Forerunners vs. latecomers—institutional competition in the German federalism during the COVID crisis

B eide, Lukas; Budzinski, Oli e ; G ebel, Thomas; Mendelsohn, Juliane
A icle — Published Ve sion
Fo e unne s s. la ecome s—ins i u ional compe i ion in
he Ge man ede alism du ing he COVID c isis
Eu opean Jou nal o Law and Economics
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: B eide, Lukas; Budzinski, Oli e ; G ebel, Thomas; Mendelsohn, Juliane (2025) :
Fo e unne s s. la ecome s—ins i u ional compe i ion in he Ge man ede alism du ing he COVID
c isis, Eu opean Jou nal o Law and Economics, ISSN 1572-9990, Sp inge US, New Yo k, NY, Vol. 59,
Iss. 1, pp. 101-132,
h ps://doi.o g/10.1007/s10657-025-09832-4
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/323359
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/
Vol.:(0123456789)
Eu opean Jou nal o Law and Economics (2025) 59:101–132
h ps://doi.o g/10.1007/s10657-025-09832-4
Fo e unne s s. la ecome s—ins i u ional compe i ion
in heGe man ede alism du ing heCOVID c isis
LukasB eide1· Oli e Budzinski1· ThomasG ebel1· JulianeMendelsohn1
Accep ed: 8 Janua y 2025 / Published online: 7 Feb ua y 2025
© The Au ho (s) 2025
Abs ac
Du ing he COVID-19 pandemic, poli ical compe i ion among he p emie s o Ge -
many’s 16 ede al s a es in ensi ied, wi h leade s s i ing o posi ion hemsel es
as p oac i e o e unne s in managing he c isis. This pape examines he iming
and de e mina ion o hese s a e leade s in announcing, legisla ing, and en o cing
COVID-19 policies, wi h a en ion o egional con ex s and speci ic de e minan s
in luencing hei ac ions. U ilizing mul iple dis inc da abases, we conduc a su -
i al analysis o assess each s a e’s poli ical esponse in ela i e e ms. Ou indings
e eal ha s a e leade s who we e ea ly ad oca es in public announcemen s and dis-
cou se did no necessa ily lead in o mal legisla ion o en o cemen o COVID-19
measu es. This s udy p o ides a nuanced iew o poli ical compe i ion in c isis go -
e nance, highligh ing he di e gence be ween poli ical he o ic and angible policy
ac ion ac oss Ge many’s ede al landscape.
Keywo ds Poli ical compe i ion· Ins i u ions· COVID· Su i al analysis
JEL Classi ica ion D7· H7· H11· H12· R5
* Thomas G ebel
[email p o ec ed]
Lukas B eide
[email p o ec ed]
Oli e Budzinski
oli e [email p o ec ed]
Juliane Mendelsohn
[email p o ec ed]
1 Ins i u e o Economics, Ilmenau Uni e si y o Technology, Eh enbe gs aße 29, 98693Ilmenau,
Thu ingia, Ge many
102
Eu opean Jou nal o Law and Economics (2025) 59:101–132
1 In oduc ion
Poli ical compe i ion wi hin ede al sys ems os e s bo h inno a ion and imi a ion
among s a e leade s, pa icula ly du ing c ises like he COVID-19 pandemic (Mis u
e al., 2023). Fede alism enables o e unne s, who ini ia e no el poli ical esponses,
and la ecome s, who adap s a egies p o en e ec i e by o he s, o shape public pol-
icy h ough a balance o p oac i e and eac i e app oaches. This dynamic, akin o
compe i ion in goods ma ke s, sugges s ha bo h s a egies can be e ec i e, wi h
o e unne s seeking i s -mo e ad an ages despi e inhe en isks, such as unp edic -
able public ecep ion (Rugge i e  al., 2024; Ba be á e al., 2019), while la ecom-
e s bene i om obse ing he challenges aced by ea ly adop e s, allowing hem o
a oid miss eps in policy implemen a ion. The u gency o a pandemic ampli ies his
in e play, as s a es wi hin a ede al sys em mus apidly espond o ex e nal shocks
(Sega o e al., 2022).
The decision o ac as a “ o e unne ” — o in oduce measu es ha a e pe cei ed
as necessa y o e ec i e—can enhance a leade ’s poli ical s anding i subsequen
e en s alida e hose ac ions. Con e sely, posi ioning onesel as a “la ecome ” by
wai ing cau iously be o e implemen ing po en ially es ic i e policies can mi iga e
isks o poli ical epu a ion. A nuanced s a egy seen du ing COVID-19 was he
p ac ice o ea ly policy announcemen s wi hou immedia e implemen a ion, allow-
ing leade s o appea esponsi e wi hou he immedia e poli ical cos s associa ed
wi h en o cemen (Sega o e al., 2022; Rugge i e al., 2024).
While he b oade dynamics o ede al compe i ion ha e long been obse ed,
he COVID-19 pandemic p o ides a unique empi ical oppo uni y o examine
o e unne s and la ecome s in Ge many’s ede al sys em, pa icula ly ac oss he
domains o public announcemen s, legisla ion, and en o cemen . Unlike cen al-
ized sys ems, he Ge man ede al s uc u e allows s a es (“Lände ”) o implemen
egion-speci ic ac ions wi hin a na ional legisla i e amewo k (Hegele & Schnabel,
2021). Al hough cen alized esponses may o e immedia e cohe ence du ing c i-
ses, ede al sys ems like Ge many’s a e a gued o be mo e adap able in add essing
egional a ia ions and unce ain ies o e ime (Bol on & Fa ell, 1990; Cepaluni
e al., 2022). Fo example, u al a eas, whe e social cos s o con ac es ic ions a e
lowe , may bene i om dis inc policy adjus men s compa ed o densely popula ed
u ban cen e s (Ca ozzi e al., 2024; Congle on, 2023). The Ge man In ec ion Con-
ol Ac — alling unde he ede al go e nmen ’s concu en legisla i e powe s (A .
74(1) N . 19, Ge man Basic Law) – explici ly pe mi s ede al s a es o implemen
such con ex - ailo ed egula ions wi hin he amewo k o ede al guidelines. How-
e e , by Ap il 2021, he Ge man Bundes egie ung shi ed om a p ima ily s a e-
coo dina ed pandemic esponse o a mo e cen alized egula o y amewo k, aiming
o educe ic ion bo h among he s a es and be ween ede al and s a e go e nmen s.
(Fä be , 2021).
Exis ing li e a u e on ede alism and c isis managemen , such as Congle on
(2023), highligh s he lexibili y ede al sys ems o e in ailo ing esponses.
Opposing iews, such as hose o Cepaluni e  al. (2022), emphasize ha high
s a e capaci y may be he essen ial ac o enabling leade s o implemen s ingen
103
Eu opean Jou nal o Law and Economics (2025) 59:101–132
measu es quickly, ega dless o ede al o uni a y sys em s uc u es. The sho -
e m ad an ages o cen alized esponses may decline as conside a ions o long-
e m social, economic, and psychological impac s necessi a e decen alized
adjus men s. In he case o Ge many, as poin ed ou , he p ocess was e e sed.
The ambigui y as o whe he a cen alized o a decen alized app oach o c i-
sis managemen is p e e able is due o he inhe en unce ain y o an unp ece-
den ed pandemic (Bjø nsko & Voig , 2022; Kollman e al., 2000). Leade s ace
unce ain y abou he mos e ec i e esponses, leading o expe imen a ion, policy
inno a ion, and an e ol ing selec ion p ocess (Ga za elli & Kee on, 2018; Samad
e al., 2022). They mus na iga e a complex se o in luences, including hei own
ideological leanings, he p e ailing public opinion, he p io i ies o hei espec-
i e pa ies, and he dynamics o hei poli ical coali ions. These ac o s gi e ise
o ya ds ick compe i ion (Ha ison, 1978; Bol on & Fa ell, 1990; Besley & Case,
1992; Fä be , 2021). Leade s seek o benchma k hei policies agains hose o
neighbo ing s a es, he eby os e ing a spi i o adap i e compe i ion.
This s udy con ibu es o he li e a u e by examining how s a e leade s wi hin
Ge many’s ede al sys em na iga ed hei oles as ei he o e unne s o la ecom-
e s in esponse o he COVID-19 pandemic. We analyze poli ical ac ions ac oss
h ee dimensions—public announcemen s, legisla i e ac ion, and en o cemen —
using a unique da ase compiled om LexisNexis, Ju is, and Des a is. This da a
enables us o quan i y compe i i e dynamics in policy-making and ca ego ize
s a e leade s acco ding o hei s a egic app oaches: “p oc as ina o s,” who make
ea ly policy announcemen s ye delay legisla i e ac ion; “a mchai ac i is s,”
who apidly legisla e bu exhibi leniency in en o cemen ; and “silen policy-
make s,” who e ec i ely lead in bo h legisla i e and en o cemen e o s wi hou
p io i izing media p esence. As such, ou analysis complemen s he “c edi and
blame”game (G ee e al., 2022) by using mo e g anula da a and digging deepe
in o he unde lying poli ico-compe i i e mechanisms. We show ha on he le el
o s a es and hei leade s, s a egies looking a he popula i y o esponse ins u-
men s also play an impo an ole as well.
Closes o ou esea ch is he s udy by B oschek (2022), which compa es h ee
esponse s a egies adop ed by Canadian p o inces du ing he pandemic: laissez-
ai e (al hough no Canadian p o ince adop ed his s a egy), mi iga ion (six p o -
inces), and con ainmen ( ou p o inces). Using a p ocess- acing app oach, he
pape es s se e al hypo heses abou whe he leade ship con ibu ed o he success
o ailu e o he pandemic esponse du ing he second and hi d wa es, and con-
cludes ha he con ainmen s a egy is supe io o comba he pandemic and ha
poli ical leade ship plays an impo an ole in bo h successes and ailu es. B oschek
(2022) calls o mo e esea ch on he ole and beha io o poli ical leade s du ing
such c ises. Ou pape ollows he call o mo e analysis o he ole o leade s and
ocuses on he mechanisms o ede al poli ical compe i ion, whe e leade s ha e an
eye on public opinion and eelec ion p obabili ies and hus may s a egically de ia e
om announcemen s, law codi ica ion, and en o cemen . In sho , we add ess he
(s a egic) inconsis ency be ween (media) announcemen s, law codi ica ion, and law
en o cemen in a ede al sys em. To he bes o ou knowledge, no p e ious wo k has
add essed his issue.
104
Eu opean Jou nal o Law and Economics (2025) 59:101–132
The emainde o his pape is s uc u ed as ollows: Sec ion2 p o ides an o e -
iew o he Ge man ede al sys em and pandemic esponse amewo k, and Sec-
ion3 p esen s desc ip i e s a is ics on o e unne s and la ecome s ac oss announce-
men , legisla ion, and en o cemen . Sec ion4 models he p opensi y o s a e leade s
o en o ce COVID policies, wi h eg ession esul s in Sec .5. Sec ion6 concludes.
2 Poli ical compe i ion ina ede al sys em
Ge many is a ede a ion o six een s a es, each o which enjoy a la ge scope o indi-
idual so e eign y. While he COVID-c ises was a na ional heal h c isis and, hus,
coo dina ed a he ede al le el, he implemen a ion and en o cemen o indi idual
measu es happened a s a e and egional le els. The indi idual s a es hus enjoyed
much poli ical disc e ion. In pa icula du ing he beginning o he pandemic his
c ea ed a di e se a ay o poli ical s a emen s being made and ules o he indi-
idual ci izens being implemen ed.
2.1 The scope o poli ical compe i ion
In o de o analyze he compe i ion leading o o e unne s and la ecome s, i helps o
p o ide an o e iew o he legal amewo k ha de e mines he e ical compe i ion
i s be ween he ede al s a e (Bund) and he indi idual s a es (Lände ), as his also
de e mines he scope o ho izon al compe i ion be ween he indi idual 16 ede al
s a es. We also show he shi s and changes in he amoun o execu i e disc e ion
a o ded o he indi idual 16 s a es du ing he pandemic, hus p o iding a imeline
o he scope o poli ical compe i ion.
Being a ede al coun y, bo h he ede al s a e (Bund) and s a es (Lände ) enjoy
legisla i e and adminis a i e compe ences. In ela ion o he s a es, he ede al le el
ei he has exclusi e, concu en , o ancilla y legisla i e powe s. Acco ding o he
cons i u ion, ede al legisla ion akes p ecedence o e s a e legisla ion (A . 31 Ge -
man Basic Law).
Fo dange ous and in ec ious diseases he e exis s a so-called concu en legis-
la i e powe (in Ge man: konku ie ende Gese zgebung). A 74 pa a. 1 N . 19 o
he Ge man Basic Law desc ibes his as he compe ence o “measu es o comba
human and animal diseases which pose a dange o he public o communicable”.1
This means ha he s a es can pass legisla ion only un il and inso a as he ede al
s a e has no done so. Wi h he Ge man Ac on he P e en ion and Con ol o In ec-
ious Diseases (In ek ionsschu zgese z, In SchG) o 20002 he ede al s a e has made
1 Fo an o icial e sion and ansla ion o he Fede al Basic Law (G undgese z), see h p:// gese ze- im-
in e ne . de/ gg/.
2 In ec ion P o ec ion Ac o July 20, 2000 (BGBl. I p. 1045), which was las amended by A icle 4 o
he law o Ma ch 18 2022 (BGBl. I p. 473). The law was in oduced as A icle 1G o . July 20, 2000 I
1045 (SeuchRNeuG) om he Bundes ag decided wi h he consen o he Fede al Council. Acco ding
o A icle 5, Pa ag aph 1, Sen ence 1, his law comes in o o ce on Janua y 1, 2001, Sec ions37 and 38
came in o o ce wi h e ec om July 26, 2000.

105
Eu opean Jou nal o Law and Economics (2025) 59:101–132
use o his powe o compe ence, hus p ecluding he ede al s a es om passing
any o mal laws o legal ac s. The indi idual s a es, howe e , implemen he p o i-
sions con ained in his ac and can pass egula ions (in Ge man:“Ve o dnungen”, i.e.
adminis a i e o de s) o do so.
Du ing he COVID-pandemic, his ede al ac (In SchG) saw many ounds o
amendmen s and is he p ima y sou ce o law go e ning all measu es and deci-
sions passed in ela ion o he pandemic. In addi ion, he ede al s a e (Bund) has
passed o he ac s and egula ions (pe aining o a el es ic ions, labou condi ions,
p o ec i e accina ions) as well.3 The ede al In SchG con ains se e al subs an i e
measu es o ules, such as social dis ancing, he wea ing o masks, he closing o
schools, o mi iga e he e ec s o he pandemic and con ol he sp ead o he i us.
Wi hin he amewo k o he ede al law, he s a es en o ce hese measu es by i s
passing speci ic egula ions (execu i e laws). A he ou se o he pandemic, he
ede al s a es enjoyed conside able disc e ion in doing so – in de ining he speci ic
measu es, hei scope and he ime-line o hei implemen a ion. Du ing he cou se
o he pandemic, howe e , he scope o he s a e egula ions dec eased signi ican ly.
The execu ion o en o cemen and hus he p ac ical implemen a ion o hese eg-
ula ions (execu i e o de s) was no done by he s a es (Lände ) hemsel es bu by
smalle adminis a i e uni s, by coun ies and dis ic s (in ci y s a es such as Be lin,
Hambu g and B emen). Se e al s a es also passed ancilla y egula ions on sanc ions
( ines) and in e p e a ions o he p ima y egula ion.4
The disc e ion gi en o he indi idual s a es (Lände ) changed signi ican ly du -
ing he cou se o he pandemic as e e mo e uni o m ules we e se on he ede al
le el which also included implemen a ion guidelines and guidance on which in e -
es s o conside when exe cising disc e ion. An assessmen /opinion o he legal se -
ice o he ede al pa liamen om 29Ma ch, 2021 ound ha he ede al s a e has
all-encompassing and a - eaching legisla i e compe ences in ela ion o all meas-
u es pe aining o he COVID-pandemic and could also speci y he implemen a ion
o hese measu es in ull de ail. This means ha he powe s and disc e ion o he
s a es could ha e been “ educed o ze o”, including measu es ela ed o schools and
heal h acili ies. By 22Ap il, 2021, he Fede al In ec ion P o ec ion Ac (In SchG)
was so de ailed and expansi e ha no signi ican compe i ion among he indi idual
s a es could s ill ake place. F om his poin onwa d, he compe i ion o announcing
and implemen ing p e en a i e measu es was la gely eplaced by he oll-ou o he
accina ion. Mos o he es ic i e measu es on ede al and s a e le el ha e now
been li ed. The legal amewo k and he in oduc ion o accina ions hus de ines
he imeline and scope o ou analysis.
Ha ing cla i ied he scope o ac ion, we now u n o he compe i i e mechanisms
and incen i es wi hin he Ge man ede al sys em.
3 COVID-19-Schu zmaßnahmen-Ausnahmen e o dnung (SchAusnahmV); Co ona i us-Ein eise e -
o dnung (Co onaEin eiseV); SARS-CoV-2-A bei sschu z e o dnung (Co ona-A bSchV); Ve o dnung
zum Ansp uch au Schu zimp ung gegen das Co ona i us SARS-CoV-2 (Co onaImp V); Ve o dnung
zum Ansp uch au Tes ung in Bezug au einen di ek en E ege nachweis des Co ona i us SARS-CoV-2
(Tes V).
4 Find a lis he e: h ps:// www. wobi ds. com/ de/ insig h s/ 2021/ ge ma ny/ co id- 19- e o dnung en- und-
e u egung en- bl.
106
Eu opean Jou nal o Law and Economics (2025) 59:101–132
2.2 Leade ship in imeso c isis
The dynamics o poli ical compe i ion in a ede al sys em a e mul idimensional and
include bo h ho izon al and e ical e ec s. Focusing on he si ua ion o s a e-le el
leade s in he Ge man ede alism and hei ho izon al compe i ion among each o he
e eals a numbe o in e es ing compe i ion mechanisms ha a e ele an o ou
pape .5 Gene al ema ks a e made in he ollowing be o e he special cha ac e is ics
and e ec s o an unexpec ed c isis like he pandemic wi h i s inhe en unce ain y
a e in oduced in he second hal o his sec ion.
Fi s , leading s a e-le el poli icians (in pa icula , s a e p emie s), compe e o
o e s wi h all he ypical ca ea s and impe ec ions o he poli ical p ocess in a ep-
esen a i e democ acy, i.e. impe ec knowledge, asymme ic in o ma ion, a ional
igno ance, e c. He e, p o iding high-quali y policy solu ions o ele an p oblems
o he popula ion na u ally ep esen s a channel o compe i ion.
Second, s a e-le el leade s s and in compe i ion wi h hei ellow s a e-le el lead-
e s as o e s may obse e and assess (bo h impe ec ly) wha happens in o he Ge -
man s a es. Thus, also he ela i e quali y o hei policies plays a ole.
Thi d, media a en ion is a ela i e pa ame e o poli ical compe i ion. By win-
ning he a en ion o hei own and he o he s a es’ popula ion, s a e leade s may
imp o e hei e-elec ion chances and, a he same ime, quali y o a ca ee on he
ede al le el by gaining popula i y ac oss he s a es. Depending on hei indi idual
p e e ences, poli ical leade s may be d i en by di e ing weigh s o ideology (poli i-
cal belie s abou he bes solu ions) and ca ee conce ns like he p obabili ies o ge -
ing e-elec ed o ge ing in o a ac i e poli ical o o he (pos -poli ician) posi ions
in he u u e. No e ha ideology and e-elec ion desi e may also co ela e in he
sense ha wi hou ge ing e-elec ed, no poli ician can implemen poli ical solu ions
acco ding o his o he ideology. No o iously, i is di icul o impossible o disen an-
gle hese elemen s o an indi idual poli ician’s u ili y unc ion. In his pape , we pu
a special emphasis on his dimension as i also sheds some ligh on he mo i a ion o
poli icians.
We assume ha s a e-le el poli ical leade s ins umen alize all h ee dimensions
o posi ion hemsel es a o ably in his ho izon al poli ical compe i ion. S a egy
choice is no i ial in a mul i-dimensional s a egy space unde impe ec in o ma-
ion, whe e o e we expec ha di e en poli ical leade s will iew di e en s a e-
gies as indi idually op imal o hem due o hei indi idual u ili y unc ions and due
o di e en weighing o he p os and cons and he dimensions ske ched abo e.
In imes o unexpec ed c isis, he knowledge p oblem is pa icula ly ele an as
he e is usually no bluep in o success ully dealing wi h he c isis p oblems. In
o he wo ds, i is ex an e unclea wha a “high quali y poli ical solu ion” may be. A
he same ime, he popula ion expec s hei leade s o p o e hemsel es as success ul
c isis manage s. The indi idual weigh o ideology and poli ical belie s in he u il-
i y unc ion o poli ical leade s may ac ually inc ease in he ace o he knowledge
5 In o de o keep ou pape compac , we do no discuss he huge body o gene al economics o ede al-
ism and poli ical economy in a ede al and democ a ic s a e. Fo a ecen heo y-d i en app oach wi h
e e ence also o gene ally ele an li e a u e see, in e alia, Congle on (2023).
107
Eu opean Jou nal o Law and Economics (2025) 59:101–132
p oblem since a eliance on one’s own belie s may ill he oid o (ex e nal, empi i-
cal, a ailable) p oblem-sol ing knowledge. Al e na i ely, and depending on he indi-
idual leade , he unce ain y may delay o e en pa alyze decision making because
o he ea o doing he “w ong” hing. Unce ain y also ela es o he ques ion o
whe he o e s will ac ually use he nex elec ion as a e e endum on he COVID-
19 policies o he espec i e poli ical leade s. Du ing he pandemic heydays, his is
clea ly one o he mos impo an issues, bu he nex elec ions may be a long way
o , and he pandemic u bulence may (o may no ) ha e eceded in impo ance o
o e s’ decisions by hen. In imes o c isis, howe e , i is unclea ex an e how long
he c isis will las and how impo an he beha io du ing he pandemic ime will be
o he nex elec ion. This is unknown bo h o he poli icians and he o e s. How-
e e , we assume ha poli ical leade s a e in e es ed in hei cu en popula i y— o
a ying deg ees (see abo e)—and i will ma e o hem how hey a e in he eyes
o he elec o a e when hey decide wha o do. Gi en he widesp ead unce ain y
on leng h and cou se o he pandemic, we hink ha his assump ion is easonable.
This is especially ue because he popula i y gained h ough inc eased media co -
e age–essen ially a popula i y ha makes a egional poli ician ge s be e known a
he ede al le el—is likely o las beyond he pandemic and will be e y help ul in
ascending in highe o ice in he u u e— ega dless o whe he he pandemic las s
and emains impo an o o e s o no .
In such a scena io, o which he COVID-19 pandemic is a ep esen a i e case,
he s a egy o ac ing as a o e unne comes wi h ad an ages and disad an ages.
A common expec a ion o poli ical leade s in imes o c isis is he e ec ha he
espec i e poli icians in cha ge ake ini ia i e and p o ide a clea di ec ion in he
c isis ha he popula ion can ollow. Thus, o e unne s can be expec ed o be p aised
by he p ess and he public o hei decisi e ac ion. They may be seen as s ong
leade s who, in spi e o obs acles, mo e o wa d wi h de e mina ion, se he cou se
o hei s a e, communica e clea ly and o en wi h he ci izens, and also enjoy
inc eased popula i y in his phase because hey “don’ alk, hey ac ”.
Howe e , while consequen and clea measu es may be popula ex an e, his may
change when he popula ion ac ually su e s om hem and when, in he cou se o
ime, doub s eme ge ha hese in e en ions we e necessa y o he “ igh ” ones. I is
possible ha o e s will hink ha poli ical leade s ha e o e eac ed because o he
panic na u e o he si ua ion.6 I ha happens, he ini ial popula i y o o e unne s
u ns ou o be sho -li ed and he mid e m e ec s may be nega i e.
La ecome s, on he o he hand, may ep esen a na u al eac ion o he unce -
ain y o he si ua ion and he lack o (knowledge abou ) op imal solu ions, hus
p omo ing a ional imi a ion o e en he d beha io . Ne e heless, hese poli ical
leade s emain in compe i ion wi h o he s and may be pe cei ed as oo cau ious,
lacking in leade ship and indecisi e in he ini ial phase o he c isis. Thei cau-
ious app oach o es ic ing ci il libe ies may ge sha ply c i icized and he e is
p essu e o ollow he o e unne s. Howe e , i he mo e cau ious app oach u ns
ou o ha e sa ed he popula ion om (pe cei ed o eal) unnecessa y ha d in e -
en ions in o hei daily li e and wel a e, hen he la ecome may gain epu a ion
6 See e.g. Jones & Baumga ne (2005).
108
Eu opean Jou nal o Law and Economics (2025) 59:101–132
as a “ hough ul pe son” p o iding measu ed policy esponses. Thus, he choice
whe he o ac immedia ely o a he wai is subjec o unce ain y.
The pic u e ge s e en mo e complica ed when mixed s a egies a e consid-
e ed. In eali y, poli ical leade s do no ha e o decide o ac as a o e unne
wi h all subsequen e ec s. Ins ead, hey could choose o be a o e unne in
announcing consequen in e en ions and measu es bu hen be ca e ul o codi y
hem in o law and/o ac ually en o ce hem. By spli ing s a egies—a o e un-
ne in announcemen s combined wi h a la ecome in codi ica ion and en o ce-
men — he ad an age o media a en ion (which always ocuses on he i s calls
o ac ions o a new measu e) may be combined wi h minimizing he dange o
ge ing sla ed by hei own popula ion su e ing ha de in e en ions han he
ci izens in o he s a es. Vice e sa, o he poli ical leade s may ind i a ac i e
o be compa a i ely silen a i s bu hen p o e o be pe sons o ac ion by being
o e unne s in codi ica ion and en o cemen o measu es, hus a oiding possi-
ble cha ac e iza ions as being “jus loudspeake s” o “populis s”. I codi ica ion
– i.e. he implemen a ion o a law/ egula ion in oducing a ce ain ins umen o
measu e—can be seen as a signal, hen spli ing s a egies be ween codi ica ion
and en o cemen may also make sense since only he en o cemen will ac ually
hu people by es ic ing hei li e and educing hei wel a e.
The dis inc ion be ween legisla ion ( he codi ica ion o speci ic measu es and hei
modes o implemen a ion in s a e-le el egula ions) and en o cemen (o hese meas-
u es) is, o he bes o ou knowledge, a no el con ibu ion o he discussion. Thus, i
is impo an o emphasize ha he dis inc ion be ween legisla ion and en o cemen is
no iden ical o he s ingency o deg ee o en o cemen o a policy measu e. Being
he i s o legisla e and/o en o ce does no au oma ically imply he adop ion o pa -
icula ly s ic policies. Du ing pe iods o exploding in ec ion a es, he e is a el-
e an likelihood o a ace owa d s ic e policies (among o e unne s), whe eas du -
ing pe iods o declining in ec ion a es—pe haps accompanied by ising accina ion
a es— he e may well be a ace owa d mo e lenien policies (among o e unne s).
I is impo an o emphasize ha he quali y o he policies—he e: a oiding dea hs
and con olling in ec ion wa es—plays he mos impo an ole. Howe e , especially
unde he assump ion o impe ec knowledge abou he “ igh ” an i-c isis policy—
he scope o s a egic ho izon al compe i ion may be used by poli ical leade s. The
ollowing empi ical in es iga ion is mean o shed ligh on he exis ence o o e un-
ne s and la ecome s as well as mixed s a egies du ing he pandemic in Ge many. O
cou se, we con ol o he main goal— he comba o dea h olls and in ec ion a es.
Be o e doing so, we p o ide a b ie o e iew o he da a used in ou s udy.
3 Da a anddesc ip i e s a is ics
As poin ed ou abo e, o e unne s and la ecome s can be iden i ied om di e en
pe spec i es. (1) who is i s (las ) in he media add essing a c isis- ela ed aspec ,
(2) who is i s (las ) o cas a policy measu e in o law, and (3) who se es as a
ole model in law en o cemen . Using di e en da a sou ces, we will y o iden-
i y he o e unne s and he la ecome s in he h ee ca ego ies: i s (las ) in media,
115
Eu opean Jou nal o Law and Economics (2025) 59:101–132
3.4 Law en o cemen andi s con ex ual ac o s
O he 24 agg ega ed measu es by he s a is ical o ice epo ed in Table 7 o he
appendix, we only use 21 ca ego ies o p o ec i e policies. Some o he measu es
had no been used by anyone and he e o e do no p o ide any u he in o ma ion
abou ede al s a es’ pandemic his o y. We ha e chosen he pe iod om Ma ch 1,
2020 o Janua y 15, 2022. Be o e his pe iod, no COVID policy had been imple-
men ed as he i us had no p o oked any poli ical ac ion. By he end o 2021, he
omic on a ian s a ed o sp ead. Since i was pe cei ed as a weake a ian , he
i us and pandemic los mo e and mo e o i s pe ilous na u e, especially as a la ge
majo i y o he popula ion had ecei ed i s second accina ion. Mo eo e , he in a-
sion o Russia in Uk aine sh unk he media- ele ance o he COVID pandemic e en
u he . Hence, we ended up wi h he in o ma ion abou 21 policy ca ego ies om
400 NUTS-3 egions om 16 s a es o 686 days, i.e., in o al 5,762,400 da a poin s.
As we a e in e es ed in o e unne s and la ecome s, we only look a he e en s
when a policy was in oduced. Wi hin he pe iod unde conside a ion, we iden i ied
47,301 momen s when a COVID policy was in oduced. Table 4 shows he a i-
ables ha we calcula ed om he a ailable in o ma ion. Because many policies we e
in oduced, abolished, and ein oduced se e al imes, we coun mo e han jus a sin-
gle in oduc ion o a gi en policy in a ce ain coun y. E e y policy was in oduced a
leas 4 imes, some e en 25 imes. The s a is ical o ice p o ides u he in o ma ion
on he NUTS-3 le el which we will also exploi .
The emaining a iables in he able a e indus y s uc u e, which anges be ween
10 and 30% and measu es he sha e in o al sales o nea -pe sonal se ices (Ho els
and es au an s, Educa ion, Heal h and Social Se ices, and A s, En e ainmen
and Rec ea ion); o al popula ion, and popula ion densi y; he numbe o ins alled
policies in all o he s a es as a measu e o imi a ion ( echnically, a measu e o spa-
ial co ela ion); he numbe o employees in pa ial employmen (i.e. Ku za bei o
in sho : KU). Be o e ou empi ical exe cise, we p esen a simple model – in line
wi h ou heo e ical e lec ions – ha mo i a es he p opensi y o egional leade s o
implemen a COVID policy.
Table 4 Summa y s a is ics on NUTS-3-le el
O e all, 400 NUTS-3- egions ha e been iden i ied. As we only ake he in o ma ion abou he day a
egion in oduces a policy, he numbe o obse a ions dec eases o 47,301
Va iables N Mean S and. de . Min Max
Numbe o policies 47,301 8.364 5.528 1 21
Numbe o p e ious policies 47,301 4.0 3.6 0.0 25.0
Indus y S uc u e 47,301 0.20 0.00 0.10 0.30
To al popula ion (NUTS-3, in housands) 47,301 210.8 231.6 34.0 3664.0
Popula ion densi y (pe squa e kilome e ) 47,301 553.6 712.8 36.0 4790.0
Dea hs 47,301 2.8 2.1 0.0 33.3
Numbe o policies in emaining s a es 47,301 1.7 0.7 0.0 3.2
Numbe o employees in KU 47,301 3076.0 8847.0 0.0 255,368.0

116
Eu opean Jou nal o Law and Economics (2025) 59:101–132
4 The ole o policy make s
In his sec ion, we model he en o cemen beha io o poli icians, i.e. he ci cum-
s ances unde which hey decide o in oduce one o he 21 p o ec ion measu es.
Du ing he COVID pandemic, policy make s had o decide on he kind and whe he
o no o in oduce p o ec i e policies. As a gauge, he in ec ion a e, which we label
i in he ollowing, was ag eed on. In Ge many’s ede al sys em, each s a e had some
scope in he decision abou he ‘ igh ’ iming o a u he measu e. Fo e unne s a e
expec ed o be among he i s o in oduce a equi ed policy measu e, compa ed
o la ecome s. Howe e , calenda ime is no he igh indica o o de e mine who
among s a e leade s can be conside ed a o e unne o a la ecome . The me e p o-
g ession o ime is inconsequen ial in he implemen a ion o a policy in e en ion
in he absence o a change in in ec ion a e: no in ec ion, no COVID policy, ega d-
less o he ime ha passes. A leade can only be conside ed a o e unne i he o
she decides o in oduce a p o ec i e policy mo e decisi ely han o he s a a gi en
le el o in ec ion, ha is, in a speci ic pandemic si ua ion. In o he wo ds, leade ship
is demons a ed by he de e mina ion o make an unpopula bu necessa y decision
a he igh (in ec ion) ime. The decision i sel depends on many de e minan s, o
which we ha e o con ol as a gued in he heo e ical pa abo e, he decision o
implemen a p o ec i e policy a a gi en in ec ion a e i has o be conside ed as a
andom a iable. The e o e, we conside he p opensi y o adop he policy m o he
leade l as a andom a iable
𝜆(i)
while con olling o con ex ual, egional ac o s.
Fo his eason, and in con as o adi ional su i al s udies, we do no use calenda
ime as he du a ion pa ame e , bu he ac ual in ec ion a e i.7
4.1 P opensi y oimplemen COVID policies
As we look o he p obabili y ha a poli ical leade in oduces a p o ec i e meas-
u e gi en in ec ion a e i and con ex ual ac o s, i is s a is ically an amoun o use
su i al analysis. The cumula i e densi y o a policy measu e ha has no ye been
implemen ed a in ec ion a e i is
P(I>i)=1−F(i)=S(i)
. Hence, he p obabili y
o s a e leade l o in oduce a COVID-policy m is8:
wi h
m
l(i)
as he p obabili y densi y unc ion o leade l wi h espec o measu e
m. Equa ion (1) is he e o e equi alen o he haza d unc ion. I is he ins an ane-
ous a e o in oducing a policy a su i al- ime i. In simple e ms, he denomina-
o in Eq.1) s a es he p obabili y ha a ce ain p o ec i e measu e m has no ye
been in oduced a he gi en in ec ion a e i; he nume a o s a es he p obabili y
(1)
𝜆
m
l(i)= lim
di→∞
P(i≤I<i+di|I≥i)
di =
m
l
(i)
1−Fm
l
(i
)
7 A u he ad an age o his p ocedu e is ha all ime- a ian a iables, which we discuss below, a e
ime-in a ian wi h espec o su i al- ime i, hough hey a e ime- a ian wi h espec o calende
ime .
8 Concep ually, ou model is aligned o he wo k by Aga wal and Go (2002).
117
Eu opean Jou nal o Law and Economics (2025) 59:101–132
ha measu e m will be ( e-)in oduced by leade l ins an ly gi en i. I exp esses he
condi ional expec a ion.
We hypo hesize ha , in addi ion o he in ec ion a e, he p obabili y o poli ical
ac ion depends on he s a e leade s l’s baseline de e mina ion
Dl(0)
, egional calen-
da - ime-in a ian ac o s
Rl
(e.g. indus y s uc u e, popula ion densi y) as well as
calenda - ime- a ian de e minan s, i.e he cu en s a e o he COVID pandemic
Il( )
such as he in ensi y o es ic ions expe ienced so a , he numbe o dea hs,
o he accina ion a e; u he mo e, he con empo aneous economic si ua ion
El( )
( he g ea e he economic damage o p e ious COVID-policies, he g ea e he
eluc ance o policymake s o inc ease es ic ions) plays an impo an ole and las
no leas , unde s ong unce ain y, he ex en o which policymake s imi a e o he
leade s L
−l(
)
due o lack o be e knowledge.
4.2 Leade s’ baseline de e mina ion
Fo eadabili y we lea e ou supe - and subsc ip s m and l, in he ollowing. We
assume leade s o ha e an indi idual and cons an baseline de e mina ion o impose
es ic ions o p o ec i e easons. This is wha makes a leade ei he a o e unne o
a la ecome in en o cemen . I is assumed as a ixed pe sonali y ai ha de e mines
he leade ’s baseline de e mina ion
Dl(0)
o in oduce a es ic i e COVID policy
measu e by e alua ing he ade o be ween ci izens’ eedom and he p o ec ion o
hei heal h. The de e mina ion changes as he pandemic p og esses. The change can
be exp essed as:
whe e he de e mina ion
Dl(i)
depends on:
which is, he leade l’s baseline de e mina ion
Dl(0)
, he de e mina ion
Dl(i−1)
gi en he p e ious in ec ion s a e
i−1
, he cu en pandemic si ua ion
Il( )
, egional
ime-in a ian speci ici ies (
Rl
), he egion’s economic si ua ion (
El( )
, i.e. numbe
o wo ke s in subsidized pa - ime employmen , hence o h: Ku za bei , KU), and
he cu en policy ac ions o emaining s a es
L−l( )
. To allow o beha io al unce -
ain y, we include a dis u bance
𝜀∼N(0, 𝜎)
– as p e iously discussed in he in o-
duc o y sec ion o his pape .
Whe eas he leade s’ baseline de e mina ion
Dl(0)
does no change, he cou se
o he pandemic will change his/he ac ual de e mina ion. When he si ua ion ge s
mo e se e e and he human cos o he pandemic inc eases (dea hs), he p opensi y
o in oduce u he measu es will inc ease (
⇒
𝜕
∕
𝜕I
l(
)
>
0
), hough making he
popula ion su e mo e and mo e.
Regional speci ici ies will also impel leade s o ac acco dingly. Fo ins ance, a
egion’s indus y s uc u e wi h a high sha e in se ices ha in ol es high- equen
human in e ac ion will inc ease he leade ’s p opensi y o pass u he COVID
(2)
ΔDl(i)=Dl(i)−Dl(i−1)
(3)
Dl
(i)=
(
D
l
(0),D
l
(i−1),I
l
( ),R
l
,E
l
( ),L
−l
( ),𝜀(i)
)
118
Eu opean Jou nal o Law and Economics (2025) 59:101–132
policies (
⇒𝜕 ∕𝜕Rl>0
), al hough he sign o he de i a i e is unclea . A high pop-
ula ion densi y equi ing mo e immedia e policy ac ions may make a leade mo e
eluc an o impose es ic ions, as he/she aces subs an ial poli ical (decline in epu-
a ion) as well as socio-economic cos ( educ ion o ci izens’ well-being). When he
pandemic s a s o a ec a egion’s economy, nega i ely, he leade ’s willingness o
inc ease he bu dening p essu e o u he policies will decline (
⇒𝜕 ∕𝜕El( )>0
).
An aspec ha also will in luence a leade ’s decision making p ocess is o imi-
a e one’s neighbo s, ha is, looking a one’s pee s and ac alike (
⇒
𝜕 ∕𝜕L
−l
>
0
)
(=imi a ion).
4.3 Leade s’ policy ac ion unc ion
Assembling all elemen s om abo e, we can now o mula e he co esponding pol-
icy ac ion unc ion (=haza d unc ion):
whe e h(i) is he haza d a e (o de e mina ion) o leade l o in oduce a u he
COVID policy a in ec ion a e i,
𝛿l
cap u es he leade ’s baseline de e mina ion
and deno es he employed haza d unc ion. In ou eg essions below, we use he
Weibull dis ibu ion as haza d unc ion:
The Weibull dis ibu ion allows us o conside di e en shapes o he de elopmen
o he baseline haza d a e. I may dec ease, inc ease, o emain he same du ing he
pandemic. Pa ame e p cap u es he co esponding end.
5 Resul s
To s a wi h, we pe o m a su i al analysis employing a non-pa ame ic cumula-
i e haza d unc ion o show s a es’ de e mina ion o in oduce a es ic i e COVID
policy wi hou conside ing any co a ia es (con ex ual a iables). A e wa ds, we un
a pa ame ic su i al model while including s a es’ con ex . I is c ucial o highligh
ha he discussion he e ocuses on he ole o law en o cemen in implemen ing
s ingen p o ec i e measu es a he sub- egional le el, pa icula ly wi hin he 400
NUTS-3 egions. In o he wo ds, we de i e s a e leade s’ de e mina ion by agg ega -
ing decisions made a he sub- egional le el.
5.1 Non‑pa ame ic cumula i e haza d unc ion
The cumula i e haza d unc ion by s a es is g aphed in Fig.4. The solid black line
indica es he baseline haza d o in oduce a COVID policy ac oss ede al s a es. The
(4)
h(i|x)= (𝛿l+𝛽R
R
+𝛽I
I
+𝛽E
E
+𝛽L
L
−l)
(5)
h0(i|x)=p
⋅
i
p
−
1
exp(
𝛿
l+
𝛽
R
R
+
𝛽
I
I
+
𝛽
E
E
+
𝛽
L
L
−l)
119
Eu opean Jou nal o Law and Economics (2025) 59:101–132
bands shaded in ligh g ay ma k he in e -qua ile ange. Upon ini ial examina ion, i
appea s ha he ou mos cau ious s a es, which may be conside ed o e unne s, a e
BY, NW, BW, and RP. The ou leas p uden s a es; HH, BE, BR, and SC, al hough
SC e u ns o he in e -qua ile ange a highe in ec ion a es.
Since we do no conside any con ex ual a iables in he igu e, he anking o
o e unne s and la ecome s is no e y di e en ia ed. S a es a e no homogeneous, as
a gued in he in oduc ion. Fo ins ance, hey di e in indus y s uc u e and popula-
ion densi y, bo h de e minan s ha may lead policy make s o di e en conclusions
ega ding COVID policies. The ac ual COVID si ua ion i sel will call o speci ic
policies. Fu he mo e, a densely popula ed egion wi h a high sha e o accina ed
people allows o less es ic i e policies han in o he egions ha lag behind in
immuniza ion co e age; in egions wi h an economically di icul si ua ion o a high
popula ion densi y, poli icians will likely hesi a e o in oduce a es ic i e and hus
addi ionally es ic i e policy measu e; he unce ain y ha is in ol ed in a hi he o
unknown pandemic will exace ba e a s ingen pu pose ul policy; he e o e, policy
Fig. 4 Cumula i e Haza d by S a es (CHF). No e: The solid line in he middle indica es he a e age base-
line haza d (
h0(i)
) as cumula i e haza d unc ion dependen on su i al i, i.e. he in ec ion a e; g ay
bands indica e he in e qua ile ange
120
Eu opean Jou nal o Law and Economics (2025) 59:101–132
make s will also look a neighbo s how hey cope wi h he pandemic (imi a ion). We
will add ess all hese de e minan s in he ollowing su i al model.
5.2 Su i al eg ession analysis
In Table5 we s a wi h a semi-pa ame ic p opo iona e haza d model (Cox, 1972).
The Cox model (Model 1) is p esen ed in he i s column o his able. As co a ia es
we included se e al g oups o a iables.
Fi s , we look a he egional cha ac e is ics. An indus y s uc u e wi h a high
sha e in nea -pe son se ices o ces policy make s o pass COVID policies and so
does a high popula ion densi y because i acili a es he sp eading o he i us. A
highe popula ion densi y, howe e , con e sely inc eases he eluc ance o in oduce
u he es ic ing policies – as a gued in he in oduc ion.
Second, he cu en COVID si ua ion in he egion, such as he accina ion a e,
he numbe o dea hs om which a egion su e s will in luence he decision making
p ocess o poli icians. The mo e dead people, he highe leade s’ de e mina ion o
in oduce u he measu es. The mo e ad anced he accina ion a e, he ewe new
s ic measu es will be in oduced.
Thi d, he in ol ed unce ain y in policy making shows up in he deg ee o imi a-
ion. Leade s o a gi en egion will in oduce COVID measu es, i he emaining
ones ha e done so. No e ha he imely associa ion is no on calenda ime bu on
su i al ime i. In o he wo ds, i o he egions in oduced ce ain measu es a a
gi en in ec ion a e i, so will he espec i e egion unde conside a ion. As he posi-
i e coe icien sugges s, he e is imi a ion (spa ial co ela ion) among egions o a
signi ican ex en .
Fou h, we include a nai e pa isan de ini ion o s a e go e nmen s (Magge i &
T ein, 2021; Russo & Ve zichelli, 2016): le , cen e , and igh , whe e he base ca -
ego y o his ca ego ical a iable ‘Poli ical s ance’ is ‘cen e ’ go e nmen . The h ee
ca ego ies a e based on whe he he go e ning coali ion du ing he pandemic can be
conside ed le , cen e , o igh . We also ook in o accoun changes in go e nmen
du ing he pandemic.
Fi h, a di icul economic si ua ion makes poli icians mo e eluc an o impose
es ic i e policies. This is wha he nega i e coe icien o he a iable labeled
‘Numbe employees in KU’ sugges s. An inc easing ‘Ku za bei ’, i.e pa ial employ-
men subsidized by he s a e, will educe policy make s’ willingness o u he
weaken he economy wi h COVID policy measu es.
In Model (1), we apply he Cox model, a semi-pa ame ic model assuming a p o-
po ional haza d unc ion. In Model (2), we employ an exponen ial haza d unc ion
as a i s obus ness es o ou esul s. The signs do no change and he coe icien s
emain s able, excep o he i s accina ion coe icien ha becomes nega i e and
signi ican . As in Model (1), we in oduce s a e dummies, which a e ixed e ec s
ha we will u ilize subsequen ly o de i e he anking o s a es – ou main objec i e
in his pape . Addi ionally, we in oduce dummies o he ype o he 23 measu es.
The d awback o he p e ious models is ha hey assume a cons an baseline haz-
a d which is a he unlikely in he case o he sp eading o a i us. An exponen ial

121
Eu opean Jou nal o Law and Economics (2025) 59:101–132
Table 5 Su i al eg ession models
S anda d e o s in pa en heses, ***
p<0.01
, **
p<0.05
, *
p<0.1
. No e: Each column p esen s eg es-
sion coe icien s: (1) a semi-pa ame ic Cox eg ession, pa ame ic su i al eg essions assuming (2) an
Exponen ial, (3) and (4) a Weibull p obabili y dis ibu ion, espec i ely. Model (4) includes an indica o
a iable o ins i u ional change (IC), when he ede al s a e es ic ed he scope o each s a e. We do no
epo he cons an e m in eg essions
Time a iable: in ec ion a e
Cox Exponen ial Weibull Weibull (IC)
(1) (2) (3) (4)
Regional cha ac e is ics
Indus y s uc u e 0.047
∗∗∗
0.059
∗∗∗
0.050
∗∗∗
0.049
∗∗∗
(0.002) (0.002) (0.002) (0.002)
Popula ion densi y
−
0.218
∗∗∗
−
0.298
∗∗∗
−
0.245
∗∗∗
−
0.231
∗∗∗
(0.008) (0.011) (0.009) (0.009)
Cu en COVID pe o mance
Numbe o 1s accina ion (= 1s ac)
−
0.001
−
0.012
∗∗∗
−
0.007
∗∗∗
−
0.048
∗∗∗
(0.002) (0.003) (0.003) (0.003)
Numbe p e ious policies
−
0.068
∗∗∗
−
0.078
∗∗∗
−
0.070
∗∗∗
−
0.082
∗∗∗
(0.002) (0.003) (0.002) (0.003)
Dea hs 26.059
∗∗∗
30.498
∗∗∗
27.996
∗∗∗
36.390
∗∗∗
(2.278) (4.249) (3.610) (3.868)
Policy imi a ion
Numbe policies emaining s a es
−
0.287
∗∗∗
−
0.424
∗∗∗
−
0.319
∗∗∗
−
0.600
∗∗∗
(0.012) (0.017) (0.014) (0.016)
Policy s ance
Le -wing
−
0.382
∗∗
−
0.514
∗∗∗
−
0.446
∗∗∗
−
0.810
∗∗∗
(0.167) (0.165) (0.137) (0.136)
Righ -wing 1.118
∗∗∗
1.784
∗∗∗
1.465
∗∗∗
1.196
∗∗∗
(0.190) (0.222) (0.184) (0.194)
Economic si ua ion
Numbe employees in KU
−
0.008
∗∗∗
−
0.011
∗∗∗
−
0.009
∗∗∗
−
0.012
∗∗∗
(0.001) (0.001) (0.001) (0.001)
Ins i u ional change
Cen ally managed = 1 0.788
∗∗∗
(0.018)
Disp opo iona e haza d
ln(p)
−
0.177
∗∗∗
−
0.159
∗∗∗
(0.004) (0.004)
S a e dummies Yes Yes Yes Yes
Policy dummies Yes Yes Yes Yes
Obse a ions 42,209 42,209 42,209 42,209
LL -401,897 -73,921 -72,663 -71,683
122
Eu opean Jou nal o Law and Economics (2025) 59:101–132
sp eading, as is he case wi h COVID will also a ec policy make s sensi i i y o
ac and igh he pandemic in o de o keep social and economic cos low. Using a
Weibull haza d unc ion will allow es ima ing a disp opo iona e haza d a e. Model
(3) epo s he co esponding esul s wi h he Weibull haza d unc ion. Again, he
es ima ed coe icien s emain obus compa ed o models (1) and (2). Pa ame e p
indica es he baseline haza d wi h
p=exp [ln(p)]=.83 <1
. The highe he in ec-
ion a e once eached, he lowe he ma ginal endency o ake addi ional ac ions
agains COVID, ce e is pa ibus. Ba el e al. (2020) name a ious ac o s ha explain
he dec easing baseline haza d a e when he in ec ion a e i inc eases: policy meas-
u es a e only e ec i e i he public complies wi h hem. As in ec ion a es ise,
people may become desensi ized o he se e i y o he si ua ion. Con inued expo-
su e o high in ec ion a es and he ongoing na u e o he pandemic may lead o
a no maliza ion e ec , whe e high numbe s become pa o he “new no mal”. As
a esul , indi iduals and communi ies may pe cei e he si ua ion as less ala ming,
educing he pe cei ed u gency o ake u he ac ion. This public pe cep ion in u n
has poli icians become mo e eluc an o ake u he ac ion.9 Mo eo e , he e a e
se e al in e ac ions be ween a iables ha we do no conside because hey a e no
ou main ocus, ha is, o de i e a anking o s a e leade s aking in o accoun con-
ex ual ac o s.10
In Model (4), we include an indica o a iable o he ime a e when he COVID
policy was cen alized end o Ap il 2021 by he Fede al In ec ion P o ec ion Ac
(In SchG). In con as o Cepaluni e al. (2022), who claim ha in he sho un a
high s a e capaci y is be e sui ed o ackle a pandemic, he Ge man case ollows
he Canadian example— he la e desc ibed in Migone (2020): a i s , local ac o s
domina e, and g adually he ede al s a e akes o e . This elimina ed la gely compe-
i ion be ween s a es. As he es ima e sugges s, cen aliza ion signi ican ly inc eased
he p opensi y o en o ce es ic i e measu es.
O e all, he esul s seem o be obus ac oss models. The coe icien s ha e he
expec ed signs: Regional cha ac e is ics inc ease he likelihood o in oducing
ano he measu e as he deg ee o social con ac implied by he indus y s uc u e
inc eases, and dec ease he likelihood a high popula ion densi ies as social accep -
abili y dec eases; he highe he accina ion a e, he less likely u he p o ec i e
measu es a e; he mo e policy measu es al eady ins alled (numbe o p e ious poli-
cies), he lowe he likelihood o u he ac ion; he mo e people die, he mo e likely
a u he p o ec i e measu e. In e ms o policy imi a ion, s a es seem o be less
s ic han hei neighbo s, a i s sigh . I is plausible ha policy imi a ion in e ac s
wi h he policy s ance o he espec i e egion’s go e nmen . Righ -wing (le -wing)
go e nmen s ac mo e (less) s ic ly han cen is go e nmen s11; a ense economic
si ua ion (Numbe o employees in KU) educes he likelihood o ano he measu e
9 See Ba el e al. (2020) o Pe he ick e al. (2021) o u he explana ions.
10 To do so, i is no necessa y o del e in o he in e dependencies o con ex ual ac o s.
11 Adding up he coe icien s o ‘Policy Imi a ion’ and ‘Policy S ance’ weigh ed by NUTS-3 sub egions
leads o an o e all posi i e policy imi a ion e ec . Since we do no in end o s udy he g anula in e de-
pendencies o poli ical beha io , bu a he he anking o s a e leade s du ing he COVID pandemic, we
do no conside u he mode a ion e ec s, he e. This will be le o a ollow-up s udy.
123
Eu opean Jou nal o Law and Economics (2025) 59:101–132
being in oduced. The cen aliza ion o powe by he ede al go e nmen signi i-
can ly inc eased he likelihood o policy measu es being implemen ed.
I should be emphasized he e ha we a e no in e es ed in he speci ici ies o poli-
cymake s’ poli ical in e dependencies in his pape , bu only wan o con ol o con-
ex ual ac o s in o de o dis inguish o e unne s om la ecome s in policy ac ion.
We now ocus on he s a e ixed e ec s, i.e., s a e dummies, which indica e he s a e-
speci ic baseline haza d o en o cing policies ela ed o he COVID pandemic. The
espec i e coe icien s allow us o ank s a es acco ding o hei baseline haza d, i.e.
he ex en o which s a es a e de e mined o s ic ly en o ce p o ec i e measu es.
Toge he wi h he ankings iden i ied wi h ega d o media announcemen and
law codi ica ion, we can now con on hese ankings wi h he ela i e pe o mance
o s a es in law en o cemen – while conside ing he he e ogenei y o s a es. Figu e5
p esen s he pe o mance o coun ies in he h ee compe i i e ields: Announce s,
Codi ie s, and En o ce s. The le e ical axis indica es he anking based on he
sea ch in (p in ) media (LexisNexis da abase), he middle axis e e s o he sea ch
in he Ju is da abase, and he igh axis o he anking in law en o cemen acco ding
o he Des a is da a, con olling o s a e he e ogenei y. The la e is based on he
s a e ixed e ec , which can be in e p e ed as he leade s’ baseline de e mina ion
o en o ce es ic i e measu es. Co espondingly, he o e unne s in announcing a e
BW, BE, SR, and BY; la ecome s a e MV, SC, BR, and NI. The i s ou o e un-
ne s in codi ying COVID-measu es a e HE, BE, BW, and MV; he la ecome s a e
BB, BR, HH, and SR. Wi h ega d o en o ce s, he o e unne s a e RP, BY, SH, and
NI, he la ecome s HH, BE, SC, BR. Figu e5 illus a es he change in ankings. MV,
Fig. 5 Rankings: Announce s, Codi ie s, and En o ce s. No e: The h ee axes ep esen he ankings o
he 16 s a es conce ning he announcemen o COVID-policy measu es, he codi ica ion o hese meas-
u es, and hei en o cemen . Dashed (solid black) lines indica e coun ies wi h dec easing (inc easing)
ankings in e ms o he h ee ankings announcemen , codi ica ion and en o cemen
124
Eu opean Jou nal o Law and Economics (2025) 59:101–132
Table 6 Fo e unne s s. La ecome s
The i s hal o he able indica es o e unne s and la ecome s acco ding o he anking hey achie ed in he h ee compe i i e ields: announce s (a), codi ie s (c), and
en o ce s (e). In he second hal o he able, he o e unne s and la ecome s a e lis ed in o de o g ea es di e ence in he pai wise compa ison o each s a e’s anking.
Compa ing anks such as
a>c
means ha he ank in a (announcing) is highe han he ank in c (codi ying). Also, see Fig.5)
C i e ion Fo e unne s La ecome s
Announce s (a) BW BE SR BY NI BR SC MV
Codi ie s (c) HE BE BW MV SR HH BR BB
En o ce s (e) RP BY SH NI BR SC BE HH
Beha io al ypes Fo e unne s La ecome s
p oc as ina o s
(a > c)
SR BY BB HH SC TH HE MV
A mchai ac i -
is s (c > e)
BE HE NW BW NI SR BB BY
Slien policy-
make s (e > a)
MV NI RP SH BW HH NW BE
131
Eu opean Jou nal o Law and Economics (2025) 59:101–132
di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/
licenses/by/4.0/.
Re e ences
Aga wal, R., & Go , M. (2002). Fi m and p oduc li e cycles and i m su i al. Ame ican Economic
Re iew, 92(2), 184–190.
Ba be á, P., Casas, A., Nagle , J., Egan, P. J., Bonneau, R., Jos , J. T., & Tucke , J. A. (2019). Who leads?
who ollows? measu ing issue a en ion and agenda se ing by legisla o s and he mass public using
social media da a. Ame ican Poli ical Science Re iew, 113(4), 883–901.
Ba el, J. J. V., Baicke , K., Boggio, P. S., Cap a o, V., Cichocka, A., Gel and, M., Sanchez, R. G., Hes-
len ield, S., Kel ne , D., Han, S., Ken ick, D. F., an de Linden, S., Lone gan, E. A., Pu yea , C.,
Ra hje, S. G., Reine o, M., Pei z, L. T., Wille , R., Jo dan, J. J., … Mo is, M. W. (2020). Using
social and beha iou al science o suppo Co id-19 pandemic esponse. Na u e Human Beha iou ,
4(5), 460–471.
Besley, T. J. and Case, A. (1992). Incumben beha io : Vo e seeking, ax se ing and ya ds ick
compe i ion.
Bjø nsko , C., & Voig , S. (2022). This ime is di e en ?—On he use o eme gency measu es du ing he
Co ona pandemic. Eu opean Jou nal o Law and Economics, 54(1), 63–81.
Bol on, P., & Fa ell, J. (1990). Decen aliza ion, duplica ion, and delay. Jou nal o Poli ical Economy,
98(4), 803–826.
B oschek, J. (2022). Fede alism, poli ical leade ship and he Co id-19 pandemic: Explaining Canada’s
ale o wo ede a ions. Te i o y, Poli ics, Go e nance, 10(6), 779–798.
Ca ozzi, F., P o enzano, S., & Ro h, S. (2024). U ban densi y and Co id-19: Unde s anding he US expe-
ience. The Annals o egional science, 72(1), 163–194.
Cepaluni, G., Do sch, M. T., & B anyiczki, R. (2022). Poli ical egimes and dea hs in he ea ly s ages o
he Co id-19 pandemic. Jou nal o public inance and public choice, 37(1), 27–53.
Ce que i, R., Coppie , R., Gi a di, A., & Ven u a, M. (2022). The soone he be e : Li es sa ed by he
lockdown du ing he Co id-19 ou b eak. The case o I aly. The Econome ics Jou nal, 25(1), 46–70.
Congle on, R. (2023). Fede alism and pandemic policies: Va ie y as he spice o li e. Public Choice,
195(1), 73–100.
Cox, D. R. (1972). Reg ession models and li e- ables. Jou nal o he Royal S a is ical Socie y: Se ies B
(Me hodological), 34(2), 187–202.
Fä be , G. (2021). Ge many’s igh agains Co id-19: The ension be ween cen al egula ion and decen-
alised managemen . In Compa a i e Fede alism and Co id-19, pages 51–69. Rou ledge.
Gaenssle, S., & Budzinski, O. (2021). S a s in social media: New ligh h ough old windows? Jou nal o
Media Business S udies, 18(2), 79–105.
Gaenssle, S. and Budzinski, O. (2024). 24 The economics o in luence s and social media s a dom. Wal-
e de G uy e GmbH & Co KG.
Ga za elli, G., & Kee on, L. (2018). Labo a o y ede alism and in e go e nmen al g an s. Jou nal o
Ins i u ional Economics, 14(5), 949–974.
Ga za elli, G., Kee on, L., & Si oe, A. A. (2022). Righ s edis ibu ion and Co id-19 lockdown policy.
Eu opean Jou nal o Law and Economics, 54(1), 5–36.
G ee , S. L., Rozenblum, S., Falkenbach, M., Löblo á, O., Ja man, H., Williams, N., & Wisma , M.
(2022). Cen alizing and decen alizing go e nance in he Co id-19 pandemic: The poli ics o c edi
and blame. Heal h Policy, 126(5), 408–417.
Ha ison, J. L. (1978). Ya ds ick compe i ion: A p ema u ely disca ded o m o egula o y elie . Tul. L.
Re ., 53, 465.
Hegele, Y., & Schnabel, J. (2021). Fede alism and he managemen o he co id-19 c isis: Cen alisa ion,
decen alisa ion and (non-) coo dina ion. Wes Eu opean Poli ics, 44(5–6), 1052–1076.
Jones, B.D. and Baumga ne , F.R. (2005). The poli ics o a en ion: How go e nmen p io i izes p ob-
lems. Uni e si y o Chicago P ess.
Ke be , W., & Budzinski, O. (2003). Towa ds a di e en ia ed analysis o compe i ion o compe i ion
laws. Zei sch i ü We bewe bs ech , 1(4), 411–448.

132
Eu opean Jou nal o Law and Economics (2025) 59:101–132
Kollman, K., Mille , J. H., & Page, S. E. (2000). Decen aliza ion and he sea ch o policy solu ions.
Jou nal o Law, Economics, and O ganiza ion, 16(1), 102–128.
Magge i, M., & T ein, P. (2021). Mo e is less: Pa isan ideology, changes o go e nmen , and policy
in eg a ion e o ms in he UK. Policy and Socie y, 40(1), 79–98.
Migone, A. R. (2020). T us , bu cus omize: Fede alism’s impac on he Canadian Co id-19 esponse.
Policy and Socie y, 39(3), 382–402.
Mis u , E. M., Gi ens, J. W., & Ma iso , D. C. (2023). Con agious Co id-19 policies: Policy di usion
du ing imes o c isis. Re iew o Policy Resea ch, 40(1), 36–62.
Pe he ick, A., Goldszmid , R., And ade, E. B., Fu s , R., Hale, T., Po , A., & Wood, A. (2021). A wo ld-
wide assessmen o changes in adhe ence o Co id-19 p o ec i e beha iou s and hypo hesized pan-
demic a igue. Na u e Human Beha iou , 5(9), 1145–1160.
Rugge i, K., S ock, F., Haslam, S. A., Cap a o, V., Boggio, P., Elleme s, N., Cichocka, A., Douglas, K.
M., Rand, D. G., Van de Linden, S., e al. (2024). A syn hesis o e idence o policy om beha -
iou al science du ing Co id-19. Na u e, 625(7993), 134–147.
Russo, F., & Ve zichelli, L. (2016). Go e nmen ideology and pa y p io i ies: he de e minan s o pub-
lic spending changes in I aly. I alian Poli ical Science Re iew/Ri is a I aliana di Scienza Poli ica,
46(3), 269–290.
Samad, A., Al Je jawi, K., & Dadich, A. (2022). C isis leade ship: Poli ical leade ship du ing he Co id-
19 pandemic. Sus ainabili y, 15(1), 266.
Sega o, C. I., dos San os, F. B. P., Bichi , R. M., & Mo andi, E. L. (2022). Inequali ies and he Co id-19
pandemic in B azil: Analyzing un-coo dina ed esponses in social assis ance and educa ion. Policy
and Socie y, 41(2), 306–320.
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps
and ins i u ional a ilia ions.