Mo isi, Da ide; Clolé y, Héloïse; Kon Kam King, Guillaume; Schaub, Max
A icle — Published Ve sion
How COVID-19 a ec s o ing o incumben s: E idence
om local elec ions in F ance
PLOS ONE
P o ided in Coope a ion wi h:
WZB Be lin Social Science Cen e
Sugges ed Ci a ion: Mo isi, Da ide; Clolé y, Héloïse; Kon Kam King, Guillaume; Schaub, Max (2024) :
How COVID-19 a ec s o ing o incumben s: E idence om local elec ions in F ance, PLOS ONE,
ISSN 1932-6203, Public Lib a y o Science (PLoS), San F ancisco, CA, Vol. 19, Iss. 3, pp. 1-19,
h ps://doi.o g/10.1371/jou nal.pone.0297432
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/312942
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
How COVID-19 a ec s o ing o incumben s:
E idence om local elec ions in F ance
Da ide Mo isiID
1,2
, He
´loïse Clole
´ yID
3
, Guillaume Kon Kam King
3,4
, Max SchaubID
5,6
*
1Depa men o Poli ical Science and Public Managemen , Uni e si y o Sou he n Denma k, Odense,
Denma k, 2Collegio Ca lo Albe o, Tu in, I aly, 3CREST, E
´cole Poly echnique, IP Pa is, Palaiseau, F ance,
4INRAE, Uni e si e
´Pa is-Saclay, Jouy-en-Josas, F ance, 5Depa men o Poli ical Science, Uni e si y o
Hambu g, Hambu g, Ge many, 6WZB Be lin Social Science Cen e , Be lin, Ge many
*max.schaub@uni-hambu g.de
Abs ac
How do o e s eac o an ongoing na u al h ea ? Do o e s sanc ion o ewa d incumben s
e en when incumben s canno be held accoun able because an un o eseeable na u al
disas e is un olding? We add ess his ques ion by in es iga ing o e s’ eac ions o he
ea ly sp ead o COVID-19 in he 2020 F ench municipal elec ions. Using a no el, ine-
g ained measu e o he ci cula ion o he i us based on excess-mo ali y da a, we ind ha
suppo o incumben s inc eased in a eas ha we e pa icula ly ha d hi by he i us. Incum-
ben s om bo h le and igh gained o es in a eas mo e s ongly a ec ed by COVID-19.
We p o ide sugges i e e idence o wo mechanisms ha can explain ou indings: an emo-
ional channel ela ed o eelings o ea and anxie y, and a p ospec i e- o ing channel,
ela ed o he abili y o incumben s o ac mo e swi ly agains he di usion o he i us han
challenge s.
1. In oduc ion
Na u al disas e s in luence o ing. A e ca as ophic e en s—such as ea hquakes, loods, and
hu icanes— o e s o en ewa d poli ical candida es and incumben s [1–5] o punish hem
elec o ally [6–11], depending on how hey esponded o such e en s. Vi ually all a ailable e i-
dence ocuses on how o e s e ospec i ely e alua e poli icians’ pe o mance a e na u al
disas e s ha e occu ed [12], and whe he hese e alua ions in luence o ing. Howe e , we do
no know how o e s espond o a h ea ening e en while i is un olding. How do o e s eac
o an ongoing na u al h ea ? Do o e s sanc ion o ewa d incumben s e en when incum-
ben s canno be held accoun able because an un o eseeable na u al disas e is s ill un olding?
We add ess his ques ion by in es iga ing o e s’ eac ions o he ea ly sp ead o COVID-19
in F ance. COVID-19 di e s om o he na u al disas e s since i is a p olonged h ea ha
does no esul in physical des uc ion bu sp eads in isibly, po en ially a ec ing all human
beings. Like o he in ec ious diseases [13], COVID-19 gene a es ea , which, we a gue,
inc eases o e s’ endency o suppo incumben candida es as a means o ob ain sa e y and
secu i y. This dynamic applied especially a he onse o he pandemic when he e was unce -
ain y ega ding he bes s a egy o con aining he i us and gua an eeing ci izens’ sa e y.
PLOS ONE
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 1 / 19
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Ci a ion: Mo isi D, Clole
´ y H, Kon Kam King G,
Schaub M (2024) How COVID-19 a ec s o ing o
incumben s: E idence om local elec ions in
F ance. PLoS ONE 19(3): e0297432. h ps://doi.
o g/10.1371/jou nal.pone.0297432
Edi o : Eugenio P o o, Uni e si y o Glasgow,
UNITED KINGDOM
Recei ed: June 17, 2023
Accep ed: Janua y 4, 2024
Published: Ma ch 19, 2024
Copy igh : ©2024 Mo isi e al. 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, 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
au ho and sou ce a e c edi ed.
Da a A ailabili y S a emen : All eplica ion
ma e ial and code ha e been published in a public
eposi o y, he e: h ps://doi.o g/10.7910/DVN/
EFDGLY.
Funding: DM ecei ed a special COVID-19 g an by
Collegio Ca lo Albe o (Tu in, I aly) o conduc ing
he esea ch included in his manusc ip . The
sponso did no play any ole in he s udy design,
da a collec ion and analysis, decision o publish, o
p epa a ion o he manusc ip . The publica ion o
his a icle was suppo ed by he Open Access
Publica ion Fund o he Uni e si y o Hambu g.
We ocus on he municipal elec ions ha ook place in F ance on 15 Ma ch 2020, when
SARS- Co -2, he i us causing COVID-19, was al eady ci cula ing in he coun y. To es ima e
he “ h ea ” o COVID-19, we de eloped a no el, ine-g ained measu e o he sp ead o he
i us ( he “p e alence” o COVID-19) a he municipali y le el, using excess-mo ali y da a
weighed by he age and sex s uc u e o he popula ion. This measu e accoun s o he In ec-
ion Fa ali y Ra io [14] and allows us o add ess he p oblem o unde epo ing wi h ega d o
he ci cula ion o he i us a he beginning o he pandemic [15]. E en i he o al numbe o
con i med cases o COVID-19 in F ance on 15 Ma ch 2020 was ela i ely low—a o al o 4532
[16]—ou es ima es indica e ha a subs an ial sha e o F ench o e s al eady knew o o he
people who su e ed om COVID-19 in hei local a ea. Speci ically, we es ima e ha he e
was a small bu sizeable g oup o municipali ies (7 pe cen o he o al), in which a o e had a
leas a 30 pe cen p obabili y o knowing a leas ano he pe son wi h COVID-19 symp oms in
hei inne ci cle o iends and acquain ances. E en i he p opo ion o people wi h COVID-
19 symp oms in a gi en municipali y is low, he p obabili y ha a andom o e knows a leas
one pe son wi h COVID-19 symp oms can be la ge (as u he explained below).
We exploi he une en di usion o he i us in he coun y o es whe he suppo o local
incumben s inc eased in municipali ies ha we e pa icula ly ha d hi by COVID-19. In pa -
icula , we es ima e he e ec o being se e ely a ec ed using a) eg ession models ha con ol
o he lagged dependen a iable and condi ion on baseline mo ali y a es, b) a di e ence-in-
di e ences app oach, and c) p opensi y sco e ma ching. In line wi h e idence o a “ ally-
a ound- he- lag” e ec a he beginning o he pandemic [17–22], ou indings show ha o e
sha es o local incumben s inc eased as he h ea o COVID-19 inc eased. In municipali ies
wi h a high ci cula ion o he i us ( en h decile), he a e age o e sha e o incumben s was
2.5 pe cen age poin s highe han in places wi h a low ci cula ion ( i s decile). The e ec s a e
insensi i e o he poli ical a ilia ion o he incumben s, meaning ha suppo o incumben s
inc eases ega dless o whe he hey a e le -wing, igh -wing, o om o he pa ies.
We also show ha he e ec o COVID-19 on o ing is mo e p onounced in a eas a) whe e
people eel pa icula ly anxious abou he pandemic, and b) whe e economic wel a e imp o ed
du ing he ou going adminis a ion. This e idence suppo s ou a gumen ha , when aced
wi h he ci cula ion o he i us, o e s, on he one hand, u ned o incumben s as a means o
educing ea and anxie y–an “emo ional channel” in line wi h e idence om he Uni ed S a es
[23,24]. In addi ion, ollowing he logic o p ospec i e o ing, we a gue ha o e s suppo
incumben s o main ain con inui y wi h he cu en adminis a ion as he mos e icien
means o ackling he i us, especially in a eas whe e he incumben mayo s had pe o med
well.
In he nex sec ion, we e iew exis ing esea ch on he e ec s o na u al disas e s on o ing,
and ad ance ou a gumen s o why he ea ly sp ead o COVID-19 should inc ease suppo
o local incumben s. We hen p o ide some b ie backg ound in o ma ion on he 2020
municipal elec ions in F ance, desc ibe ou da a and me hods, and p esen he esul s. In he
conclusion, we summa ize ou indings and discuss he implica ions and he limi a ions o ou
s udy.
2 Na u al h ea s and suppo o incumben s
Resea ch shows ha o e s punish o c edi poli ical candida es depending on how hese can-
dida es espond o se e e wea he phenomena. Fo example, s udies indica e ha incumben
poli icians lose suppo i hei esponse o na u al disas e s is pe cei ed as inadequa e [6,8,
11,25]. Vo e s also punish incumben s in he a e ma h o se e e wea he e en s ha cause
economic damage, such as o nadoes, loods and ex eme ain all [4,7,9,10], and e en in he
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 2 / 19
Compe ing in e es s: NO au ho s ha e compe ing
in e es s.
case o e en s ha a e clea ly ou o poli icians’ con ol, such as sha k a acks [26]. On he
o he hand, he e is e idence ha o e s ewa d go e nmen s and incumben poli icians who
a e seen as ha ing esponded well o na u al disas e s [1–3,5,12].
The p ima y mechanism desc ibing o e s’ eac ion a e a disas e has occu ed is e o-
spec i e o ing [27]: on elec ion day, o e s sanc ion o ewa d poli icians depending on how
well hey esponded o he damage caused by he na u al ca as ophe. A he beginning o he
COVID-19 pandemic, howe e , i was unclea how o bes espond o he ci cula ion o he
i us. As an in isible h ea ha does no lead o physical des uc ion and pe sis s o an inde i-
ni e pe iod, COVID-19 canno be ackled wi h he same means used a e a na u al ca as o-
phe. Relie -spending and he ypes o econs uc ion unde aken a e na u al disas e s we e
no iable op ions o incumben poli icians a he beginning o he pandemic. This is espe-
cially ue o local mayo s who lacked he legal au ho i y o in oduce s ong con ainmen
measu es, such as lockdowns. Thus, he ini ial sp ead o COVID-19 should no ha e a ec ed
o ing o local incumben s h ough e ospec i e e alua ions since i was unclea how poli i-
cians ough o ha e esponded o he pandemic.
In he absence o e ospec i e e alua ions, h ea ening e en s can s ill induce people o
exp ess g ea e suppo o incumben poli ical leade s ia emo ional eac ions, he so-called
‘ ally-a ound- he- lag’ e ec [28–30]. This e ec has been iden i ied in ela ion o e o is
a acks (see o example [31–35]) and mili a y c ises [28], bo h o which ha e been ound o
lead o a boos in p esiden ial popula i y and go e nmen suppo . Recen s udies ha e a gued
ha he h ea o COVID-19 should ha e simila ally e ec s [21], a leas a he beginning o
he pandemic. Indeed, e idence indica es ha suppo o wo ld leade s [17–19,22] and us
in go e nmen s [20,36–38] inc eased a he beginning o he COVID-19 pandemic. We do
no know, howe e , whe he app o al a ings ac ually ansla ed in o o ing beha iou (bu
see [39]), and i is unclea wha mechanisms could connec he sp ead o he i us o o ing
beha iou .
We a gue ha he ea ly sp ead o COVID-19 should lead o e s o inc ease suppo o
local incumben s mainly h ough wo channels. Simila o ally e ec s, he i s channel is
emo ional, al hough in ol ing a di e en ype o emo ion: anxie y (ins ead o ange o pa i-
o ic a ousal). I has been a gued ha ally e ec s a e ac i a ed mos ly by a sen imen o pa io -
ism and by ange owa ds a clea ly iden i iable ex e nal agg esso , such as a e o is g oup
[29,40]. The h ea o COVID-19, howe e , should no ac i a e he same eelings o ange and
pa io ism, since he e is no delibe a e a ack behind he ci cula ion o he i us (i.e., he i us
did no “in end” o a ack a popula ion, as e o is s do), and he e is no clea ly iden i iable
human enemy agains whom he popula ion should mobilize.
Ins ead, we a gue ha COVID-19 igge s ea and anxie y, as indica ed by psychological
s udies [41,42] and by ecen e idence in he U.S. [23] (see also [21]). As wi h he Ebola i us
when i eached he US in 2014 [13], a he beginning o he pandemic, COVID-19 ep esen ed
an in isible h ea agains which he e was no clea emedy, hus po en ially inducing ea in
he popula ion. Di e en s ands o heo ies indica e ha sen imen s o ea and anxie y lead
o isk-a e sion ( o a e iew, see [43]), which, in u n, should ansla e in o inc eased suppo
o incumben candida es, as hey ep esen a sa e , s a us quo op ion compa ed o he isk o
elec ing an as-ye -un es ed challenge .
Fu he mo e, acco ding o e o -managemen [44] and mo i a ed-social-cogni ion heo y
[45,46], people wan o see he wo ld as a secu e place. The e o e, in he p esence o a ea -
inducing h ea such as COVID-19, o e s should u n hei suppo owa ds incumben poli-
icians who o e an ac ual o symbolic sense o sa e y and secu i y because hey ep esen he
s a us quo. Indeed, e idence om he Uni ed S a es indica es ha he sp ead o COVID-19
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 3 / 19
and lockdown measu es a he beginning o he pandemic inc eased suppo o candida es
ep esen ing he “s a us quo” [24].
Al hough hese di e en heo e ical backg ounds speak in a ou o a posi i e associa ion
be ween he ci cula ion o COVID-19 and suppo o incumben candida es, a nega i e asso-
cia ion can also be concei able o wo di e en easons. Fi s , gi en accumula ed e idence
ha “ ea , h ea , and anxie y o en con ibu e o igh -wing ex emism” [47], i is possible ha
he sp ead o COVID-19 shi s o e s away om incumben s, i incumben s a e p edomi-
nan ly no ex emis s. Al hough his heo e ical possibili y is plausible, i seems highly unlikely
o occu in he con ex o local elec ions in F ance, whe e he o e sha e o ex eme igh -
wing candida es is negligible.
Second, i is concei able ha whe e he i us ci cula es mo e, i is mos ly isk ake s who go
o o e, because isk-a e se o e s decide o s ay home in he i s place. Thus, in hese a eas
we should obse e a decline in bo h u nou and o e o incumben s. Al hough his heo e ical
channel is also plausible, i es s on he assump ion ha isk ake s a e pa icula ly likely o
choose challenge s o e incumben s. Howe e , i is also possible ha isk ake s p e e incum-
ben s o e challenge s o pu ely ins umen al easons ela ed o he p ospec i e channel
desc ibed below, in line wi h e idence ha ins umen al isk aking is linked o mo e a ional
decisions [48] and ha isk ake s “ end o decide mo e s ic ly on he basis o cos -bene i
conside a ions” [49].
The discussion o hese al e na i e scena ios leads o he second channel ela ed o a simple
e sion o p ospec i e o ing, ha is, o ing on he basis o how candida es will pe o m in he
upcoming adminis a ion. This channel is based on he simple conside a ion ha du ing a c i-
sis “i is no a good ime o a change.” In line wi h he idea o p ospec i e o ing [39,50,51],
o e s compa e how challenge s e sus incumben s will pe o m in ela ion o con aining he
i us. An icipa ing ha he h ea o COVID-19 is going o las , and gi en ha i ually no
candida e has a ack eco d in dealing wi h a pandemic, o e s migh conside ha incum-
ben s will pe o m be e simply because hey ha e had se e al yea s o expe ience wi h he
local adminis a ion and can swi ly mobilize all he a ailable local esou ces. Fo ins ance,
F ench mayo s had o coope a e wi h exis ing local ac o s (schools, local NGOs, and olun-
ee s) du ing he pandemic o sha e in o ma ion, dis ibu e masks, o coo dina e wi h o he
adminis a ions. Vo e s could in e p e he poli icians’ abili y o wo k wi h local ac o s as a sig-
nal o be e u u e managemen o he c isis. Because hey al eady knew hese local ac o s, he
incumben s p obably had an ad an age o e he o he candida es. Fo hese easons, in a sce-
na io o a looming h ea , o e s migh p e e o main ain con inui y wi h he cu en adminis-
a ion as he mos e icien means o ackling he i us and es o ing sa e y. On he o he
hand, choosing a challenge o e he incumben migh delay he necessa y esponse o he
pandemic due o he pe iod o ansi ion o he new adminis a ion. Fu he mo e, o e s
migh simply conside ha in a si ua ion o high unce ain y p oduced by he ci cula ion o
he i us, elec ing a challenge would add u he unce ain y ega ding how he local adminis-
a ion will deal wi h he pandemic. Thus, hey migh p e e incumben s as a sa e way o p e-
se ing public sa e y.
Bo h he emo ional channel and he p ospec i e- o ing channel lead us o he expec a ion
ha suppo o incumben s should inc ease in a eas whe e he sp ead o COVID-19 was
highe a he ea ly s age o he pandemic. Acco ding o hese heo e ical channels, suppo o
incumben s should inc ease ega dless o hei poli ical a ilia ion, as o e s h ea ened by
COVID-19 u n o incumben s mainly because hey ep esen he s a us quo. This said, p e i-
ous s udies indica e ha i is mos ly igh -wing incumben s ha gain om mili a y h ea s
because hei agg essi e de ense s a egies a e seen as mo e in line wi h ci izens’ need o p o-
ec ion compa ed o o he pa ies’ s a egies [52,53]. Al hough he sp ead o COVID-19 di e s
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 4 / 19
om mili a y h ea s, i igh -wing incumben s a e be e sui ed o educing anxie y and
add essing ci izens’ need o sa e y and secu i y, we migh expec he e ec o be s onges o
igh -wing incumben s. This expec a ion is also in line wi h ci ed e idence ha ea and anxi-
e y con ibu e o a shi owa ds igh -wing poli ical a i udes ( o a e iew, see [47]).
3 Municipal elec ions in F ance
On Ma ch 15, 2020 F ench o e s wen o he polls o choose he mayo s o all o he nea ly
35,000 municipali ies in he coun y. Municipal elec ions in F ance a e held e e y six yea s,
and, depending on he size o he municipali y, consis o one o up o wo ounds. In small
municipali ies wi h unde 1,000 inhabi an s, o e s choose lis s o candida es and a e allowed
o cas a o e o mo e han a single candida e in a majo i a ian sys em. In addi ion, he candi-
da es in owns wi h less han 1000 inhabi an s a e no equi ed o decla e any poli ical a ilia-
ion. In la ge owns and ci ies wi h mo e han 1,000 inhabi an s, elec ions consis o one o
wo ounds. In he i s ound, all lis s o candida es s anding o he elec ion compe e agains
each o he . In case no lis ob ains mo e han 50 pe cen o he o e, a un-o elec ion is held in
which all he lis s ha ob ained mo e han 10 pe cen o he o es compe e again o elec ion.
This second ound is usually held wo weeks a e he i s ound.
In 2020 he i s ound o he municipal elec ions was held as scheduled, despi e a pa ial
lockdown in some places. Howe e , as he numbe o cases o COVID-19 con inued o
inc ease, he day a e he i s ound o he elec ions, P esiden Mac on announced ha he
second ound would be pos poned o an unspeci ied da e, which was la e se o June 28, 2020.
Compa ed o he p e ious local elec ions in 2014, u nou d opped subs an ially. While in
2014, 63.5 pe cen o eligible o e s wen o cas hei ballo in he i s ound o he elec ions,
on Ma ch 15, 2020 less han hal o eligible o e s (44.7 pe cen ) wen o o e. This d op in pa -
icipa ion was likely due o he ci cula ion o he i us, as con i med by ecen s udies [54–56].
4 Da a and me hods
4.1 Vo ing and census da a
We e ie e elec o al da a o he municipal elec ions in 2020 and 2014 om he Minis y o
he In e io [57–59]. We conside only he municipali ies ha in bo h elec ion ounds had
mo e han 1000 inhabi an s due o he di e ences in elec o al ules explained abo e. We also
exclude o e seas e i o ies and he ci ies o Pa is, Ma seille and Lyon, since he o ing sys em
also di e s in hese a eas. A e excluding hese obse a ions, we a e le wi h a sample o 8,193
municipali ies.
To de e mine which candida es unning in 2020 we e incumben s, we ma ched he names
o hose who we e elec ed mayo in 2014 wi h he names o he 2020 candida es. This ope a-
ion gi es us ou e ec i e sample o 4,952 municipali ies (a ound 56%) in which an incumben
was unning in 2020. Balance s a is ics indica e ha municipali ies wi h and wi hou incum-
ben s a e s a is ically indis inguishable ac oss all ou indica o s, apa om popula ion densi y
and u nou in 2014, as municipali ies wi h an incumben a e mo e densely popula ed and
ha e a highe u nou in 2014 han municipali ies wi hou an incumben (see S1 Table in S1
File). Ou ou come o in e es is he sha e o o es o he incumben s in he i s ound o he
municipal elec ions in 2020. C ucially, in all eg ession models we include he o e sha e o
he same candida es in 2014, which allows us o es ima e whe he he sp ead o COVID-19
in luences he change in o e sha es o incumben s in 2020 compa ed o 2014.
We also e ie e in o ma ion abou he poli ical a ilia ion o he incumben candida es o
es o he e ogeneous e ec s depending on he incumben s’ poli ical leaning. In 2014, all can-
dida es compe ing in municipali ies abo e 1000 inhabi an s had o indica e hei poli ical
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 5 / 19
a ilia ion, which was hen classi ied by he Minis y o he In e io . We use his in o ma ion o
dis inguish be ween incumben s ela ed o le -wing pa ies (22%), igh -wing pa ies (47%),
and a esidual ca ego y including mos ly candida es wi h no clea le - igh a ilia ion and a
ew cen is candida es (31%). In S3 Table (S1 File) we p o ide de ailed in o ma ion abou
each poli ical g oup.
Fu he mo e, we collec census da a a he municipali y le el om he F ench na ional
ins i u e o s a is ics o he yea 2017 o con ol o key socio-economic indica o s ha migh
con ound he ela ionship be ween he sp ead o COVID-19 and o ing o incumben s [60].
Speci ically, in each model we include measu es o popula ion densi y, male o e emale a io,
he sha e o people aged abo e 65, he sha e o immig an s, he sha e o blue-colla wo ke s,
he sha e o unemployed inhabi an s, he median income o households in he municipali y,
and he sha e o people wi h a junio high school deg ee (“CAP” o “BEP” deg ee) and wi h a
bachelo ’s deg ee. We use he logged ans o ma ion o each o hese a iables (apa om he
median income) in o de o accoun o he p esence o clea ou lie s. Fu he mo e, we con ol
o he le el o u nou in 2014 o accoun o imbalances in o e s’ elec o al pa icipa ion a
he municipali y le el, and o he numbe o candida es unning in he i s ound o he elec-
ions. The la e a iable is pa icula ly impo an , since he incumben o e sha e is pa ially
de e mined by he le el o elec o al compe i ion in a gi en municipali y (wi h highe o e
sha e whe e ewe candida es compe e). A di ec ed acyclic g aph (DAG) jus i ying he choice
o con ol a iables is included in he supplemen a y S2 Fig in S1 File. Fo summa y s a is ics,
see S2 Table in S1 File.
4.2 Es ima ing he sp ead o COVID-19
Ou co e independen a iable is a municipali y’s deg ee o a ec edness by COVID-19. As no
such measu e was eadily a ailable, we es ima ed i ou sel es using s anda d epidemiological
me hods. The aim o his sec ion is o p o ide an o e iew o how his measu e is cons uc ed
(addi ional echnical de ails a e a ailable in Appendix C (S1 File)). Speci ically, we sough o
cons uc a ine-g ained measu e o he sp ead o Co id-19 a he municipali y le el a he
ime o he elec ion, which we deno e Co
m,20
. This quan i y is hen used as an independen
a iable o p edic he o e sha e ecei ed by incumben s, wi h he aim o measu e he impac
o COVID-19 on o ing beha iou .
Es ima ing he sp ead o COVID-19 a he municipali y le el is challenging because o sca -
e ed da a a ailabili y and he low eliabili y o common measu es. Fo ins ance, epo ed
cases o COVID- 19 a he beginning o he pandemic we e sensi i e o di e ences in es ing
equencies ac oss he coun y and unde es ima ed he ci cula ion o he i us [15], while da a
abou hospi alisa ions a e no a ailable a he municipali y le el since hospi als a e loca ed
only in a ew ci ies. To es ima e he sp ead o COVID-19 a he municipali y le el, we he e o e
u ilized exhaus i e mo ali y eco ds a he municipali y le el o es ima e COVID-19 p e a-
lence in each municipali y a he ime o he elec ion. We ook in o accoun he di e en ial
p obabili y o dying depending on age and sex ( he In ec ion Fa ali y Ra io (IFR, [14])), which
essen ially weigh s he excess mo ali y da a by he age and sex s uc u e o he popula ion a
he municipali y le el.
Mo e p ecisely, we i s compu ed he excess mo ali y (o excess haza d) a he municipal-
i y le el by collec ing all mo ali y eco ds be ween 2015 and 2020 [61]. We es ima ed excess
mo ali y a ound he ime o he i s ound o elec ions by compa ing mo ali y igu es o
hose o he same pe iod 2015–2019, hus a oiding issues o seasonal a iabili y. Excess mo al-
i y has been p esen ed as a eliable me hod o es ima e he numbe o COVID-19 cases because
mo ali y appea s ai ly s able o e a pe iod o a ew yea s, making ex eme mo ali y e en s
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 6 / 19
clea ly isible in he o m o peaks [62,63]. Yea ly in luenza ou b eaks, o ins ance, a e clea ly
iden i iable in coun y-wide weekly mo ali y eco ds (see S4 Fig in S1 File), while agg ega ed
mo ali y a es seem easonably s able o e he las hal -decade o he pe iod conside ed (see
S5 Fig in S1 File). Gi en ha he e is a delay be ween in ec ion and dea h, we conside all
dea hs occu ing be ween elec ion day (Ma ch 15, 2020) o six weeks la e , which should p o-
ide in o ma ion abou he numbe o in ec ed a he ime o he elec ion. The e is a isk o
simul anei y bias: (i) he sp ead o he disease can a ec he o e sha e o incumben s, bu (ii)
u nou —and hus o e sha es o incumben s—i sel can a ec he sp ead o COVID-19 [64].
In a supplemen a y analysis, we conside a sho e ime window o ou weeks, less likely o be
a ec ed by he simul anei y issue. We ob ain subs an ially simila esul s wi h his ou -week
window (see S7 Table in S1 File).
Schola s ha e a gued ha excess dea h may be ela ed indi ec ly o COVID-19 h ough he
sa u a ion o hospi al acili ies, in e up ion o p e en a i e p og ammes o he impac o
social es ic ions on physical and men al heal h [62]. Howe e , i seems unlikely ha hese
ac o s played a s ong ole a he e y beginning o he pandemic. Ins ead, esea che s ha e
a gued ha he p obabili y o dying om COVID-19 once in ec ed by he i us (In ec ion
Fa ali y Ra io, IFR) is hea ily de e mined by age and sex [14] and ha excess mo ali y should
be weighed by hese ac o s i we in end o use i as a p oxy o he ci cula ion o he i us. In
his sense, a 5% excess mo ali y in a ela i ely young ci y indica es ha SARS-Co -2, he i us
causing COVID-19, is ci cula ing mo e han in a ela i ely old ci y wi h he same le el o excess
dea h since young people a e less likely o die om COVID-19 han elde ly people. The e o e,
we build a model o es ima e COVID-19 p e alence ( he p opo ion o he popula ion in ec ed
by COVID-19) a he municipali y le el om mo ali y da a and use i as a no el measu e
desc ibing he sp ead o COVID-19, which weigh s excess dea h by age and sex-speci ic in ec-
ion a ali y a ios o es ima e he numbe o cases mo e p ecisely.
Speci ically, we use a ype o Poisson mo ali y model whe e he municipali y-le el baseline
haza d o e he 6 weeks pe iod is conside ed cons an be ween 2015 and 2019 and dea h
coun s in 2020 esul om he baseline haza d plus an excess haza d due o COVID-19. Excess
haza d is ela ed o p e alence modula ed by he age and sex-speci ic In ec ion Fa ali y Ra io
om [14]. We ob ain a municipali y-le el measu e o p e alence, along wi h age and sex-spe-
ci ic baseline haza ds o each municipali y. Popula ion da a is ob ained om Census da a
om 2010 o 2017 [65] and hen ei he linea ly o log-linea ly ex apola ed o 2018–2020 (see
de ails in Appendix C.2 (S1 File)). To deal wi h he p esence o ou lie s among he municipali-
ies, in he eg ession models we use ei he a decile e sion ( ea ed as a con inuous a iable)
o a qua ile e sion o ou measu e ( o es o non-linea e ec s). In supplemen a y analyses,
we show ha subs an ially simila esul s a e ob ained using a logged ans o ma ion o a quin-
ile e sion o ou measu e (see S4 and S5 Tables in S1 File).
Al hough in he majo i y o F ench municipali ies he SARS-Co -2 i us was ei he no
p esen o ba ely sp eading on elec ion day on 15 Ma ch 2020, we es ima e ha in a small bu
sizeable numbe o ci ies o e s had a ela i ely high p obabili y o knowing o he people wi h
COVID-19 symp oms. Assuming ha he a e age o e knows a ound 15 people, and assum-
ing homogenei y among o e s, based on ou p e alence measu e we es ima e ha on 15
Ma ch 2020 he e we e 654 ci ies (7 pe cen o he o al) and 255 ci ies (3 pe cen o he o al)
in which a pe son had a p obabili y o a leas 30 pe cen o 50 pe cen espec i ely o knowing
a leas one pe son wi h COVID-19 symp oms in hei ci cle o iends and acquain ances.
This is a e y conse a i e es ima e, conside ing a small social ne wo k o 15 people ins ead o
he ne wo k o 150 people calcula ed by Dunba [66], o accoun o he ac ha no all
COVID-19 in ec ed people may expe ience symp oms o sha e ha in o ma ion in hei social
ne wo k. This s ems om he ac ha e en i only ew people su e om COVID-19 in a
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 7 / 19
municipali y, he p obabili y ha a andom o e knows a leas one pe son wi h COVID-19
symp oms can be la ge—a coun e -in ui i e inding in line wi h he so-called “bi hday pa a-
dox” [67,68]. These basic es ima es sugges ha on elec ion day he e was al eady a subs an ial
numbe o o e s who we e awa e ha he i us causing COVID-19 was ci cula ing in hei
local a ea o esidence.
4.3 Es ima ing he e ec on incumben o ing
In o de o es ima e he e ec o COVID-19 on incumben o e sha es, we es ima e OLS
eg ession models in he o m
Im;d;20 ¼aþbCo m;20 þgXc; 1þZIm;d;14 þldþ�m;d; ð1Þ
whe e I
m,d,20
is he o e sha e ecei ed by he incumben in a gi en municipali y min 2020;
Co
m,20
measu es he sp ead o COVID-19 in a gi en municipali y in 2020; X
c, −1
is a ec o
including he municipali y-le el census a iables p esen ed abo e (measu ed in 2017), in addi-
ion o he le el o u nou in 2014 and a measu e o baseline mo ali y. The baseline mo ali y
is he municipali y-speci ic p obabili y o dying du ing he pe iod p io o he onse o he pan-
demic. The cons uc ion o bo h Co
m,20
and he baseline mo ali y was discussed in he p e i-
ous sec ion. Con olling o he baseline mo ali y is especially impo an since i cap u es
a ious con ibu ions o mo ali y—bo h obse ed and unobse ed— he eby adjus ing he
o he wise qui e di e se se o municipali ies in e ms o expec ed u u e mo ali y. I
m,d,14
is he
incumben ’s o e sha e in he elec ion 2014, and λ
d
a e coun y (Dépa emen ) ixed e ec s ha
emo e ime-in a ian ac o s a his le el om ou es ima ions.
Ou pa ame e o in e es is β. Since ou models include he lagged dependen a iable (i.e.,
o e o incumben s in 2014), βmeasu es he change in incumben suppo be ween he wo
elec ions ela ed o he sp ead o COVID-19, all else being equal. We op ed o his speci ica-
ion because i allows o mo e lexibili y in he way he independen a iable can be speci ied.
Below we show ha ou esul s also hold in a di e ence-in-di e ences amewo k, when
ma ching municipali ies using p opensi y sco es and a punishing calipe , and a e obus o a
placebo es .
5 Resul s
Fig 1 p esen s he esul s o ou eg ession models. Ou model explains a la ge pa o he a i-
a ion in o es o incumben s in 2020 (R-squa ed = 0.59). Suppo o incumben s in 2020 is
highe in municipali ies wi h highe popula ion densi y, highe sha e o people wi h a bache-
lo ’s deg ee, lowe sha e o people wi h a junio high school deg ee, lowe sha e o people aged
abo e 65, and lowe median income o households. As expec ed, o e sha e o incumben s
inc eases as elec o al compe i ion dec eases, as cap u ed by he numbe o candida es unning
in each municipali y ( o comple e esul s, see S4 Table in S1 File). Despi e he high explana-
o y powe o he eg ession model, ou measu e o he sp ead o COVID-19 has a posi i e
e ec : o e sha es o incumben s a e subs an ially highe whe e COVID-19 is mo e
widesp ead.
The e ical ba s in he le -hand plo o Fig 1 show he p edic ed o e sha e o incumben s
in 2020 a each decile o COVID-19 p e alence, ea ed as a ca ego ical p edic o , holding all
o he ac o s cons an . Despi e a ia ion, we obse e a oughly linea inc ease mo ing om
places wi h low o high ci cula ion o COVID-19. I we ea COVID-19 p e alence as a linea
p edic o ( he do ed line in he le -hand plo ), we ind ha suppo o local incumben s is
2.5 pe cen age poin s highe in he municipali ies wi h high ci cula ion o he i us ( en h dec-
ile) han in he municipali ies wi h low ci cula ion ( i s decile)–an inc ease ha is s a is ically
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 8 / 19
we lack indi idual-le el da a ha can add ess hese mechanisms di ec ly. We also call o u -
he esea ch in o he na u e o ally-a ound- he- lag e ec s. While ou indings a e in line
wi h e idence o ally e ec s a he beginning o he pandemic [17–20,22,36–38], we a gue
ha he h ea o COVID-19 di e s om he ype o h ea s ha , acco ding o p e ious s udies,
lead people o ally a ound he lag, such as e o is a acks and mili a y c ises. In pa icula ,
COVID-19 should no ac i a e he same eelings o ange and e enge ha ha e been heo ized
o explain ally e ec s [29], since he i us lacks he in en ion o a ack a popula ion. Fu he
esea ch should es whe he acing a na u al h ea (such as COVID-19) o a human h ea
(such as a e o is a ack) is associa ed o di e en emo ional s a es o ea o ange , which, in
u n, can lead o di e en beha iou al ou comes ( o a e iew, see [43]).
Las ly, we should no o e s a e he “powe ” o he incumbency s a us in in luencing o ing
beha iou . Al hough he e ec s we de ec can change he ou come o elec ions, hey a e likely
o be sho -li ed, in line wi h esea ch on he e ec s o e o is a acks [31] and COVID-19 as
well [76]. As he pandemic un olds, a numbe o conside a ions p obably in luence o e s’
decisions, including he e alua ion o how incumben s and poli ical ep esen a i es esponded
o he sp ead o he i us [77]. Indeed, e idence indica es ha suppo o go e nmen s in
Wes e n democ acies has dec eased du ing he cou se o he pandemic [78]. In his sense, o -
e s a e no blind o he pe o mance o hei poli ical ep esen a i es. A e he ini ial emo-
ional eac ion o he sp ead o COVID-19, i is likely ha o he conside a ions shape o ing
decisions.
Suppo ing in o ma ion
S1 File. Suppo ing in o ma ion including addi ional analyses.
(PDF)
Acknowledgmen s
We hank ou h ee e iewe s and con e ence pa icipan s a he Collegio Ca lo Albe o and
he Eu opean Poli ical Science Associa ion’s Annual Mee ing 2021 o commen s on ea lie
e sions o his esea ch.
Au ho Con ibu ions
Concep ualiza ion: Da ide Mo isi, He
´loïse Clole
´ y, Guillaume Kon Kam King, Max Schaub.
Da a cu a ion: Da ide Mo isi, He
´loïse Clole
´ y, Guillaume Kon Kam King, Max Schaub.
Fo mal analysis: Da ide Mo isi, He
´loïse Clole
´ y, Guillaume Kon Kam King, Max Schaub.
Funding acquisi ion: Da ide Mo isi, Guillaume Kon Kam King, Max Schaub.
In es iga ion: Da ide Mo isi.
Me hodology: Da ide Mo isi, He
´loïse Clole
´ y, Guillaume Kon Kam King, Max Schaub.
P ojec adminis a ion: Da ide Mo isi.
Resou ces: Da ide Mo isi.
W i ing – o iginal d a : Da ide Mo isi, He
´loïse Clole
´ y, Guillaume Kon Kam King, Max
Schaub.
W i ing – e iew & edi ing: Da ide Mo isi, He
´loïse Clole
´ y, Guillaume Kon Kam King, Max
Schaub.
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 15 / 19
Re e ences
1. Bech el MM, Hainmuelle J. How Las ing Is Vo e G a i ude? An Analysis o he Sho - and Long-Te m
Elec o al Re u ns o Bene icial Policy. Ame ican Jou nal o Poli ical Science. 2011; 55(4):852–868.
2. Belloc M, D ago F, Galbia i R. Ea hquakes, Religion, and T ansi ion o Sel -Go e nmen in I alian Ci -
ies. The Qua e ly Jou nal o Economics. 2016; 131(4):1875–1926.
3. Healy A, Malho a N. Myopic Vo e s and Na u al Disas e Policy. Ame ican Poli ical Science Re iew.
2009; 103(3):387–406.
4. Hee sink B, Pe e son BD, Jenkins JA. Disas e s and Elec ions: Es ima ing he Ne E ec o Damage
and Relie in His o ical Pe spec i e. Poli ical Analysis. 2017; 25(2):260–268.
5. Masie o G, San a ossa M. Na u al Disas e s and Elec o al Ou comes. Eu opean Jou nal o Poli ical
Economy. 2021; 67(Janua y):101983.
6. Aka ca AT, Tansel A. Social and Economic De e minan s o Tu kish Vo e Choice in he 1995 Pa lia-
men a y Elec ion. Elec o al S udies. 2007; 26(3):633–647.
7. Cole S, Healy A, We ke E. Do Vo e s Demand Responsi e Go e nmen s? E idence om Indian
Disas e Relie . Jou nal o De elopmen Economics. 2012; 97(2):167–181.
8. E iksson LM. Winds o Change: Vo e Blame and S o m Gud un in he 2006 Swedish Pa liamen a y
Elec ion. Elec o al S udies. 2016; 41:129–142.
9. Gaspe JT, Ree es A. Make I Rain? Re ospec ion and he A en i e Elec o a e in he Con ex o Na u-
al Disas e s. Ame ican Jou nal o Poli ical Science. 2011; 55(2):340–355.
10. Healy AJ, Malho a N, Mo CH. I ele an E en s A ec Vo e s’ E alua ions o Go e nmen Pe o -
mance. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica. 2010; 107
(29):12804–12809. h ps://doi.o g/10.1073/pnas.1007420107 PMID: 20615955
11. Lay JC. Race, Re ospec i e Vo ing, and Disas e s: The Re-Elec ion o C. Ray Nagin a e Hu icane
Ka ina. U ban A ai s Re iew. 2009; 44(5):645–662.
12. Bech el MM, Mannino M. Re ospec ion, Fai ness, and Economic Shocks: How Do Vo e s Judge Policy
Responses o Na u al Disas e s? Poli ical Science Resea ch and Me hods. 2022; 10(2):260–278.
13. Campan e FR, Depe is-Chau in E, Du an e R. The Vi us o Fea : The Poli ical Impac o Ebola in he
U.S. Na ional Bu eau o Economic Resea ch; 2020. 26897.
14. O’D iscoll M, Ribei o Dos San os G, Wang L, Cummings DAT, Azman AS, Pai eau J, e al. Age-Speci ic
Mo ali y and Immuni y Pa e ns o SARS-CoV-2. Na u e. 2021; 590(7844):140–145. h ps://doi.o g/10.
1038/s41586-020-2918-0 PMID: 33137809
15. Russell TW, Hellewell J, Ja is CI, an Zand oo K, Abbo S, Ra nayake R, e al. Es ima ing he In ec-
ion and Case Fa ali y Ra io o Co ona i us Disease (COVID-19) Using Age-Adjus ed Da a om he
Ou b eak on he Diamond P incess C uise Ship, Feb ua y 2020. Eu osu eillance. 2020; 25(2000256).
h ps://doi.o g/10.2807/1560-7917.ES.2020.25.12.2000256 PMID: 32234121
16. Cen e o Sys ems Science and Enginee ing (CSSE) a Johns Hopkins Uni e si y. COVID-19 Da a
Reposi o y by he Cen e o Sys ems Science and Enginee ing (CSSE) a Johns Hopkins Uni e si y;
2021. h ps://gi hub.com/CSSEGISandDa a/COVID-19.
17. Bol D, Giani M, Blais A, Loewen PJ. The E ec o COVID-19 Lockdowns on Poli ical Suppo : Some
Good News o Democ acy? Eu opean Jou nal o Poli ical Resea ch. 2021; 60(2):497–505.
18. De V ies CE, Bakke BN, Hobol SB, A ceneaux K. C isis Signaling: How I aly’s Co ona i us Lockdown
A ec ed Incumben Suppo in O he Eu opean Coun ies. Poli ical Science Resea ch and Me hods.
2021; 9(3):451–467.
19. Die z M, Roß eu sche S, Sche e P, S o
¨ sand LC. Rally E ec in he Co id-19 Pandemic: The Role o
A ec edness, Fea , and Pa isanship. Ge man Poli ics. 2023; 32(4):643–663.
20. K i zinge S, Foucaul M, Lacha R, Pa heymu lle J, Plescia C, B oua d S. ‘Rally Round he Flag’: The
COVID-19 C isis and T us in he Na ional Go e nmen . Wes Eu opean Poli ics. 2021; 44(5–6):1205–
1231.
21. Sch a D. Poli ical T us du ing he Co id-19 Pandemic: Rally a ound he Flag o Lockdown E ec s?
Eu opean Jou nal o Poli ical Resea ch. 2021; 60:1007–1017. h ps://doi.o g/10.1111/1475-6765.
12425 PMID: 33362332
22. Yam KC, Jackson JC, Ba nes CM, Lau J, Qin X, Lee HY. The Rise o COVID-19 Cases Is Associa ed
wi h Suppo o Wo ld Leade s. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es
o Ame ica. 2020; 117(41):25429–25433. h ps://doi.o g/10.1073/pnas.2009252117 PMID: 32973100
23. Bech el MM, Ob och a W, Ta i s M. Can Policy Responses o Pandemics Reduce Mass Fea ? Jou nal
o Expe imen al Poli ical Science. 2022;: 1–11. h ps://doi.o g/10.1017/XPS.2022.7
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 16 / 19
24. Bisbee J, Honig D. Fligh o Sa e y: COVID-Induced Changes in he In ensi y o S a us Quo P e e ence
and Vo ing Beha io . Ame ican Poli ical Science Re iew. 2021;: 1–17.
25. Aka ca AT, Tansel A. Vo e Reac ion o Go e nmen Incompe ence and Co up ion Rela ed o he 1999
Ea hquakes in Tu key. Jou nal o Economic S udies. 2016; 43(2):309–335.
26. Achen CH, Ba els LM. Democ acy o Realis s: Why Elec ions Do No P oduce Responsi e Go e n-
men . P ince on, NJ: P ince on Uni e si y P ess; 2016.
27. Fio ina MP. Re ospec i e Vo ing in Ame ican Na ional Elec ions. New Ha en: Yale Uni e si y P ess;
1981.
28. Bake WD, Oneal JR. Pa io ism o Opinion Leade ship? Jou nal o Con lic Resolu ion. 2001; 45
(5):661–687.
29. Lambe AJ, Scho JP, Sche e L. Th ea , Poli ics, and A i udes: Towa d a G ea e Unde s anding o
Rally-’ ound- he-Flag E ec s. Cu en Di ec ions in Psychological Science. 2011; 20(6):343–348.
30. Muelle JE. P esiden ial Popula i y om T uman o Johnson. Ame ican Poli ical Science Re iew. 1970;
64(1):18–34.
31. Dinesen PT, Jæge MM. The E ec o Te o on Ins i u ional T us : New E idence om he 3/11 Mad id
Te o is A ack. Poli ical Psychology. 2013; 34(6):917–926.
32. Gaines BJ. Whe e’s he Rally? App o al and T us o he P esiden , Cabine , Cong ess, and Go e n-
men since Sep embe 11. PS: Poli ical Science & Poli ics. 2002; 35(3):531–536.
33. He he ing on MJ, Nelson M. Ana omy o a Rally E ec : Geo ge W. Bush and he Wa on Te o ism. PS:
Poli ical Science & Poli ics. 2003; 36(1):37–42.
34. Hin son J, Vaishna M. Who Rallies a ound he Flag? Na ionalis Pa ies, Na ional Secu i y, and he
2019 Indian Elec ion. Ame ican Jou nal o Poli ical Science. 2023; 67:342–357.
35. Pe in AJ, Smolek SJ. Who T us s? Race, Gende , and he Sep embe 11 Rally E ec among Young
Adul s. Social Science Resea ch. 2009; 38(1):134–145. h ps://doi.o g/10.1016/j.ss esea ch.2008.09.
001 PMID: 19569296
36. Baekgaa d M, Ch is ensen J, Madsen JK, Mikkelsen KS. Rallying a ound he Flag in Times o COVID-
19: Socie al Lockdown and T us in Democ a ic Ins i u ions. Jou nal o Beha io al Public Adminis a ion.
2020; 3(2):1–12.
37. Esaiasson P, Sohlbe g J, Ghe se i M, Johansson B. How he Co ona i us C isis A ec s Ci izen T us
in Ins i u ions and in Unknown O he s: E idence om ‘ he Swedish Expe imen ’. Eu opean Jou nal o
Poli ical Resea ch. 2020;1–13.
38. Nielsen JH, Lind all J. T us in Go e nmen in Sweden and Denma k du ing he COVID-19 Epidemic.
Wes Eu opean Poli ics. 2021; 44(5–6):1180–1204.
39. Leininge A, Schaub M. S a egic Alignmen in Times o C isis: Vo ing a he Dawn o a Global Pan-
demic. Poli ical Beha io . 2023.
40. Vasilopoulos P, Ma cus GE, Valen ino NA, Foucaul M. Fea , Ange , and Vo ing o he Fa Righ : E i-
dence om he No embe 13, 2015 Pa is Te o A acks. Poli ical Psychology. 2019; 40(4):679–704.
41. Hyland P, She lin M, McB ide O, Mu phy J, Ka a zias T, Ben all RP, e al. Anxie y and Dep ession in
he Republic o I eland du ing he COVID-19 Pandemic. Ac a Psychia ica Scandina ica. 2020; 142
(3):249–256. h ps://doi.o g/10.1111/acps.13219 PMID: 32716520
42. Pe zold MB, Bendau A, Plag J, Py kosch L, Masca ell Ma icic L, Be zle F, e al. Risk, Resilience, Psy-
chological Dis ess, and Anxie y a he Beginning o he COVID-19 Pandemic in Ge many. B ain and
Beha io . 2020; 10(9):e01745. h ps://doi.o g/10.1002/b b3.1745 PMID: 32633464
43. Wagne M, Mo isi D. Anxie y, Fea , and Poli ical Decision Making. Ox o d Resea ch Encyclopedia o
Poli ics. 2019;.
44. G eenbe g J, Solomon S, Pyszczynski T. Te o Managemen Theo y o Sel -Es eem and Cul u al
Wo ld iews: Empi ical Assessmen s and Concep ual Re inemen s. In: Zanna MP, edi o . Ad ances in
Expe imen al Social Psychology. ol. 29. Academic P ess; 1997.: 61–139.
45. Bonanno GA, Jos JT. Conse a i e Shi among High-Exposu e Su i o s o he Sep embe 11 h Te -
o is A acks. Basic and Applied Social Psychology. 2006; 28(4):311–323.
46. Jos JT, Glase J, K uglanski AW, Sulloway FJ. Poli ical Conse a ism as Mo i a ed Social Cogni ion.
Psychological Bulle in. 2003; 129(3):339–375. h ps://doi.o g/10.1037/0033-2909.129.3.339 PMID:
12784934
47. Jos JT. Ange and Au ho i a ianism Media e he E ec s o Fea on Suppo o he Fa Righ —Wha
Vasilopoulos e al. (2019) Really Found. Poli ical Psychology. 2019; 40(4):705–711.
48. Zaleskiewicz T. Beyond Risk Seeking and Risk A e sion: Pe sonali y and he Dual Na u e o Economic
Risk Taking. Eu opean Jou nal o Pe sonali y. 2001; 15(1 SUPPL.):S105–S122. h ps://doi.o g/10.
1002/pe .426
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 17 / 19
49. Nadeau R, Ma in P, Blais A. A i ude owa ds Risk-Taking and Indi idual Choice in he Quebec Re e -
endum on So e eign y. B i ish Jou nal o Poli ical Science. 1999; 29:523–539.
50. Locke bie B. P ospec i e Economic Vo ing in U. S. House Elec ions, 1956–88. Legisla i e S udies
Qua e ly. 1991; 16(2):239–261.
51. Rosema M. Pa isanship, Candida e E alua ions, and P ospec i e Vo ing. Elec o al S udies. 2006; 25
(3):467–488.
52. Be ebi C, Klo EF. A e Vo e s Sensi i e o Te o ism? Di ec E idence om he Is aeli Elec o a e.
Ame ican Poli ical Science Re iew. 2008; 102(3):279–301.
53. Ge mansky A, Zei zo T. Te o ism and Vo ing: The E ec o Rocke Th ea on Vo ing in Is aeli Elec-
ions. Ame ican Poli ical Science Re iew. 2014; 108(3):588–604.
54. Le omain E, Vannoo enbe ghe G. Vo ing unde Th ea : E idence om he 2020 F ench Local Elec ions.
Eu opean Jou nal o Poli ical Economy. 2022; 75:102204. h ps://doi.o g/10.1016/j.ejpoleco.2022.
102204 PMID: 36440377
55. Nou y A, F an cois A, Ge gaud O, Ga el AHow Does COVID-19 A ec Elec o al Pa icipa ion? E idence
om he F ench Municipal Elec ions. PLOS ONE. 2021; 16(2):e0247026. h ps://doi.o g/10.1371/
jou nal.pone.0247026 PMID: 33626074
56. Picchio M, San olini R. The COVID-19 Pandemic’s E ec s on Vo e Tu nou . Eu opean Jou nal o Poli i-
cal Economy. 2022; 73:102161. h ps://doi.o g/10.1016/j.ejpoleco.2021.102161 PMID: 34975184
57. Minis è e de l’In e
´ ieu e des Ou e-me . Elec ions Municipales 2020—Re
´sul a s 1e Tou ; 2020. Da a.
Gou .F /F /Da ase s/Elec ions-Municipales-2020-Resul a s-1e -Tou /.
58. Minis è e de l’In e
´ ieu e des Ou e-me . Municipales 2020—Re
´sul a s 2nd Tou ; 2020. Da a.Gou .F /
F /Da ase s/Municipales-2020-Resul a s-2nd-Tou /.
59. Minis è e de l’In e
´ ieu e des Ou e-me . Elec ions Municipales 2014; 2014. Da a.Gou .F /F /Da a-
se s/?o ganiza ion=534 91a3a7292c64a77 53&q=%{{C3}}%{{A9lec ions}}%20 municipales%202014.
60. Ins i u na ional de la s a is ique e des e
´ udes e
´conomiques. S a is iques e e
´ udes; 2020. h ps://www.
insee. / /s a is iques.
61. Ins i u na ional de la s a is ique e des e
´ udes e
´conomiques. Fichie des pe sonnes de
´ce
´de
´es—De
´cès,
h ps://www.da a.gou . / /da ase s/ ichie -des-pe sonnes-decedees/, Downloaded on July 6, 2020;
2020.
62. Beaney T, Cla ke JM, Jain V, Goles aneh AK, Lyons G, Salman D, e al. Excess Mo ali y: The Gold
S anda d in Measu ing he Impac o COVID-19 Wo ldwide? Jou nal o he Royal Socie y o Medicine.
2020; 113(9):329–334. h ps://doi.o g/10.1177/0141076820956802 PMID: 32910871
63. Ka linsky A, Kobak D. T acking Excess Mo ali y ac oss Coun ies du ing he COVID-19 Pandemic wi h
he Wo ld Mo ali y Da ase . eLi e. 2021; 10:e69336. h ps://doi.o g/10.7554/eLi e.69336 PMID:
34190045
64. Cassan G, Sangnie M. The Impac o 2020 F ench Municipal Elec ions on he Sp ead o COVID-19.
Jou nal o popula ion economics. 2022; 35:963–988. h ps://doi.o g/10.1007/s00148-022-00887-0
PMID: 35345551
65. Ins i u na ional de la s a is ique e des e
´ udes e
´conomiques. S a is iques Locales—Densi e
´de Popula-
ion, h ps://s a is iques-locales.insee. /# iew=map1&c=indica o , Downloaded on July 10, 2020; 2020.
66. Dunba RIM. Neoco ex Size as a Cons ain on G oup Size in P ima es. Jou nal o Human E olu ion.
1992; 22(6):469–493. h ps://doi.o g/10.1016/0047-2484(92)90081-j
67. Von Mises R. Uebe Au eilungs- und Bese zungswah scheinlichkei en. Re ue de la Facul e
´des sci-
ences de l’Uni e si e
´d’Is anbul. 1939; 4:145–163.
68. Bo ja MC, Haigh J. The Bi hday P oblem. Signi icance. 2007; 4(3):124–127. h ps://doi.o g/10.1111/j.
1740-9713.2007.00246.x
69. T icaud C. Be e Alone? E idence on he Cos s o In e municipal Coope a ion. Condo ce Cen e o
Poli ical Economy; 2019.
70. IPSOS. Da aco id—Ba omè e COVID-19, Da a.Gou .F /F /Da ase s/Da aco id-Ba ome e- COVID-
19/; 2020.
71. Jennings W. Co id-19 and he ‘Rally-Round- he-Flag’ E ec . UK in a Changing Eu ope. 2020;.
72. Alesina A, Passa elli F. Loss A e sion in Poli ics. Ame ican Jou nal o Poli ical Science. 2019; 63
(4):936–947.
73. Mo gens e n S, Zechmeis e E. Be e he De il You Know han he Sain You Don’ ? Risk P opensi y
and Vo e Choice in Mexico. The Jou nal o Poli ics. 2001; 63(1):93–119.
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 18 / 19
74. Mo isi D, Jos JT, Panagopoulos C, Val onen J. Is The e an Ideological Asymme y in he Incumbency
E ec ? E idence om U.S. Cong essional Elec ions. Social Psychological and Pe sonali y Science.
2021; 0(0):1–11.
75. Qua one GA, T e sky A. Con as ing Ra ional and Psychological Analyses o Poli ical Choice. The
Ame ican Poli ical Science Re iew. 1988; 82(3):719–736.
76. Johansson B, Hopmann DN, Sheha a A. When he Rally-a ound- he-Flag E ec Disappea s, o : When
he COVID-19 Pandemic Becomes “No malized”. Jou nal o Elec ions, Public Opinion and Pa ies.
2021; 31(S1):321–334. h ps://doi.o g/10.1080/17457289.2021.1924742
77. Beche M, Longue Ma x N, Pons V, B oua d S, Foucaul M, Galasso V, e al. COVID-19, Go e nmen
Pe o mance, and Democ acy: Su ey Expe imen al E idence om 12 Coun ies. Na ional Bu eau o
Economic Resea ch; 2021. 29514.
78. Jø gensen F, Bo A, Lindhol MF, Bang M, Jø gensen F, Bo A. Public Suppo o Go e nmen
Responses agains COVID-19: Assessing Le els and P edic o s in Eigh Wes e n Democ acies du ing
2020 du ing 2020. Wes Eu opean Poli ics. 2021; 44(5–6):1129–1158.
PLOS ONE
How COVID-19 a ec s o ing o incumben s
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0297432 Ma ch 19, 2024 19 / 19
