da a me hods o a new measu emen o populis
discou se, which emphasizes he mo al no ion o populism.
Theo e ical accoun s o populism s ess he cen ali y o he mo ally-cha ged Mani-
chean dis inc ion be ween he people and he eli es. Tha is, populis s e e o he
people by po aying hem as homogeneous, supe io and good, while he eli es a e p e-
sen ed as co up , e il and degene a ed. While heo e ically well-es ablished, his mo al
© 2024 The Au ho (s). Published by In o ma UK Limi ed, ading as Taylo & F ancis G oup
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 .
CONTACT Sophia Hunge [email p o ec ed]
POLITICAL RESEARCH EXCHANGE
2024, VOL. 6, 2370308
h ps://doi.o g/10.1080/2474736X.2024.2370308
dis inc ion has no a elled o quan i a i e measu emen s o populism. This poses
se e al p oblems, as no only populism, bu also hick ideologies a e based on a di e en-
ia ion be ween he people and an ou -g oup, e.g. in he case o na ionalism he people
unde s ood as ‘na i es’ s. immig an s as ‘non-na i es’. This example is pa icula ly c ucial,
as populis adical igh pa ies expe ienced a leap o suppo in ecen yea s in Eu ope. I
a gue ha his no ma i e dis inc ion be ween he wo g oups needs o be aken mo e
se iously in measu ing populism and in he s udy o i s consequences, e.g. i s e ec on
o he pa ies, o e s, o he media
In o de o g ound my populism measu emen ho oughly in he heo y, I p opose a
wo-s ep dic iona y app oach. The i s s ep employs wo baseline dic iona ies, one o
‘ he people’ and one o ‘ he eli e’, which consis o mo ally neu al wo ds o bo h
g oups, such as ‘eli e’, ‘poli icians’, ‘ci izens’, o ‘ axpaye s’. These baseline dic iona ies
a e used o de ec neu al men ions o each o he g oups and o ex ac ing he
sec ion o snippe s o he speeches in which hese men ions occu . In he second s ep,
hese snippe s a e hen analysed using ame dic iona ies, again one o each g oup.
The ame dic iona ies con ain wo ds and ph ases which can be used o p esen he
g oups as mo ally supe io , i.e. he people, o as mo ally co up ed, i.e. he eli e.
My app oach comes wi h se e al ad an ages. Fi s , by using a di e en measu emen o
an i-eli ism and people-cen ism, i allows o assessing how hese di e en ea u es o popu-
lism a e used by di e en ac o s. Second, by aking he mo al dis inc ion be ween he people
and he eli es in o accoun , i p o ides a way o measu ing populism ha is less p one o be
con la ed wi h e e ences o he people in na ionalis , socialis , o o he ways. Thi d, a e
ha ing es ablished and alida ed my measu e, I compa e my discou se-based measu es o
ca ego iza ions o populis pa ies based on expe s in o de o assess i ‘populis is wha
populis s do’. Following Fo es Gump’s amous dic um o ‘S upid is as s upid does’, I
assess whe he pa ies, ha a e conside ed as populis use mo e populis he o ic based
on my measu e and es whe he he e a e di e ences based on adicali y and ideology.
Las ly, he combina ion o base and ame dic iona ies p o ides a s aigh - o wa d app oach
o quan i a i e ame analysis mo e gene ally. This migh be o in e es o a b oade ange o
schola ship beyond populism esea ch, o ins ance o assessing how speci ic g oups o
people, e.g. emale poli icians, expe s, e c., a e p esen ed in di e en se s o ex s.
Populism as a mo alized discou se
The idea ional de ini ion o populism
The concep o populism has o en been c i icized o lacking concep ual cla i y and is
subjec o i id academic deba es abou a wo hwhile de ini ion. Howe e , se e al schola s
(Aslanidis 2016, 89; Bonikowski e al. 2019, 62; Hawkins and Ro i a Kal wasse 2017, 527;
Mudde and Ro i a Kal wasse 2018, 2–3; Rooduijn 2019, 363) ha e no ed a end
owa ds ag eeing on an idea ional de ini ion. These app oaches sha e some concep ions
o populism: (1) he an agonism be ween he people and he eli es as populism’s co e
concep , (2) he idea ha populism lacks s ong p og amma ic appeals and is mos o en
combined wi h ‘ ull’ ideologies, as well as (3) a ‘deg eeis ’ unde s anding o populism.
While he ‘idea ional camp’ includes schola s ha de ine populism as ‘ aming de ice’
(Bonikowski e al. 2019, 62), ‘discu si e ame’ (Aslanidis 2016, 98), ‘communica ion s yle’
2 S. HUNGER
(B ubake 2017a, 2017b), ‘communica ion phenomenon’ (V eese e al. 2018), o ‘wo ld-
iew’ (Hawkins 2009), mos amously ph ased by Mudde (2004, 543) desc ibing populism
‘ hin ideology’ ha :
conside s socie y o be ul ima ely sepa a ed in o wo homogeneous and an agonis ic g oups,
he ‘pu e people’ and ‘ he co up eli e’, and which a gues ha poli ics should be an
exp ession o he olon è gènè ale (gene al will) o he people.
A u he common denomina o o di e en idea ional de ini ions is he unde s anding o
populism as discu si e ea u e a he han a ixed policy o ien a ion: ‘unlike a [ hick] ideol-
ogy, populism is a la en se o ideas o a wo ld iew ha lacks signi ican exposi ion and
(…) and is usually low on policy speci ics’ (Hawkins 2009, 6). While ull ideologies esul in
p og amma ic o ien a ions o pa ies, he populism becomes isible h ough he use o a
speci ic discou se (Hawkins 2010, 10).
Closely connec ed o he idea ional unde s anding o populism is he deba e a ound
he ‘deg ee-ism’ o populism. Aslanidis (2016, 96) p oposes o conside populism as
g adual a he han dicho omous since his is ‘mo e o less, how he concep has been
ope a ionalised in he g owing quan i a i e li e a u e’ (Be nha d, K iesi, and Webe
2015; see e.g. Jage s and Walg a e 2007; Reungoa 2010; Rooduijn and Pauwels 2011).
This allows o a ying deg ees o populis discou se ac oss di e en poli ical ac o s
and con ex s. Simila ly, Cammack (2000, 155) makes a case o a he using he adjec i e
‘populis ’ ins ead o he noun as a ‘quali ie o some subs an i e poli ical p ojec ’. Deegan-
K ause and Haugh on (2009, 822) also sugges o shi ‘ou unde s anding o populism
(…) o a desc ip ion o pa y appeals a he han pa ies hemsel es’. Addi ionally, ea -
ing populism solely as g aded allows o a mo e nuanced s udy o he phenomenon and
dec eases no ma i e biases and he compulsion o academics o ake sides in an essen-
ialis s uggle (Aslanidis 2016).
The deg eeis unde s anding o populism is cen al o his s udy. I allows o de elop-
ing a measu emen ha is independen om ‘ex e nal’ ca ego iza ions o ac o s in o a
bina y measu e o populis and non-populis . On he o he hand, i makes i c ucial o
de elop a measu emen which is s ongly based on heo e ical assump ions ins ead o
elying on a ‘we know i when we see i ’-logic.
Bo h he opposi ion o he ‘e il’ eli es and he p aise o he ‘good’ people a e well-
es ablished co e ea u es o populis discou se. While nea ly all ideologies appeal o
‘ he people’ in a speci ic way, e.g. socialism o na ionalism, he mo al dis inc ion
be ween he people and he eli es is essen ial o populism. This mo al ele a ion o he
people, howe e , equi es a e e ences ca ego y, an ‘o he ’, an an i-pole. This e e ence
ca ego y is embodied by he ‘mo ally degene a ed’ eli es and hence comple es his
dyadic ela ionship. In he ollowing sec ion, I discuss ha e e ences o bo h g oups
and hei mo alized aming a e inex icable ma ke s o populism, ha a e bo h necessa y
bu no su icien (see also Table 1).
Table 1. Mo ali y and an agonis ic g oups as ma ke s o populis discou se.
Mo alized language
An agonis ic g oups: eli e s. people
Yes No
Yes populis no su icien
No no su icien no populis
POLITICAL RESEARCH EXCHANGE 3
The cen ali y o mo al aming o populis discou se
Ac o s, who employ populis discou se, need o disc imina e be ween he people and he
eli es and hey do so by p esen ing one side as pu e, good and i uous and he o he side
as e il, co up and o en. Va ious au ho s ha e emphasized his no ma i e dis inc ion
be ween bo h g oups as cen al o populism, as i is de ined as ‘a Manichean and mo a-
lis ic discou se ha di ides socie y in o el pueblo and oliga chy […]’ (De la To e 2000, 4).
In a simila ein, Mülle (2017, 19–20) de ines populism as ‘a pa icula mo alis ic imagin-
a ion o poli ics, a way o pe cei ing he poli ical wo ld ha se s a mo ally pu e and ully
uni ied (…) people agains eli es who a e deemed co up o in some o he way mo ally
in e io ’. Wiles (1969, 167) connec s he idea ional na u e o populism and i s mo al
no ion s a ing ha populism ‘is mo alis ic a he han p og amma ic’. Simila ly, Panizza
(2005, 22) a gues ha populis s subs i u e a mo al discou se o a poli ical discou se in
o de o be able ‘ o alk poli ics while denouncing i as a di y game’. Tha is, by employing
a mo alis ic discou se ha di e en ia es hem om o he poli ical ac o s, hey can be pa
o poli ics while condemning he whole poli ical sphe e and i s ac o s. By using uni e sal
abs ac ions, hey ‘con as he high mo al g ounding o [ hei ] message wi h he co up-
ion and be ayal o he poli ical es ablishmen ’ (Panizza 2005, 22-23), In his mo alis ic
logic, he eli es, e.g. o he poli ical ac o s, a e p esen ed as enemies who a e no ‘sha ing
a common symbolic space wi hin which he con lic akes place’ (Mou e 2000, 20). This
makes comp omise o e en dialogue wi h he opponen impossible which is unaccep able
o populis pa ies (De la To e 2000). Panizza (2005, 22–23) also poin s ou ha mo al
di ides a e used by populis s in o de o disquali y ad e sa ies he eby making legi ima e
dissen impossible. Mudde dedica es a signi ican sha e o his chap e on he idea ional
app oach o populism o mo ali y, which he calls ‘ he essence o he populis di ision’
(2017, 29). Thus, pu i y and au hen ici y a e o him a he co e o he dis inc ion be ween
he wo g oups. This mo alis ic dis inc ion is, howe e , ied o he p esence o an i-
eli ism and people-cen ism, in o de o be conside ed as populism, as Table 1 shows.
Despi e his mul i ude o schola s emphasizing he mo alized no ion o populism,
S a akakis and Jäge (2017, 12) a gue ha mo ali y is ‘ he shaky basis’ o ‘ he “new” main-
s eam in con empo a y populism s udy’ which is bo h unde -de ined and oo b oad o be
a dis inc ea u e o populis he o ic. Mo eo e , hey (S a akakis and Jäge 2017, 13–14)
c i icize ha he mo ally cha ged dis inc ion be ween ‘e il’ and ‘good’ has been employed
by poli icians o mani old ideologies, especially as i is ‘mo e o less una oidable in any
poli ical conjunc u e, and especially p onounced in c ucial u ning poin s’. They
p o ide se e al exempla y quo es by poli icians, e.g. by Tha che : ‘I am in poli ics
because o he con lic be ween good and e il, I belie e ha in he end good will
iumph’. While his s a emen ce ainly includes mo al elemen s, i alls sho o quali ying
as populis as he an agonis ic di ide. Tha che e e s o ‘good’ and ‘bad’ in an unspeci ic
way, which is om e e ences o people o eli es. Mo alized language is ce ainly
employed by mani old ac o s (Jaspe 1992; see e.g. Jung 2019) and pu i y and idealiza ion
may be used o a ious poli ical a gumen s (e.g. du ing he ecen poli ical c isis in
Eu ope, see Na o zky (2016)), bu S a akakis and Jäge (2017) neglec he necessa y com-
bina ion o mo alized aming wi h he e e ences o he people and he eli es. In his
ega d, hei c i ique o mo ali y is unde -speci ied and does no apply o a sha ed heo -
e ical unde s anding o populism.
4 S. HUNGER
The an agonism be ween he people and he eli es
Howe e , mo ali y and mo alized discou se alone a e no su icien ma ke s o populis
discou se, he same applies o sole e e ences o he people and he eli es. I is no
jus ba e e e ences o ‘ he people’ and ‘ he eli es’ ha cha ac e ize populism, hese
g oups a e p esen ing and aming hem in a mo ally cha ged way. Fi s , ‘appeals o
he people’ as ma ke a e no explici enough abou how ‘ he people’ a e pe cei ed by
populis s. Re e ing o he people is a ai ha is sha ed by poli icians om nea ly all pol-
i ical camps and colou . In poli ical discou se, a ious unde s andings o ‘ he people’ may
be a play: o ins ance, in na ionalis ic, economic, cul u al, o poli ical e ms (Meny and
Su el 2002). Radical igh -wing ac o s p esen he ‘people as a na ion’, while adical le
ac o s a he emphasize class (K iesi 2014, 362). Wha se s populism apa is how hese
g oups a e cha ged wi h meaning which is co e o his ideology (Zehn e 2023). While
a ious cons uc ions o he people can be ound in mode n democ acies, I a gue ha
he cons uc ion o he people as mo ally supe io , hones and i uous lies a he co e
o populism, which is hen o be con as ed by he ejec ion o he ‘e il eli es’.
Many au ho s (Hawkins 2009; e.g. Jage s and Walg a e 2007; Reungoa 2010; Rooduijn
and Pauwels 2011) ha e used ‘appeals o he people’ as iden i ie o people-cen ism
When measu ing appeals o he people, Jage s and Walg a e (2007, 339) s ess ha e e -
ences o he popula ion (g oup) as an insepa able uni y mos ly p eceded by a de ini e
a icle (e.g. he o e , he people, he consume ) ‘can be ega ded as he mos solid indi-
ca o o populism’. Reungoa (2010) ope a ionalizes people-cen ism as he equen use
o ‘ e ms ha e e ed o he popula ion (popula ion as a whole o popula ion ca ego ies)’
bu does no speci y any mo al e alua ion o cons uc ion as supe io and he e o e migh
o e es ima e people-cen ism. Aslanidis (2017, 9) poin s ou ha wi h his app oach
‘people-cen ism ends o ge o e -coded, since all e e ences o ‘ he people’ a e coded
indisc imina ely in o his ca ego y’. When no conside ing his mo al aming o he an ag-
onis ic g oups, we migh a he be measu ing concep s o he han populism, e.g.
na ionalism.
The necessa y combina ion o e e ences o he an agonis ic g oups and a mo alized
aming o hese equi es ca e ul ope a ionaliza ion. In he nex sec ion, I shall p esen
my app oach which akes bo h ma ke s in o accoun .
The mo al dis inc ion be ween he people and he eli es – a ool o
iden i ying populis discou se?
Measu ing populism using dic iona ies
Populism has o en been measu ed in poli ical ex s o all kinds, since common
app oaches o measu ing ideology, such as using oll-call o es, a e no ans e able o
he concep o populism. Nex o ex -based measu es, expe su eys a e inc easingly
used o measu e populism in poli ical pa ies (see e.g. Meije s and Zaslo e 2021).
Howe e , he su eys a e no sui able o e ospec i ely assess how populism de eloped
(Le ick and Ola a ia-Gambi 2020).
Measu ing populism using au oma ed ex analysis has shown o be a edious ask, ha
comes – due o he concep ual complexi y o populism – wi h many pi alls and chal-
lenges (Jankowski and Hube 2023). Recen ad ances in he social sciences ha e also
POLITICAL RESEARCH EXCHANGE 5
spa ked app oaches o measu es populism using supe ised machine lea ning. Di Cocco
and Monechi (2022) and Celico, Rode, and Rod iguez-Ca eño (2024) use ex e nally-
de ined (bina y and con inuous) measu es o pa ies’ populism and ain classi ie s on
his in o ma ion. This pa y-based app oach, howe e , does no allow o dis inguish
be ween populism and o he pa y cha ac e is ics, such as communica ion s yle o
hick ideology, in he p edic ion and is hus la gely de ached om a heo e ically-based
concep ion o populism (see also Jankowski and Hube 2023).
E ha d e al. (2023) and Bonikowski, Luo, and S uhle (2022) use ans o me models in
o de o measu e pa ies’ le el o populism, in he Ge man Bundes ag and US elec o al
campaign espec i ely. Weigh ing be ween dic iona ies and supe ised machine-lea ning
is s ongly ied o he le el o anspa ency he classi ica ion should en ail.
Weigh ing be ween dic iona ies and supe ised machine-lea ning is s ongly ied o
he le el o anspa ency he classi ica ion should en ail. Supe ised machine-lea ning
app oaches a e o en black-boxes o esea che s and eade s alike and ‘due o [ hei ]
complexi y [ hey] lack anspa ency o alida ing he esul s su icien ly. Hence, inhe -
en ly in e p e able models should be p e e ed o highly complex asks, simila o he
ask a hand’ (Rudin 2019). Since I aim o implemen he heo e ically well-es ablished
cen ali y o a mo al aming in o a alid measu emen o populism, o me anspa ency
ou weighs he bene i s o ans o me models. While o en ea u e sco es a e no dis-
cussed in machine-lea ning app oaches, using a dic iona y model gi es us ull anspa -
ency and con ol o which ea u es a e ied o ou concep o in e es . I belie e his is
pa icula ly aluable in my case, whe e my con ibu ion also aims a showing he impo -
ance o mo ali y in o de o measu e populism.
Some o he a o emen ioned s udies use dic iona ies o iden i y and measu e populis
discou se (Bonikowski and Gid on 2016; Rooduijn and Pauwels 2011; see also G ündl
2022, o an applica ion B eye 2022). Dic iona ies d aw on wo d equencies ha a e
used in a espec i e ex . They a e designed o cap u e ca ego ies o a heo e ical
concep ‘by alloca ing wo ds o hese ca ego ies using a combina ion o a p io i and
empi ical c i e ia’ (La e and Ga y 2000, 626). Applying a dic iona y, i.e. a lis o heo e i-
cally mo i a ed keywo ds, hen esul s in coun s o hese keywo ds’ occu ences in
di e en ex s. By assessing he equency in which he keywo ds a e used, i measu es
o which ex en a ce ain concep is p e alen in a documen (G imme and S ewa
2013, 8). Hence, designing such a dic iona y equi es a deep unde s anding o he
concep a hand, as i is only alid wi h a p ecise choice o keywo ds and i migh o he -
wise esul in nume ous alse posi i es (Ruedin and Mo ales 2017).
Fu he mo e, quali a i e, in-dep h assessmen and alida ion a e c ucial in o de o
assess he numbe o alse posi i es. Dele ing and adding keywo ds based on his,
howe e , may esul in a he ad hoc dic iona ies o ei he ex eme b e i y o excessi e
leng h (Aslanidis 2017). Hence, he e ms included in populism dic iona ies o en seem
oo b oad, e.g. class, poli ic*; p opaganda; e e end*; egime, people (Rooduijn and
Pauwels 2011), as he e is no s ong heo e ical eason why – o ins ance using
he wo d poli ics – quali ies o a populis s a emen . O he keywo ds seem e y a bi-
a y and ailo ed o a e y speci ic con ex , such as loophole and long nose (Bonikowski
and Gid on 2016), o dono s (Oli e and Rahn 2016). Many o hese wo ds a e no
heo e ically connec ed o he concep o populism and hence cap u e o he
phenomena.
6 S. HUNGER
I base my app oach on s ong heo e ical assump ions, i.e. he cen ali y o mo al
aming o populis discou se. I a gue ha his heo e ical ounda ion allows us o
choose keywo ds o he wo e e ence g oups and hei espec i e aming ha
pe mi s us o iden i y populis discou se. This app oach d aws on a communica ion-
cen ed a he han an ac o -cen ed unde s anding o populism (S anye , Salgado, and
S ömbäck 2016, 354). While he i s ocuses on he heo e ical cha ac e is ics o populis
communica ion, he la e uses he cha ac e is ics o populis poli ical ac o s as a s a ing
poin . The ac o -cen ed app oach assumes ha we al eady know who populis ac o s a e
and d aws conclusions abou populis communica ion based on hese ac o s’ poli ical
communica ion. Focusing on a p e-de ined se o populis ac o s and he way hey alk,
comes a isk, I a gue, o con la ing populis discou se wi h he p og amma ic p o ile o
he populis s’ hos ideologies. Fu he mo e, an ac o -based app oach is a odds wi h an
unde s anding o populism as a ma e o deg ee, as discussed in he p e ious sec ion.
I we use a se o ac o s as a heu is ic o measu e populis discou se, we discoun he
possibili y ha populis discou se can a y ac oss se ings, con ex s, issues and ac o s.
Iden i ying he mo al dis inc ion in populis discou se h ough a wo-s ep
dic iona y
In o de o use he no ma i e dis inc ion be ween he eli e and he people as a ool o
iden i ying populis discou se, I sugges using a no el wo-s ep dic iona y app oach,
ha allows o de ec ing whe he one o he g oups is e e ed o in a mo ally cha ged
way. Unlike mos p e ious s udies, I use sepa a e dic iona ies o cap u e he e e ence
o he wo g oups. This allows o measu ing people-cen ism and an i-eli ism, i.e. wo
co e ea u es o populism, sepa a ely and hus o assessing di e ences in he numbe
o e e ences ac oss ac o s. My wo-s ep dic iona y app oach is se up as ollows: Fi s , I
design wo baseline dic iona ies o e e ences o he people and he eli es. This allows
me o iden i y he e e ences o he wo g oups as dis inc ea u es and o show how
hey a e used by di e en pa ies as well as when hey occu oge he . S udying mic o-
le el populis a i udes, Schulz e al. (2017, 2) make a simila a gumen s a ing ‘ ha a uni-
dimensional model ails o adequa ely desc ibe populis a i udes, as i does no accoun
o he di e en poli ical ideas ha ha e been iden i ied as dis inc ye co ela ed ace s o
a populis ideology’. As a gued abo e, men ioning he eli es and he people is no ine-
g ained enough o measu e populis discou se. As e e ing o he people can be done in
neu al, na ionalis and many o he ways, I sugges ha he cons uc ion o he people as
mo ally supe io cap u es populism mo e p ecisely. Simila ly, no all e e ences o he eli e
a e necessa ily populis , i.e. nega i e, as poli icians could e e o he eli e posi i ely as
well. In o de o cap u e his mo al aming, I de elop wo addi ional ame dic iona ies
which aim o cap u e whe he he people a e p esen ed in a mo ally supe io and he
eli es in a mo ally degene a ed manne .
In he i s s ep, he baseline dic iona ies a e ailo ed o iden i y e e ences o he people
o he eli es (see Table 2). I c ea ed hem d awing on p e ious au oma ed app oaches and
codebooks o hand-coding (Bonikowski and Gid on 2018; Hamelee s and Vliegen ha
2019; Jage s and Walg a e 2007; Reungoa 2010; Rooduijn and Pauwels 2011). Nex , I
assessed he quali y o he wo dic iona ies by looking a a sample o appea ances o
e e y wo d in he people – and he eli e-dic iona y and judged whe he hey cap u e
POLITICAL RESEARCH EXCHANGE 7
he igh concep s. I excluded wo ds as alse posi i es i hey we e mos ly used o e e o
speci ic, na ow g oups o people o o he concep s ha a e no ela ed o ‘ he people’ o
‘ he eli es’. Fo ins ance, some s udies claim ha cas e is o en used in populis discou se
in o de o e e o he eli es (Rooduijn 2015). Howe e , mos o he occu ences o he
wo d cas e we e in ac e e ing o he Indian cas e sys em. Hence, he keywo d ‘cas e’
was emo ed om he dic iona y. Some keywo ds we e included using globing, i.e. wild-
ca d cha ac e s (depic ed by an as e isk *), hus allowing o any s ing ollowing he
speci ied keywo ds. This means ha he keywo d eli * will pick up eli e, eli ism, eli is ,
eli e- ocused, and so on. O he wo ds, e.g. esiden , a e only included in speci ic g amma-
ical o ms, as he keywo d wi h an as e isk o he wise would pick up wo ds such as ‘ esi-
den ial’ disp opo ionally. The inal se o wo ds o he base-dic iona ies is depic ed in
Table 2. Some o he keywo ds a e qui e b oad and neu al, e.g. ci izen, while o he s –
especially o he eli e – ca y a alue judgemen , o ins ance, c onies. I conside hese
mo ally cha ged keywo ds as su icien ma ke s o a mo al aming. These wo ds a e
highligh ed in bold in Table 2. Using he baseline dic iona ies, I ind a o al o 9817 e e -
ences o he eli e and 133,625 e e ences o he people. The o e ep esen a ion o e e -
ences o he people se es as an indica ion ha e e ing o he people is done by mos
poli icians in a mul i ude o ways, ha a e no populis . Table A2 in he appendix shows
he equency o he indi idual keywo ds o people and eli es in he co pus.
As I ha e a gued abo e, e e ences o he people and he eli es ough o be p esen ed
in a mo alized way in o de o be conside ed populis he o ic. To pe o m he nex s ep o
my analysis, i.e. de e mining i he wo g oups a e amed in a mo ally-cha ged way, I
selec ‘snippe s’ a ound he occu ences o he people- and he eli e-dic iona y. This
means, ha I c ea e a new ex co pus, ha does no consis o he ull speeches
anymo e, bu only o he pa s o he speeches in which we ind a e e ence o he
people o he eli es. Conside ing he leng h o he speeches and he sen ences wi hin
hem, I decided o selec a window o eigh wo ds be o e and a e each keywo d.
Addi ionally, he leng h o he ame window is based on he Ge man elec ion mani es os
since 1998. The Mani es o P ojec (Lehmann e al. 2023) uses he sou ce in o de o code
pa ies’ issue posi ion. The main coding uni in he p ojec a e so-called quasi-sen ences,
whe e one quasi-sen ence includes exac ly one posi ional s a emen on one issue. I
assessed he summa y s a is ics o he quasi-sen ences coded in all Ge man mani es os
in ha ime ame. The median leng h is 13 wo ds, he mean is 14.71 wo ds. I used his
Table 2. Keywo ds o e e ences o he people and he eli e.
People Eli e
ci izen popula ion’s appa a chik* eu ophil*
ci izen’s esiden a is oc a * inancial sha ks
ci izens esiden ’s bad apples* inancie *
ci izens’ esiden s banks ma ia
eu opeans esiden s’ bosses ma ia’s
eu opeans’ axpaye bosses’ old eli es
amilies axpaye ’s B ussels ma ia oliga ch*
amily axpaye s bu eauc a poli ical class*
man on he s ee * axpaye s’ ci cles* poli ician*
o dina y pe son* he public c onies powe monge
he public’s people uling ci cles eli *
people’s popula ion es ablishmen uling class
es ablishmen ’s sycophancy eu oc a*
8 S. HUNGER
al eady discussed abo e – a discussion on populism esea ch ha could be summa ized as
‘Populis is as populis does?’ O pu di e en ly: How do we de ine whe he an ac o o a
pa y is populis ? Gump’s answe implies ha a pe son should a he be judged by
ac ions and no he appea ance. T ans e ed o he s udy o pa ies and pa y ac o s, his
calls o pe cei ing populism as a ‘desc ip ion o pa ies’ appeals a he han pa ies hem-
sel es’ (Deegan-K ause and Haugh on 2009, 822) as al eady discussed in sec ion 2.
Table 5 shows he o al numbe o speeches gi en by MEPs o pa ies ha a e classi ied
as populis o non-populis ollowing Rooduijn e al. (2019). The h ee las columns show
he classi ica ion o populis speech based on my measu e, and he measu es o Bonikowski
and Gid on (2018) and Rooduijn and Pauwels (2011), sepa a ely o populis and non-popu-
lis pa ies. O all speeches in he sample, abou 20 pe cen we e gi en by ac o s belonging
o pa ies ha a e adi ionally conside ed populis . The i s di e ence ha s icks ou is he
misma ch ega ding he o al sha e o speeches classi ied as populis be ween he
measu es. While he dic iona y by Rooduijn and Pauwels (2011) hi s in nea ly 40 pe cen
o he speeches, my measu es classi y a ound 2.5 pe cen as populis and he app oach
o Bonikowski and Gid on (2018) classi ies only a ound 1.4 pe cen o all speeches as popu-
lis .
3
This se es as an addi ional indica ion ha Rooduijn and Pauwels (2011) app oach
d as ically o e es ima es populism. This is in line wi h he expec a ion o he EP as a
ha d es o populism. While he sha e o populis discou se is simila o Bonikowski and
Gid on’s (2018) app oach, hei small accu acy se es as an indica ion ha hey iden i y
di e en speeches han my app oach. Looking a he se o pa ies ha a e classi ied as
populis a p io i based on Rooduijn e al. (2019), all h ee measu es iden i y a highe
sha e o populis he o ic among hem compa ed o he pa ies ca ego ized as non-popu-
lis . Hence, i seems ha he pa ies gene ally seen as populis in academic li e a u e also
show highe le els o populis he o ic in he EP. Howe e , he di e ence be ween bo h
g oups is ai ly small, o in he case o Rooduijn and Pauwels (2011) e en negligible.
Co e ea u es – an i-eli ism and people-cen ism
Nex , I del e mo e in o he co e ea u es an i-eli sm and people-cen ism and b eak up my
measu emen . O he han many p e ious app oaches, ha ing sepa a e dic iona ies p o ides
he oppo uni y o s udy how hey ela e o each o he . In o de o assess how hei use
coincides, I p esen wo di e en igu es. Fi s , I show how o en bo h concep s a e p esen
in he e y same speech. Second, I show whe he ac o s who a e using people-cen ism also
use an i-eli ism, i.e. whe he he use o e e ences o bo h g oups is co ela ed wi hin ac o s.
Table 6 shows he p esence o an i-eli ism and people-cen ism in single speeches. The
wo-by- wo able shows ha people-cen ism is a ound six imes mo e common han
an i-eli ism. Addi ionally, he able allows o assess he co-occu ence o an i-eli ism
and people-cen ism wi hin a single speech. This is, howe e , an ex emely a e e en .
I appea s in 61 speeches in he whole co pus i.e. in only 0.03 pe cen o he speeches.
Table 5. Classi ica ion o speeches o di e en dic iona ies.
Populis pa y Own app oach Bonikowski & Gi don Rooduijn & Pauwels
No Coun 3623 2003 53,282
No P opo ion 0.025 0.014 0.371
Yes Coun 1148 880 13,715
Yes P opo ion 0.032 0.025 0.389
POLITICAL RESEARCH EXCHANGE 15
Nex , I u n o he MEP-le el, assessing whe he he e is any ela ionship be ween he
use o he wo co e ea u es an i-eli ism and people-cen ism wi hin indi iduals. Figu e 1
p esen s as sca e plo showing he pe cen age o speeches in which MEPs used he co e
ea u es o populism. The pe cen age o speeches including people-cen ism is shown on
he x-axis, while he y-axis shows he pe cen age o an i-eli ism. The black do s ep esen
MEPs who a e classi ied as populis a p io i ollowing Rooduijn e al. (2019), while each
g ey do ep esen s a non-populis MEP. Addi ionally, he g aph p esen s he i ed
lines o bo h g oups sepa a ely, again populis MEPs in black and non-populis MEPs
in g ey.
4
The g aph con i ms he indings al eady p esen ed abo e: people-cen ism is
a mo e used han an i-eli ism. Howe e , he e seems o be a posi i e co ela ion o
he usage o bo h co e concep s o populis pa ies, while his co ela ion is nea ly
absen o non-populis pa ies.
These indings speak o a concep ual issue which has spa ked a deba e among popu-
lism schola s: Do ma ke s o populism ha e o co-occu in he same speech o ex o is i
‘enough’ i an ac o uses bo h co e ea u es sepa a ely? Rooduijn, de Lange, and an de
B ug (2014, 567) a gue ha ‘i is he combina ion o people-cen ism and an i-eli ism ha
de ines populism. Only i a c i ique on he (bad) eli e coincides wi h an emphasis on he
(good) people, can we speak o populism’. Howe e , i hey e- un hei analyses emo ing
his condi ion, hei esul s a e obus .
Engesse e al. (2017) a gue o a ‘ agmen ed concep ’, i.e. he co e ea u es o popu-
lism do no necessa ily ha e o be p esen in one speech by a poli ician in o de o con-
side he speech populis . Along simila lines, E ns e al. (2017) show ha he co e
ea u es occu in a agmen ed manne , bu s ill emphasize he impo ance o combining
hem. The esul s p esen ed abo e suppo his iew, showing ha a co-occu ence o
bo h ea u es is a he a e in he EP, bu he use o bo h ea u es by a single MEP is co -
ela ed o MEPs belonging o pa ies ha a e classi ied as populis .
Populism le and igh ?
Las ly, I p esen e idence ha populism is mo e commonly used by pa ies on he inges
o he poli ical spec um. Se e al p e ious s udies (Be nha d and K iesi 2019; E ns e al.
2017; Rooduijn and Akke man 2017) ha e a gued and shown ha pa ies a he inges
o he poli ical sys em end o use populism mo e han hose posi ioned in he cen e.
Figu e 2 shows he use o an i-eli ism, people-cen ism, my combined measu es and
he measu es by Rooduijn and Pauwels (2011) and Bonikowski and Gid on (2018). The
pa ies a e clus e ed along hei le - igh dimension using he Pa lGo (Dö ing and
Manow 2016) le - igh measu es, whe e 0 deno es he le end and 10 he igh end
o he spec um. While Pa lGo p o ides he mos exhaus i e measu e o pa ies’
le - igh posi ion, I also include he ca ego y ‘NA’ in o de o show ha he missings
do no in oduce addi ional bias.
Table 6. C oss abula ion o an i-eli ism and people-cen ism being p esen in he same speech.
An i-eli ism People-cen ism
Yes No
Yes 61 594
No 4055 174,195
16 S. HUNGER
Panel A clea ly shows ha a - igh pa ies use an i-eli ism o he la ges ex en , ol-
lowed by a -le pa ies. Unsu p isingly, mains eam pa ies use e y li le an i-eli ism,
gi en ha hey a e conside ed pa o ‘ he es ablishmen ’ hemsel es. Fo people-cen-
ism (panel B), he di e ences sh ink. While he a -le and he a - igh use appeals o
he people nea ly equally o en, mains eam pa ies also do employ he people-cen is
he o ic, howe e o a smalle ex en . This pa ly con i ms Ma ch’s (2017, 290) indings
o ‘demo icism (i.e. closeness o ‘o dina y’ people wi hou his an agonis ic iden i y)’ as
Figu e 1. Sca e plo o pe cen age o people-cen ism and an i-eli ism in MEPs’ speeches.
Figu e 2. Box plo s o pa ies by le - igh dimension.
POLITICAL RESEARCH EXCHANGE 17
a common ai o mains eam pa ies’ he o ic. My combined measu es (panel C) s ill
show ha adical le and adical igh pa ies use mo e populism han cen is pa ies.
Tu ning o he ou comes o he o he wo dic iona ies, we also see he in e se
u-shaped cu e in Bonikowski and Gid on’s (2018) (panel D) measu e. I is wo h poin ing
ou ha he e adical le pa ies u n ou o be mo e populis han adical igh pa ies. Fo
he measu e by Rooduijn and Pauwels (2011) (panel E) he di e ences be ween pa ies o
di e en ideological o ien a ions nea ly anish.
Conclusion
In his a icle, I ha e a gued ha he mo al dis inc ion be ween he people and he eli es is a
cen al componen o populis discou se. While his is heo e ically well-es ablished, quan-
i a i e app oaches o measu ing populism ha e o en neglec ed his mo al di ide. Build-
ing on his cen ali y o he wo an agonis ic g oups o populism discou se, I de eloped a
no el wo-s ep dic iona y app oach o measu e populis discou se. My app oach comes
wi h se e al ad an ages. Fi s , as my dic iona ies a e well-g ounded in he li e a u e,
hey a e less a isk o being con ex -dependen and hence less ad hoc ashioned.
Second, as I designed he app oach based on heo e ical assump ions and no ac o -
based, i is less p one o con la e populism wi h empi ically co-occu ing concep s, such
as na ionalism and adical igh ideology (Hunge and Pax on 2022). Addi ionally, combin-
ing dic iona ies o e e ences o g oups and hen iden i ying hei aming wi h a second
se o dic iona ies is a esea ch design which can be ans e ed o o he ields o s udy.
I alida ed my measu es ca e ully, agains a hand-coded gold s anda d and compa ed
hei pe o mance o o he populism dic iona ies. While he accu acy is no sky ocke ing,
my dic iona ies ou pe o m p e ious app oaches and a e hence a aluable ool o iden-
i ying populis discou se. Bonikowski and Gid on (2018) ob ain e y simila esul s in
e ms o he sha e o speeches con aining populis discou se, howe e , hei accu acy
measu es di e signi ican ly om my app oach. This migh be due o he ac ha hei
dic iona y mos ly includes wo ds ela ed o an i-eli ism a he han people-cen ism.
Hence, i is e y likely ha hei app oach is no sui able o measu ing people-cen ism.
A alid au oma ed app oach o measu e populis discou se is pa icula ly aluable o
la ge se s o ex s, as manual app oaches o en equi e an unmanageable e o , Zulia-
nello, Albe ini, and Ceccobelli (2018) o ins ance hand-coded 24,240 Facebook pos s. I
use my measu es o analyse he occu ence o populis discou se in he Eu opean Pa lia-
men om 1999 o 2014 d awing on a co pus o 178,905 speeches held by 1704 indi idual
MEPs. A compa ison o a p ede ined classi ica ion o populis pa ies shows ha indeed
‘Populis is as populis does’, meaning ha populis pa ies use mo e populis speech
han mains eam pa ies, howe e only o a small ma gin. The wo co e ea u es o popu-
lism in e ac wi h each o he on MEP le el, howe e , hei co-occu ence in one single
speech is empi ically sca ce. Ne e heless, ac o s adi ionally de ined as populis a e
mo e p one o use bo h ea u es ac oss hei speeches han mains eam pa y pa liamen-
a ians. Las ly, I show ha adical le and adical igh ac o s use populis discou se o a
g ea e ex en han mains eam pa ies. This di e ence is, howe e , mo e p onounced o
an i-eli ism han o people-cen ism.
Apa om se ing as a ool o iden i y populis discou se, showing how mo alized
aming is cen al o populism p o ides u he a enues o u u e esea ch. The i s
18 S. HUNGER
possible expansion is o s udy he e ec o mo al aming on indi iduals. This migh well
speak o schola ship on indi iduals’ mo i a ion o o e o populis pa ies (A zheime
2009; Bakke , Rooduijn, and Schumache 2016; Dunn 2015). We migh a gue ha aside
om classical o ing heo ies, populis s manage o ac i a e people’s mo al con ic ions
(see e.g. Ryan (2014); Ryan (2017)) by using hei mo alized aming and hence gain
hei o es (see e.g. Hamelee s, Bos, and de V eese 2017). Second – and connec ed o
schola ship on indi idual-le el populis a i udes (see e.g. Akke man, Mudde, and
Zaslo e 2014; Elcha dus and Sp uy 2016; Schulz e al. 2017) we could ask which ole
he mo alized aming and mo al con ic ions play in hese populis a i udes.
No es
1. P e ious hand-coded analyses (see e.g. E ns e al. (2017, 1359) show ha ‘The dimension o
es o ing so e eign y is almos absen ’.
2. In o de o exclude in e posed ques ions which a e o en echnical o p ocedu al, I in oduce
a minimum leng h o 26 wo ds. The maximum speech leng h is 2528 wo ds, wi h a mean o
210 wo ds. Addi ionally, I emo e speeches on o eign policy, as hey o en con ain e e ences
o people and eli es o non-Eu opean coun ies deba ed (Wa anabe 2018), which amoun s o
43,309 speeches. The emo al o speeches on o eign policy was done using he newsmap
dic iona y c ea ed by Wa anabe (2018). Speci ically, I emo ed all speeches wi h e e ences
o non-Eu opean coun ies wi h excep ion o he US.
3. The hand-coded andom sample con ains 6 pe cen o populis speeches. O he app oaches
ind simila numbe s o di e en ex co po a: Be nha d and K iesi (2019) in 5 pe cen o
pa ies’ p ess eleases in 11 coun ies, E ns e al. (2017) in abou 10 pe cen o hei analysed
wi e and acebook pos s, Zulianello, Albe ini, and Ceccobelli (2018) in abou 1.9 pe cen o
he Facebook pos s in hei sample.
4. Fo he sake o eadabili y, I emo ed ou lie s om he g aph, hey we e howe e included in
he es ima ion o he i ed lines.
Acknowledgemen s
Fo help ul commen s on p e ious e sions o his a icle, he au ho wishes o hank Pa is Aslanidis,
End e Bo ba h, Bjö n B eme , The esa Gessle , Edga G ande, Swen Hu e , Ellen Imme gu , Heiko
Klü e , Hanspe e K iesi, Sa ah de Lange, Daniel Saldi ia Gonza i, Ma iken an de Velden, Te esa
Völke , Ma a Webe , Lisa Zehn e , he edi o s and h ee anonymous e iewe s o PRX, as well as
he pa icipan s o he ollowing wo kshops and con e ences, whe e he a icle was p esen ed and
ecei ed help ul eedback: EPSA Annual Gene al Con e ence 2017, LSE G adua e Con e ence “Popu-
lism: Le Righ Cen e” 2017, and he Colloquium o he Zen um ü Zi ilgesellscha s o schung.
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 ha suppo he indings o his s udy a e openly a ailable in [Ha a d Da a e se] a
h ps://doi.o g/10.7910/DVN/9UGJVM, e e ence numbe [9UGJVM].
<will be c ea ed (including eplica ion ma e ial) upon publica ion>
The aw da a om he EU Pa liamen a y Speeches can be und ia h ps://gi hub.com/Con-
nec ed-Poli ics-Lab/EUDD/ ee/main/Da ase s/EP_Co pus
POLITICAL RESEARCH EXCHANGE 19
Re e ences
Adcock, R., and D. Collie . 2001. “Measu emen Validi y: A Sha ed S anda d o Quali a i e and
Quan i a i e Resea ch.” The Ame ican Poli ical Science Re iew 95 (3): 529–546. doi:10.1017/
S0003055401003100
Akke man, A., C. Mudde, and A. Zaslo e. 2014. “How Populis a e he People? Measu ing Populis
A i udes in Vo e s.” Compa a i e Poli ical S udies 47 (9): 1324–1353. doi:10.1177/
0010414013512600
A mony, A. C., and V. A mony. 2005. “Indic men s, My hs, and Ci izen Mobiliza ion in A gen ina: A
Discou se Analysis.” La in Ame ican Poli ics & Socie y 47 (4): 27–54. doi:10.1111/j.1548-2456.2005.
b00327.x
A zheime , K. 2009. “Con ex ual Fac o s and he Ex eme Righ Vo e in Wes e n Eu ope, 1980-2002.”
Ame ican Jou nal o Poli ical Science 53 (2): 259–275. doi:10.1111/j.1540-5907.2009.00369.x
Aslanidis, P. 2016. “Is Populism an Ideology? A Re u a ion and a new Pe spec i e.” Poli ical S udies 64
(1_suppl): 88–104. doi:10.1111/1467-9248.12224
Aslanidis, P. 2017. “Measu ing Populis Discou se wi h Seman ic Tex Analysis: An Applica ion on
G ass oo s Populis Mobiliza ion.” Quali y & Quan i y 52: 1241–1263.
Bakke , B. N., M. Rooduijn, and G. Schumache . 2016. “The Psychological Roo s o Populis Vo ing:
E idence om he Uni ed S a es, he Ne he lands and Ge many.” Eu opean Jou nal o Poli ical
Resea ch 55 (2): 302–320. doi:10.1111/1475-6765.12121
Benoi , K., K. Wa anabe, H. Wang, P. Nul y, A. Obeng, S. Mülle , A. Ma suo, e al. 2018. “Quan eda: An
Package o he Quan i a i e Analysis o Tex ual Da a.” Jou nal o Open Sou ce So wa e 3 (30):
774. doi:10.21105/joss.00774
Be nha d, L., and H. K iesi. 2019. “Populism in Elec ion Times: A Compa a i e Analysis o 11 Coun ies
in Wes e n Eu ope.” Wes Eu opean Poli ics 42 (6): 1188–1208. doi:10.1080/01402382.2019.
1596694
Be nha d, L., H. K iesi, and E. Webe . 2015. “The Populis Discou se o he Swiss People’s Pa y.” In
Eu opean Populism in he Shadow o he G ea Recession, edi ed by H. K iesi, and T. S. Pappas,
125–139. Colches e : ECPR P ess.
Bonikowski, B., and N. Gid on. 2016. “The Populis S yle in Ame ican Poli ics: P esiden ial Campaign
Rhe o ic, 1952-1996.” Social Fo ces 94 (4): 1593–1621. doi:10.1093/s /so 120
Bonikowski, B., and N. Gid on. 2018. “Populism in Legisla i e Discou se: E idence om he Eu opean
Pa liamen , 1999-2004.” Wo king Pape .
Bonikowski, B., D. Halikiopoulou, E. Kau mann, and M. Rooduijn. 2019. “Populism and Na ionalism in
a Compa a i e Pe spec i e: A Schola ly Exchange.” Na ions and Na ionalism 25 (1): 58–81. doi:10.
1111/nana.12480
Bonikowski, B., Y. Luo, and O. S uhle . 2022. “Poli ics as Usual? Measu ing Populism, Na ionalism, and
Au ho i a ianism in U.S. P esiden ial Campaigns (1952–2020) wi h Neu al Language Models.”
Sociological Me hods & Resea ch 51 (4): 1721–1787. doi:10.1177/00491241221122317
B ack, N. 2015. “The Roles o Eu oscep ic Membe s o he Eu opean Pa liamen and Thei
Implica ions o he EU.” In e na ional Poli ical Science Re iew 36 (3): 337–350. doi:10.1177/
0192512115571590
B eye , M. 2022. “Populis Posi ions in Pa y Compe i ion: Do Pa ies S a egically Va y Thei Deg ee
o Populism in Reac ion o Vo e and O ice Loss?” Pa y Poli ics 29 (4): 672–684.
B ubake , R. 2017a. “Why Populism?” Theo y and Socie y 46 (5): 357–385. doi:10.1007/s11186-017-
9301-7
B ubake , R. 2017b. “Be ween Na ionalism and Ci iliza ionism: The Eu opean Populis Momen in
Compa a i e Pe spec i e.” E hnic and Racial S udies 40 (8): 1191–1226. doi:10.1080/01419870.
2017.1294700
Caiani, M., and D. della Po a. 2011. “The Eli is Populism o he Ex eme Righ : A F ame Analysis o
Ex eme Righ -Wing Discou ses in I aly and Ge many.” Ac a Poli ica 46 (2): 180–202. doi:10.1057/
ap.2010.28
Cammack, P. 2000. “The Resu gence o Populism in La in Ame ica.” Bulle in o La in Ame ican
Resea ch 19 (2): 159–161. doi:10.1111/j.1470-9856.2000. b00096.x
20 S. HUNGER
Celico, A., M. Rode, and I. Rod iguez-Ca eño. 2024. “Will he Real Populis s Please S and up? A
Machine Lea ning Index o Pa y Populism.” Eu opean Jou nal o Poli ical Economy 82: 102529.
doi:10.1016/j.ejpoleco.2024.102529
C anme , M. 2011. “Populis Communica ion and Publici y: An Empi ical S udy o Con ex ual
Di e ences in Swi ze land.” Swiss Poli ical Science Re iew 17 (3): 286–307. doi:10.1111/j.1662-
6370.2011.02019.x
C oss, J., and D. G eene. 2016. “Replica ion Da a o : Explo ing he Poli ical Agenda o he Eu opean
Pa liamen Using a Dynamic Topic Modeling App oach.” Ha a d Da a e se 25 (1): 77–94. doi:10.
7910/DVN/LVHLZK
Deegan-K ause, K., and T. Haugh on. 2009. “Towa d a Mo e Use ul Concep ualiza ion o Populism:
Types and Deg ees o Populis Appeals in he Case o Slo akia.” Poli ics & Policy 37 (4): 821–841.
doi:10.1111/j.1747-1346.2009.00200.x
De la To e, C. 2000. Populis Seduc ion in La in Ame ica: The Ecuado ian Expe ience. A hens: Ohio
Uni e si y P ess.
Di Cocco, J., and B. Monechi. 2022. “How Populis a e Pa ies? Measu ing Deg ees o Populism in
Pa y Mani es os Using Supe ised Machine Lea ning.” Poli ical Analysis 30 (3): 311–327. doi:10.
1017/pan.2021.29
Dö ing, H., and P. Manow. 2016. Pa liamen s and Go e nmen s Da abase (Pa lGo ): In o ma ion on
Pa ies, Elec ions and Cabine s in Mode n Democ acies. De elopmen Ve sion.
Dunn, K. 2015. “P e e ence o Radical Righ -Wing Populis Pa ies among Exclusi e-Na ionalis s and
Au ho i a ians.” Pa y Poli ics 21 (3): 367–380. doi:10.1177/1354068812472587
Elcha dus, M., and B. Sp uy . 2016. “Populism, Pe sis en Republicanism and Declinism: An Empi ical
Analysis o Populism as a Thin Ideology.” Go e nmen and Opposi ion 51 (1): 111–133. doi:10.
1017/go .2014.27
Engesse , S., N. E ns , F. Esse , and F. Büchel. 2017. “Populism and Social Media: How Poli icians
Sp ead a F agmen ed Ideology.” In o ma ion, Communica ion & Socie y 20 (8): 1109–1126.
doi:10.1080/1369118X.2016.1207697
E ha d, L., S. Hanke, U. Reme , A. Falenska, and R. Heibe ge . 2023. PopBERT. De ec ing populism and
i s hos ideologies in he Ge man Bundes ag. a Xi p ep in a Xi :2309.14355.
E ns , N., S. Blassnig, S. Engesse , F. Büchel, and F. Esse . 2019. “Populis s P e e Social Media O e
Talk Shows: An Analysis o Populis Messages and S ylis ic Elemen s Ac oss six Coun ies.”
Social Media + Socie y 5 (1): 1–14. doi:10.1177/2056305118823358
E ns , N., S. Engesse , F. Büchel, S. Blassnig, and F. Esse . 2017. “Ex eme Pa ies and Populism: An
Analysis o Facebook and Twi e Ac oss six Coun ies.” In o ma ion, Communica ion & Socie y
20 (9): 1347–1364. doi:10.1080/1369118X.2017.1329333
Eu opean Pa liamen . 2012. Amendmen o Rule 181 on Ve ba im Repo s and Rule 182 on
Audio isual Reco d o P oceedings.
G abbe, H., and N. G oo . 2014. “Populism in he Eu opean Pa liamen : Wha Implica ions o he
Open Socie y?” The In e na ional Spec a o 49 (4): 33–46. doi:10.1080/03932729.2014.961768
G aham, J., J. Haid , and B. A. Nosek. 2009. “Libe als and Conse a i es Rely on Di e en Se s o Mo al
Founda ions.” Jou nal o Pe sonali y and Social Psychology 96 (5): 1029–1046. doi:10.1037/
a0015141
G eene, D., and J. C oss. 2017. “Explo ing he Poli ical Agenda o he Eu opean Pa liamen Using a
Dynamic Topic Modeling App oach.” Poli ical Analysis 25(1): 77–94.
G imme , J., and B. M. S ewa . 2013. “Tex as Da a: The P omise and Pi alls o Au oma ic
Con en Analysis Me hods o Poli ical Tex s.” Poli ical Analysis 21 (3): 267–297. doi:10.1093/
pan/mps028
G ündl, J. 2022. “Populis Ideas on Social Media: A Dic iona y-Based Measu emen o
Populis Communica ion.” New Media & Socie y 24 (6): 1481–1499. doi:10.1177/
1461444820976970
Haid , J., and J. G aham. 2007. “When Mo ali y Opposes Jus ice: Conse a i es Ha e Mo al
In ui ions Tha Libe als may no Recognize.” Social Jus ice Resea ch 20 (1): 98–116. doi:10.
1007/s11211-007-0034-z
POLITICAL RESEARCH EXCHANGE 21
Haid , J., and C. Joseph. 2005. “The Mo al Mind: How Fi e Se s o Inna e In ui ions Guide he
De elopmen o Many Cul u e-Speci ic Vi ues, and Pe haps E en Modules.” In The Inna e
Mind, edi ed by P. Ca u he s, S. Lau ence, and S. P. S ich, 367–391. New Yo k: Ox o d
Uni e si y P ess.
Hamelee s, M., L. Bos, and C. H. de V eese. 2017. “They did i ”: The E ec s o Emo ionalized Blame
A ibu ion in Populis Communica ion.” Communica ion Resea ch 44 (6): 870–900. doi:10.1177/
0093650216644026
Hamelee s, M., and R. Vliegen ha . 2019. “The Rise o a Populis Zei geis ? A Con en Analysis o
Populis Media Co e age in Newspape s Published Be ween 1990 and 2017.” Jou nalism
S udies 1–18.
Hawkins, K. A. 2009. “Is Chá ez Populis ? Measu ing Populis Discou se in Compa a i e Pe spec i e.”
Compa a i e Poli ical S udies 42 (8): 1040–1067.
Hawkins, K. A. 2010. Venezuela’s Cha ismo and Populism in Compa a i e Pe spec i e. Camb idge
Uni e si y P ess.
Hawkins, K. A., and C. Ro i a Kal wasse . 2017. “Wha he (Idea ional) S udy o Populism Can Teach
us, and Wha i Can’ .” Swiss Poli ical Science Re iew 23 (4): 526–542. doi:10.1111/sps .12281
Hunge , S., and F. Pax on. 2022. “Wha ’s in a Buzzwo d? A Sys ema ic Re iew o he S a e o
Populism Resea ch in Poli ical Science.” Poli ical Science Resea ch and Me hods 10 (3): 617–633.
doi:10.1017/ps m.2021.44
Jage s, J., and S. Walg a e. 2007. “Populism as Poli ical Communica ion S yle: An Empi ical S udy o
Poli ical Pa ies’ Discou se in Belgium.” Eu opean Jou nal o Poli ical Resea ch 46 (3): 319–345.
doi:10.1111/j.1475-6765.2006.00690.x
Jankowski, M., and R. Hube . 2023. “When Co ela ion is no Enough: Valida ing Populism Sco es
om Supe ised Machine-Lea ning Models.” Poli ical Analysis 31 (4): 591–605. doi:10.1017/pan.
2022.32
Jaspe , J. M. 1992. “The Poli ics o Abs ac ions: Ins umen al and Mo alis Rhe o ics in Public
Deba e.” Social Resea ch 59 (2): 315–344.
Jung, J.-H. 2019. “The Mobilizing E ec o Pa ies’ Mo al Rhe o ic.” Ame ican Jou nal o Poli ical
Science 64 (2): 341–355.
Kan ola, J., and C. Mille . 2021. “Pa y Poli ics and Radical Righ Populism in he Eu opean Pa liamen :
Analysing Poli ical G oups as Democ a ic Ac o s.” JCMS: Jou nal o Common Ma ke S udies 59 (4):
782–801. doi:10.1111/jcms.13181
K iesi, H. 2014. “The Populis Challenge.” Wes Eu opean Poli ics 37 (2): 361–378. doi:10.1080/
01402382.2014.887879
La e , M., and J. Ga y. 2000. “Es ima ing Policy Posi ions om Poli ical Tex s.” Ame ican Jou nal o
Poli ical Science 44 (3): 619–634. doi:10.2307/2669268
Lehmann, P., S. F anzmann, T. Bu s , S. Regel, F. Rie hmülle , A. Volkens, B. Be nha d, and L. Zehn e .
2023. The Mani es o Da a Collec ion. Mani es o P ojec (MRG/CMP/MARPOR) . Ve sion 2023a. Be lin:
Wissenscha szen um Be lin ü Sozial o schung (WZB) / Gö ingen: Ins i u ü
Demok a ie o schung (I Dem).
Le ick, L., and M. Ola a ia-Gambi. 2020. “Hindsigh Bias in Expe Su eys: How Democ a ic
C ises In luence Re ospec i e E alua ions.” Poli ics 40 (4): 494–509. doi:10.1177/
0263395720914571
Manucci, L., and E. Webe . 2017. “Why he big Pic u e Ma e s: Poli ical and Media Populism in
Wes e n Eu ope Since he 1970s.” Swiss Poli ical Science Re iew 23 (4): 313–334. doi:10.1111/
sps .12267
Ma ch, L. 2017. “Le and Righ Populism Compa ed: The B i ish Case.” The B i ish Jou nal o Poli ics
and In e na ional Rela ions 19 (2): 282–303. doi:10.1177/1369148117701753
Meije s, M. J., and A. Zaslo e. 2021. “Measu ing Populism in Poli ical Pa ies: App aisal o a new
App oach.” Compa a i e Poli ical S udies 54 (2): 372–407. doi:10.1177/0010414020938081
Mény, Y., and Y. Su el. 2000. Pa le Peuple, Pou le Peuple: Le Populisme e les Démoc a ies. Pa is:
Faya d.
Meny, Y., and Y. Su el. 2002. “The Cons i u i e Ambigui y o Populism.” In Democ acies and he
Populis Challenge, edi ed by Y. Meny, and Y. Su el, 1–21. London: Palg a e.
22 S. HUNGER
Mou e, C. 2000. The Democ a ic Pa adox. London: Ve so.
Mudde, C. 2004. “The Populis Zei geis .” Go e nmen and Opposi ion 39 (4): 541–563. doi:10.1111/j.
1477-7053.2004.00135.x
Mudde, C. 2017. “Populism: An Idea ional App oach.” In The Ox o d Handbook o Populism, edi ed by
C. Ro i a Kal wasse , P. Tagga , P. Ochoa Espejo, and P. Os iguy, 28–47. New Yo k: Ox o d
Uni e si y P ess.
Mudde, C., and C. Ro i a Kal wasse . 2018. “S udying Populism in Compa a i e Pe spec i e:
Re lec ions on he Con empo a y and Fu u e Resea ch Agenda.” Compa a i e Poli ical S udies
51 (13): 1667–1693. doi:10.1177/0010414018789490
Mülle , J.-W. 2017. Wha is Populism?. London: Penguin UK.
Na o zky, S. 2016. “Be ween Inequali y and Injus ice: Digni y as a Mo i e o Mobiliza ion Du ing he
C isis.” His o y and An h opology 27 (1): 74–92. doi:10.1080/02757206.2015.1111209
Oli e , J. E., and W. M. Rahn. 2016. “Rise o he T umpen olk: Populism in he 2016 Elec ion.” The
ANNALS o he Ame ican Academy o Poli ical and Social Science 667 (1): 189–206. doi:10.1177/
0002716216662639
Panizza, F. 2005. Populism and he Mi o o Democ acy. London: Ve so.
Pauwels, T. 2011. “Measu ing Populism: A Quan i a i e Tex Analysis o Pa y Li e a u e in Belgium.”
Jou nal o Elec ions, Public Opinion and Pa ies 21 (1): 97–119. doi:10.1080/17457289.2011.539483
Reungoa , E. 2010. “An i-EU Pa ies and he People: An Analysis o Populism in F ench
Eu omani es os.” Pe spec i es on Eu opean Poli ics and Socie y 11 (3): 292–312. doi:10.1080/
15705854.2010.503034
Rooduijn, M. 2015. “The Rise o he Populis Radical Righ in Wes e n Eu ope.” Eu opean View 14 (1):
3–11. doi:10.1007/s12290-015-0347-5
Rooduijn, M. 2019. “S a e o he Field: How o S udy Populism and Adjacen Topics? A Plea o Bo h
Mo e and Less Focus.” Eu opean Jou nal o Poli ical Resea ch 58 (1): 362–372. doi:10.1111/1475-
6765.12314
Rooduijn, M., and T. Akke man. 2017. “Flank A acks: Populism and Le -Righ Radicalism in Wes e n
Eu ope.” Pa y Poli ics 23 (3): 193–204. doi:10.1177/1354068815596514
Rooduijn, M., S. L. de Lange, and W. an de B ug. 2014. “A Populis Zei geis ? P og amma ic
Con agion by Populis Pa ies in Wes e n Eu ope.” Pa y Poli ics 20 (4): 563–575. doi:10.1177/
1354068811436065
Rooduijn, M., and T. Pauwels. 2011. “Measu ing Populism: Compa ing wo Me hods o Con en
Analysis.” Wes Eu opean Poli ics 34 (6): 1272–1283. doi:10.1080/01402382.2011.616665
Rooduijn, M., S. Van Kessel, C. F oio, A. Pi o, and S. L. de Lange. 2019. The PopuLis : An
O e iew o Populis , Fa Righ , Fa Le and Eu oscep ic Pa ies in Eu ope. h p://www.popu-
lis .o g.
Rudin, C. 2019. “S op Explaining Black box Machine Lea ning Models o High S akes Decisions and
use In e p e able Models Ins ead.” Na u e Machine In elligence 1 (5): 206–215. doi:10.1038/
s42256-019-0048-x
Ruedin, D., and L. Mo ales. 2017. “Es ima ing Pa y Posi ions on Immig a ion: Assessing he
Reliabili y and Validi y o Di e en Me hods.” Pa y Poli ics 25 (3): 1–12.
Ryan, T. J. 2014. “Reconside ing Mo al Issues in Poli ics.” The Jou nal o Poli ics 76 (2): 380–397.
doi:10.1017/S0022381613001357
Ryan, T. J. 2017. “No Comp omise: Poli ical Consequences o Mo alized A i udes.” Ame ican Jou nal
o Poli ical Science 61 (2): 409–423. doi:10.1111/ajps.12248
Schul e-Cloos, J. 2018. “Do Eu opean Pa liamen Elec ions Fos e Challenge Pa ies’ Success
on he Na ional Le el?” Eu opean Union Poli ics 19 (3): 408–426. doi:10.1177/
1465116518773486
Schulz, A., P. Mülle , C. Scheme , D. S. Wi z, M. We s ein, and W. Wi h. 2017. “Measu ing Populis
A i udes on Th ee Dimensions.” In e na ional Jou nal o Public Opinion Resea ch 30 (2): 316–326.
Slapin, J. B., and S.-O. P oksch. 2010. “Look Who’s Talking: Pa liamen a y Deba e in he Eu opean
Union.” Eu opean Union Poli ics 11 (3): 333–357. doi:10.1177/1465116510369266
S anye , J., S. Salgado, and J. S ömbäck. 2016. “Populis Ac o s as Communica o s o Poli ical Ac o s
as Populis Communica o s. C oss-Na ional Findings and Pe spec i es.” In Populis Poli ical
POLITICAL RESEARCH EXCHANGE 23
Communica ion in Eu ope, edi ed by T. Aalbe g, F. Esse , C. Reinemann, J. S ömbäck, and C. H. de
V eese, 353–364. New Yo k: Rou ledge.
S a akakis, Y., and A. Jäge . 2017. “Accomplishmen s and Limi a ions o he “new” Mains eam in
Con empo a y Populism S udies.” Eu opean Jou nal o Social Theo y 21 (4): 547–565.
V eese, C. H., F. de Esse , T. Aalbe g, C. Reinemann, and J. S anye . 2018. “Populism as an Exp ession
o Poli ical Communica ion Con en and S yle: A new Pe spec i e.” The In e na ional Jou nal o
P ess/Poli ics 23 (4): 423–438. doi:10.1177/1940161218790035
Wa anabe, K. 2018. “Newsmap - a Semi-Supe ised App oach o Geog aphical News Classi ica ion.”
Digi al Jou nalism 6 (3): 294–309. doi:10.1080/21670811.2017.1293487
Wiles, P. 1969. “A Synd ome, no a Doc ine.” In Populism: I s Meaning and Na ional Cha ac e is ics,
edi ed by G. Ionescu, and E. Gellne , 166–179. Macmillan.
Zehn e , L. 2023. “Iden i y Th ough Dis inc ion. Con ex ualizing in- and Ou g oups in
Populis Poli ical Communica ion.” Pape P esen ed a he ECPR Gene al Con e ence 2023.
Zulianello, M., A. Albe ini, and D. Ceccobelli. 2018. “A Populis Zei geis ? The Communica ion
S a egies o Wes e n and La in Ame ican Poli ical Leade s on Facebook.” The In e na ional
Jou nal o P ess/Poli ics 23 (4): 439–457. doi:10.1177/1940161218783836
Appendix
F ame dic iona ies
Table A1. Keywo ds o posi i e and nega i e aming o he people and he eli e.
nega i e
ame posi i e ame
abuse* dep a * hung y o
powe
poli ical class sins be ay* noble
aliena e desec a * igno e* powe
g abbing
spu n blameless o dina y man
annihila e dese ed immo al* powe
hung y
s ain* b a e o dina y men
apos asy dese e * impai powe
monge
s omp b a ely o dina y
pe son
apos a e dese ing impe ialis * powe -
g abbing
sycophany cleanliness o dina y
appa a chik* des oy impie y powe -
hung y
ain * decency p aisewo hy
a ogan* de imen * impious p i ilege a nish* decen p incipled
bad disc imna * impos e p i ileges ai o * de en* easonable
bad apples disgus * indecen* p o an* ansg ess* digni* e ined
be ay* dishones inequi able p o liga e ashy e hic* he good
bias* disloyal* inhuman* a age eache * expense o up igh
bigo * disobe* ins able e use eason* ai ups anding
blemish disp opo ion insubo dina * e use* unaccep able* o ge alo
B ussels
ma ia
dis espec * insu gen enegade unca ing o ge ing alou *
con agio* dissocia e in empe a e epuls* unchas e ha d ea ned i uous
co up * dys unc ional* jil * uin* unclean* ha d wo king
cowa d* endange * lawless* uling ci cles undemoc a ic* ha d-ea ned
c onies enem* lobby* uling ci cles unelec ed ha d-wo king
c ooked* eu oc a * misc ean sedi i* unequal* ha dwo king
cynical* eu ophil* mu inous sel -
in e es ed
un ai h ul hones *
damag* e il noncon o mis sel -sa is ied unheal hy hono *
debase* a ou i ism no in ouch sel -se ing un esponsi e honou *
debauche* il h* obscen* seques e unsc upulos igno *
decei * aud obs uc sick useless in e es
de ec o g eed o end* sin enal loyal
de ian* g oss old eli es sin ul* wan ona loyal *
de ile* ha m ul* oppose sinned weal hy
(Con inued)
24 S. HUNGER
