Bei ne, John; Pa k, Donghyun; Saadaoui, Jamel; Uddin, Mohammed Gazi Salah
Wo king Pape
Impac o clima e isk on iscal space: Do poli ical s abili y
and inancial de elopmen ma e ?
ADB Economics Wo king Pape Se ies, No. 748
P o ided in Coope a ion wi h:
Asian De elopmen Bank (ADB), Manila
Sugges ed Ci a ion: Bei ne, John; Pa k, Donghyun; Saadaoui, Jamel; Uddin, Mohammed Gazi Salah
(2024) : Impac o clima e isk on iscal space: Do poli ical s abili y and inancial de elopmen
ma e ?, ADB Economics Wo king Pape Se ies, No. 748, Asian De elopmen Bank (ADB), Manila,
h ps://doi.o g/10.22617/WPS240467-2
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/310382
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 ps://c ea i ecommons.o g/licenses/by/3.0/igo/
ASIAN DEVELOPMENT BANK
ASIAN DEVELOPMENT BANK
6 ADB A enue, Mandaluyong Ci y
1550 Me o Manila, Philippines
www.adb.o g
ADB ECONOMICS
WORKING PAPER SERIES
NO. 748
Oc obe 2024
Impac o Clima e Risk on Fiscal Space
Do Poli ical S abili y and Financial De elopmen Ma e ?
This pape explo es how clima e isks ad e sely a ec iscal space using panel local p ojec ions co e ing
199 economies spanning om 1990 o 2022. The indings highligh he impac on economies mos
ulne able o clima e change. The esul s sugges ha ac o s such as poli ical s abili y and inancial
de elopmen ha e he po en ial o alle ia e hese e ec s. I e eals ha he in luence o clima e isk on iscal
capaci y is mo e signi ican in si ua ions o limi ed iscal space. Implemen ing iscal consolida ion eme ges
as a c ucial ac o in mi iga ing he nega i e impac o clima e isks on iscal capaci y, wi h poli ical s abili y
and inancial de elopmen also playing pi o al oles.
Abou he Asian De elopmen Bank
ADB is commi ed o achie ing a p ospe ous, inclusi e, esilien , and sus ainable Asia and he Paci ic,
while sus aining i s e o s o e adica e ex eme po e y. Es ablished in 1966, i is owned by 69 membe s
—49 om he egion. I s main ins umen s o helping i s de eloping membe coun ies a e policy dialogue,
loans, equi y in es men s, gua an ees, g an s, and echnical assis ance.
IMPACT OF CLIMATE RISK ON
FISCAL SPACE
DO POLITICAL STABILITY AND FINANCIAL
DEVELOPMENT MATTER?
John Bei ne, Donghyun Pa k, Jamel Saadaoui, and Gazi Salah Uddin
ASIAN DEVELOPMENT BANK
The ADB Economics Wo king Pape Se ies
p esen s esea ch in p og ess o elici commen s
and encou age deba e on de elopmen issues
in Asia and he Paci ic. The iews exp essed
a e hose o he au ho s and do no necessa ily
e lec he iews and policies o ADB o
i s Boa d o Go e no s o he go e nmen s
hey ep esen .
ADB Economics Wo king Pape Se ies
John Bei ne, Donghyun Pa k, Jamel Saadaoui,
and Gazi Salah Uddin
No. 748 | Oc obe 2024
John Bei ne (jbei [email protected] g) is a p incipal economis
and Donghyun Pa k (dpa [email protected] g) is an economic
ad iso a he Economic Resea ch and De elopmen
Impac Depa men , Asian De elopmen Bank. Jamel
Saadaoui (saadaoui@unis a. ) is a ull p o esso o
economics a he Uni e si é Pa is 8. Gazi Salah Uddin
([email p o ec ed]) is an associa e p o esso a
Linköping Uni e si y.
Impac o Clima e Risk on Fiscal Space:
Do Poli ical S abili y and Financial De elopmen Ma e ?
C ea i e Commons A ibu ion 3.0 IGO license (CC BY 3.0 IGO)
© 2024 Asian De elopmen Bank
6 ADB A enue, Mandaluyong Ci y, 1550 Me o Manila, Philippines
Tel +63 2 8632 4444; Fax +63 2 8636 2444
www.adb.o g
Some igh s ese ed. Published in 2024.
ISSN 2313-6537 (p in ), 2313-6545 (PDF)
Publica ion S ock No. WPS240467-2
DOI: h p://dx.doi.o g/10.22617/WPS240467-2
The iews exp essed in his publica ion a e hose o he au ho s and do no necessa ily e lec he iews and policies
o he Asian De elopmen Bank (ADB) o i s Boa d o Go e no s o he go e nmen s hey ep esen .
ADB does no gua an ee he accu acy o he da a included in his publica ion and accep s no esponsibili y o any
consequence o hei use. The men ion o speci ic companies o p oduc s o manu ac u e s does no imply ha hey
a e endo sed o ecommended by ADB in p e e ence o o he s o a simila na u e ha a e no men ioned.
By making any designa ion o o e e ence o a pa icula e i o y o geog aphic a ea in his documen , ADB does no
in end o make any judgmen s as o he legal o o he s a us o any e i o y o a ea.
This publica ion is a ailable unde he C ea i e Commons A ibu ion 3.0 IGO license (CC BY 3.0 IGO)
h ps://c ea i ecommons.o g/licenses/by/3.0/igo/. By using he con en o his publica ion, you ag ee o be bound
by he e ms o his license. Fo a ibu ion, ansla ions, adap a ions, and pe missions, please ead he p o isions
and e ms o use a h ps://www.adb.o g/ e ms-use#openaccess.
This CC license does no apply o non-ADB copy igh ma e ials in his publica ion. I he ma e ial is a ibu ed
oano he sou ce, please con ac he copy igh owne o publishe o ha sou ce o pe mission o ep oduce i .
ADB canno be held liable o any claims ha a ise as a esul o you use o he ma e ial.
Please con ac pubsma ke [email protected] g i you ha e ques ions o commen s wi h espec o con en , o i you wish
oob ain copy igh pe mission o you in ended use ha does no all wi hin hese e ms, o o pe mission o use
heADB logo.
Co igenda o ADB publica ions may be ound a h p://www.adb.o g/publica ions/co igenda.
ABSTRACT
We analyze he ela ionship be ween clima e isk and iscal space in a sys ema ic and
igo ous way. To do so, we use panel local p ojec ions o examine he ole o poli ical
s abili y and inancial de elopmen in he ela ionship. Fo a sample o 199 economies in
1990–2022, we i s empi ically con i m ha clima e isks ad e sely a ec iscal space.
We ind ha such e ec s a e mos p onounced o he economies ha a e mos ulne able
o clima e change. Howe e , ou e idence indica es ha poli ical s abili y and inancial
de elopmen can mi iga e such e ec s. We also iden i y nonlinea i ies in he clima e isk–
iscal space nexus. Mo e speci ically, he impac o clima e isk on iscal space is g ea e
when iscal space is mos cons ained—i.e., in he uppe quan ile o he dis ibu ion. While
iscal consolida ion is he key o mi iga ing he ad e se e ec o clima e isks on iscal
space, ou esul s sugges bo h poli ical s abili y and inancial de elopmen can con ibu e
as well.
Keywo ds: clima e isk, ins i u ional quali y, iscal space, bond yields, so e eign a ings
JEL codes: F32, F41, F62
1. In oduc ion
Clima e isks, which e e o he po en ial ad e se socioeconomic impac s o clima e
change, en ail subs an ial iscal isks, especially h ough hei ad e se e ec s on iscal
space. Fo ins ance, a big disas e caused by clima e change is likely o necessi a e la ge
iscal ou lays o elie and eco e y e o s. O ex eme hea esul ing om global wa ming
could cause ex ensi e ag icul u al damage, o cing go e nmen s o p o ide subsidies o
ha d-hi a me s. A a b oade le el, public spending on clima e change adap a ion and
mi iga ion is coming o ep esen one o he bigges sou ces o iscal demand a ound he
wo ld. In combina ion wi h o he la ge looming iscal demands, such as hose ela ed o
popula ion aging, clima e change- ela ed iscal expendi u es pose a majo h ea o iscal
space and iscal sus ainabili y in he u u e.
The p ima y o iginal con ibu ion o ou pape o he li e a u e is ha we empi ically
examine he link be ween clima e isk and iscal space in a sys ema ic and igo ous way.
To do so, we i s in es iga e whe he clima e isk has an ad e se e ec on iscal space
and whe he his e ec depends on he ela i e amoun o iscal space. We hen examine
he ole o poli ical s abili y and inancial de elopmen in mi iga ing such clima e- ela ed
iscal isks.
Poli ical s abili y is likely o mi iga e hese isks since i inc eases he likelihood o mo e
sus ainable iscal policy, o example in he o m o a mo e obus medium- e m iscal
amewo k. As a esul , a mo e s able poli ical en i onmen is likely o educe he impac
o clima e shocks and o he shocks on iscal sus ainabili y. In addi ion, poli ical s abili y is
conduci e o mo e ca e ul, a ional, and cos -e ec i e go e nmen planning in esponse
o po en ial clima e shocks, which will help p ese e iscal space.
Financial de elopmen is also expec ed o mi iga e clima e- ela ed iscal isks. In
pa icula , in a inancially well-de eloped economy, i ms and households will ha e access
o insu ance and o he inancial ins umen s ha p o ec hem om he nega i e e ec s
o clima e shocks. This, in u n, educes he need o la ge iscal ou lays and hus
mi iga es he nega i e e ec on iscal space. Fu he mo e, inancial de elopmen
inc eases he amoun o c edi a ailable o i ms and households o help hem cushion
he impac o po en ial clima e shocks.
An addi ional, indi ec , channel h ough which clima e isk may ad e sely a ec iscal
space is ia i s impac on economic unce ain y. The e is g owing consensus ha clima e
change poses a se ious isk o humani y, a consensus ha is bu essed by eal-wo ld
clima e e en s such as eco d-high empe a u es hi ing he wo ld in 2024. Ne e heless,
he e emains signi ican unce ain y abou he u u e ajec o y o clima e change as well
as i s impac on economic ac i i y. In ui i ely, poli ical s abili y and inancial de elopmen
can mi iga e he ad e se impac o unce ain y on he economy. Poli ical s abili y p e en s
poli ical unce ain y om exace ba ing economic unce ain y and hus enables economic
agen s o espond be e o he la e . Financial de elopmen p o ides economic agen s
wi h a ious inancial ins umen s, o example hedging ins umen s such as u u es
con ac s, o p o ec hemsel es agains economic unce ain y. The e o e, bo h poli ical
2
s abili y and inancial de elopmen can help mi iga e he nega i e economic e ec s o
unce ain y associa ed wi h clima e change.
To empi ically examine he ole o poli ical s abili y and inancial de elopmen on he
clima e– iscal nexus, we pe o m a c oss-coun y analysis. Tha is, we in es iga e
whe he poli ically mo e s able and inancially mo e de eloped economies a e less
ulne able o clima e- ela ed iscal isks. To do so, we use a la ge global sample o 199
economies o he pe iod om 1990 o 2022. Ou empi ical analysis is based on panel
local p ojec ions, and wo measu es o iscal space—namely, so e eign bond yields and
a ings on o eign cu ency long- e m so e eign deb . The wo measu es a e widely used
in he li e a u e, and hey e lec he inancial ma ke s’ assessmen o an economy’s iscal
space.
Ou empi ical analysis in ol es a wo-s age es ima ion. In he i s s age, we in es iga e
he ela ionship be ween clima e ulne abili y and iscal space. We ind ha clima e
ulne abili y ad e sely a ec s iscal space and ha he e ec s a e mos p onounced o
hose economies ha a e mos ulne able o clima e change. The analysis o he i s
s age con i ms he indings o he exis ing li e a u e, al hough ou coun y sample is
subs an ially la ge han ha o o he s udies. In he second s age, we empi ically
in es iga e he ole o poli ical s abili y and inancial de elopmen in mi iga ing he
nega i e spillo e s o clima e exposu e o iscal space. This is he main o iginal
con ibu ion o ou pape o he li e a u e. We also assess whe he he ad e se e ec o
clima e isk on iscal space is nonlinea in he sense ha he e ec depends on iscal
space. Tha is, we analyze whe he he e ec depends on he ex en o which an economy
is iscally cons ained.
Ou e iew o he li e a u e yields only one majo s udy ha is somewha ela ed o ou
pape —namely, You e al. (2014). In his, he au ho s empi ically in es iga e he link
be ween democ acy, inancial openness, and global ca bon dioxide (CO2) emissions.
They also examine whe he he impac o independen a iables on CO2 emissions a ies
h oughou he CO2 emission dis ibu ion. Tha is, hey analyze whe he he impac o
hese a iables on CO2 emissions depends on a coun y’s CO2 emissions le el. Thei
c oss-coun y empi ical analysis co e s a global sample o 98 ad anced and de eloping
economies and spans he pe iod om 1995 o 2005. The au ho s ind ha , among he
economies ha emi he mos , mo e democ acy educes CO2 emissions bu g ea e
inancial openness does no .
Ou pape is undamen ally di e en , no wi hs anding some supe icial simila i ies.
Fi s , ou poli ical a iable is no democ acy bu poli ical s abili y, de ined as absence o
domes ic and ex e nal con lic . Second, we look a clima e isks, which e e o he
po en ial nega i e impac o clima e change, ins ead o CO2 emissions. Clima e isk is a
much b oade concep han CO2 emissions, which ep esen a speci ic componen o he
global en i onmen al c isis. A deepe di e ence is ha , in ou pape , clima e isk is an
independen a he han a dependen a iable. Finally, ou key a iable o in e es is iscal
space a he han clima e isk. Mo e p ecisely, we examine he ex en o which poli ical
s abili y mode a es he ad e se e ec s o clima e isk on iscal space. In ligh o he huge
3
challenge ha clima e change poses o iscal sus ainabili y, ou pape helps us iden i y
he ac o s ha ende he challenge mo e manageable. We also look a nonlinea i ies bu
in e ms o iscal space a he han clima e isk. Speci ically, we es whe he he impac
o iscal space on clima e isk depends on he ex en o which an economy is iscally
cons ained.
O e all, we ind poli ical s abili y educes he ad e se spillo e e ec s o clima e isks
on iscal space. Mo e p ecisely, ou e idence indica es ha clima e isks a e associa ed
wi h lowe -bond- isk-p emium lowe and highe so e eign a ings in economies ha su e
less om bo h ex e nal and in e nal con lic . In addi ion, we ind ha inancial de elopmen
weakens he link be ween clima e isks and iscal space. Financially mo e de eloped
economies do no expe ience a clima e- ela ed bond isk p emium o a pe sis en
de e io a ion o so e eign a ings owing o clima e ulne abili y. Finally, we iden i y
asymme ic e ec s in he sense ha he mos iscally cons ained economies a e subjec
o he la ges clima e- ela ed isk p emia. To sum up, we ind ha inancial de elopmen
and poli ical s abili y can se e as impo an bu e s agains he ad e se iscal impac s o
clima e change.
The es o he pape is s uc u ed as ollows: Sec ion 2 e iews he li e a u e; Sec ion
3 ou lines he da a and me hodology; Sec ion 4 epo s and discusses he empi ical
indings; and Sec ion 5 concludes.
2. Li e a u e Re iew
In his pape , we in es iga e he global challenge o clima e isks and clima e change,
ocusing on i s impac on iscal space. P e ious s udies show ha clima e change is no
jus an en i onmen al issue. I also has a signi ican and ad e se impac on economic
g ow h (Oppenheime e al. 2004; Tol e al. 2004; Mendelson, Dina , and William 2006;
Di enbaugh and Bu ke 2019; Dasgup a, Emme ling, and Shayegh 2023) and
exace ba es inequali y in de eloping economies (Cappelli, Cos an ini, and Consoli 2021;
Dasgup a, Emme ling, and Shayegh 2023). Gi en ha de eloping economies a e he
mos ulne able o clima e change, s a egic esou ce alloca ion is impe a i e o enhance
hei esilience (Paglialunga, Co e i, and Zan ei 2022). This includes implemen ing
adap a ion policies and isk educ ion measu es and expanding access o p ecau iona y
ools and heal h se ices (Paglialunga, Co e i, and Zan ei 2022; Ce ik and To a Jalles
2023). Fu he mo e, edis ibu ion and he in oduc ion o social sa e y ne s mus be
ensu ed in he a ec ed economies (Ce ik and To a Jalles 2023).
Clima e change insu ance unds, in es men s in economic de elopmen , and c oss-
na ional compensa ion o low-la i ude economies mo e p one o clima e change shocks
p omo e economic esilience in de eloping economies agains clima e ulne abili ies
(Mendelson, Dina , and William 2006). To be e adap o and mi iga e he socioeconomic
impac o clima e change and ising empe a u es, economies mus possess a high
adap i e capaci y (Tol e al. 2004), a di e si ied economy (Dissa 2003), poli ical s abili y
(Dell, Jones, and Olken 2012), and s ong ins i u ional leade ship (Pike, Dawley, and
Tomaney 2010). Un o una ely, hese a ibu es a e o en lacking in de eloping
4
economies, owing o inancial cons ain s and ad e se geog aphical condi ions.
Addi ionally, hese economies end o ha e mo e limi ed iscal space, meaning hei
go e nmen s a e less capable o assis ing hose a ec ed by clima e change shocks
(Ce ik and To a Jalles 2023).
One s and o li e a u e examines he asymme ic e ec s o clima e ulne abili y and
esilience on so e eign isk and public inance. Fo example, Bei ne, Renzhi, and Volz
(2021) gauge he impac o clima e ulne abili y and esilience on so e eign bo owing
cos s using a panel da ase om 40 ad anced and eme ging economies. Thei indings
sugges ha clima e ulne abili y signi ican ly in luences so e eign bo owing cos s mo e
so han do esilience ac o s, wi h bond yields inc easing p og essi ely o highly clima e-
ulne able economies. Ce ik and To a Jalles (2022) expand his discou se by analyzing
da a om 98 ad anced and de eloping economies be ween 1995 and 2017. They ind
ha bo h clima e ulne abili y and esilience a ec go e nmen bo owing cos s, wi h he
impac mo e p onounced in de eloping economies owing o hei weake adap i e
capaci ies and highe so e eign isk cos s.
In a simila ein, Boi an and Ma chewka-Ba kowiak (2022) analyze he impac o
a ious clima e change me ics—pe o mance, exposu e o ex eme e en s, ulne abili y,
eadiness, and clima e deb — on go e nmen bo owing cos s in Eu opean Union
economies om 2000 o 2020. They ind ha economies wi h highe clima e ulne abili y
and lowe capaci y o manage clima e disas e s ace highe so e eign isk p emia,
unde sco ing he impo ance o e ec i e clima e disas e managemen in main aining
a o able bo owing condi ions. Zenios (2022) add esses he b oade implica ions o
clima e isks o iscal s abili y, pa icula ly in ad anced economies. He a gues ha clima e
isks o iscal s abili y emain unanswe ed, p o iding e idence o di e gen clima e isks
ac oss ad anced economies. Mo eo e , his indings delinea e he ansmission channels
h ough which clima e change impac s public inance, emphasizing he need o
comp ehensi e isk managemen s a egies.
Ca a ini, Heu el, and Melkadze (2023) del e in o he impac o mac oeconomic
s abili y on clima e isk. In pa icula , hey examine whe he ambi ious clima e policies can
induce mac oeconomic ins abili y and p opose e icien clima e and mac op uden ial
policies o manage hese isks o e he long un and ac oss business cycles. Thei
indings sugges well-designed clima e policies can mi iga e long- e m iscal isks wi hou
des abilizing he economy.
O he s udies ha e highligh ed addi ional ac o s in he ami ica ions o clima e isk. Fo
ins ance, Ce ika and To a Jalles (2023) explo e he social dimension o clima e
ulne abili y by examining he link be ween clima e isk and economic inequali y. Thei
s udy e eals ha a 1% inc ease in clima e change ulne abili y esul s in a 1.5% inc ease
in inequali y, as measu ed by he Gini coe icien . Yang, Capo in, and Jiménez-Ma in
(2024) con ibu e o he discou se by in es iga ing clima e ansi ion isk spillo e s among
six majo inancial ma ke s om 2013 o 2021. They iden i y he Uni ed S a es as he
p ima y ne con ibu o o clima e ansi ion isk, and he People’s Republic o China
(PRC) and Japan as ne isk ecipien s. Thei s udy highligh s ha clima e isk spillo e s
11
In Figu es 3 and 4, we plo he co ela ion be ween ou wo iscal space a iables and
he ulne abili y sco e. We obse e ha his g aphical e idence poin s owa d a posi i e
co ela ion be ween clima e ulne abili y isks and bonds yields and a nega i e co ela ion
be ween clima e ulne abili y isks and so e eign a ings. These obse a ions a e in
alignmen wi h empi ical indings in he ela ed li e a u e (e.g., Bei ne, Renzhi, and Volz
2021; Ce ik and To a Jalles 2022).
Figu e 3: Sca e Plo o Vulne abili y Sco e and Bond Yields
IMF = In e na ional Mone a y Fund, JPM EMBI = J.P. Mo gan Eme ging Ma ke Bond Index, ND-GAIN =
No e Dame Global Adap a ion Ini ia i e.
Sou ce: Au ho s’ calcula ions.
Figu e 4: Sca e Plo o Vulne abili y Sco e and So e eign Ra ings
ND-GAIN = No e Dame Global Adap a ion Ini ia i e.
Sou ce: Au ho s’ calcula ions.
12
3.2. Me hodology
We use he local p ojec ions (LP) app oach (Jo dà 2005; Jo dà and Taylo 2024) o
empi ically examine he e ec s o ulne abili y shocks on wo measu es o iscal space.2
The LP app oach p esen s se e al ad an ages, including enabling (i) he es ima ion o
impulse esponses di ec ly a each ho izon, wi h no c oss-pe iod es ic ions, (ii) a simple
in e ence o impulse esponse coe icien s, (iii) ease o applica ion o nonlinea
amewo ks, and (i ) ease o scaling o panel da a. Rega ding ou esea ch ques ion, all
ea u es o he LP app oach will help us p o ide dynamic e idence on he impac o
ulne abili y shocks on iscal space a iables. We can o mula e he LP app oach as
ollows:
𝑦𝑦𝑖𝑖,𝑡𝑡+ℎ =𝑏𝑏ℎ𝑆𝑆𝑖𝑖,𝑡𝑡+𝛾𝛾ℎ𝑦𝑦𝑖𝑖,𝑡𝑡−1 +𝛼𝛼′𝒛𝒛𝑖𝑖,𝑡𝑡−1 +𝑣𝑣𝑖𝑖,𝑡𝑡+ℎ
IRF(ℎ) = 𝑏𝑏
ˆℎ
(1)
whe e 𝑦𝑦 is he dependen a iable o in e es , ℎ is he ime ho izon, 𝑆𝑆 is he impulse
a iable (a uni shock on he ulne abili y sco e), 𝐳𝐳 is a ec o o con ol a iables, IRF
deno es he impulse esponse unc ion, and 𝑣𝑣 is he e o e m. In ou case, he
dependen a iable will be ei he go e nmen bond yields o so e eign a ings (Kose e
al. 2022). The con ol a iables will be based on Bei ne, Renzhi, and Volz (2021) and
include domes ic ac o s: cu en accoun balance, consume p ice in la ion, go e nmen
deb , and de ici ; and global ac o s: US bond yields, he MSCI wo ld index, and he VIX.
In o de o imp o e iden i ica ion, we also con ol o h ee ypes o c ises (banking,
cu en , and deb ), d awn om Lae en and Valencia (2020).
4. Empi ical Resul s
4.1. Panel Local P ojec ion Reg essions
In Figu e 5, ou baseline case ac oss all economies indica es a s a is ically signi ican
p emium on so e eign bond yields as a esul o clima e isk ulne abili y, e lec ing he
su plus e u n demanded by in es o s o holding ha deb . Fu he , we spli he sample
be ween low and high clima e isk ulne abili y depending on he alue o he ulne abili y
sco e. Fo he less clima e- ulne able economies, a s a is ically signi ican e ec is no
ound. This is in line wi h economic in ui ion—i.e., low le els o clima e exposu e will no
lead o clima e- ela ed p emia on so e eign bonds. Fo economies ha a e highly
exposed o clima e change, he impac on bond yields is signi ican , as expec ed.
In e es ingly, he e ec is b oadly in line wi h ha o he panel as a whole in e ms o
magni ude, sugges ing he economies ha a e highly ulne able o clima e change may
be d i ing he o e all esul s.
2 We implemen he S a a package o Uga e Ruiz (2023) o es ima e he panel LP esul s.
13
In Figu e 6, we pe o m he same baseline analysis o so e eign a ings, ou second
measu e o iscal space. We ind a consis en esul o ha ca ied ou on bond yields,
whe eby a clima e ulne abili y shock will lead o a pe sis en dec ease in he so e eign
a ings o he ull sample and he highly clima e- ulne able economies. Fo he less
ulne able economies, we do no obse e such a pe sis en de e io a ion in so e eign
a ings, as expec ed.
In he ollowing subsec ions, we in es iga e he in luence o ins i u ional a iables and
he impac o inancial de elopmen in ligh o he p elimina y e idence in Table 2. In ac ,
he le el o inancial and ins i u ional is signi ican ly lowe in he g oup o highly ulne able
economies. Thus, we go beyond he ecen li e a u e (e.g., Bei ne, Renzhi, and Volz 2021;
Ce ik and Jalles To a 2022) on he iscal space–clima e nexus in he ollowing by
in es iga ing a ious o m o nonlinea i ies in he clima e- ela ed p emia on so e eign
bond yields and a ings hanks o in e ac ion e ms and quan ile local p ojec ions.
Figu e 5: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Vulne abili y is de ined as below/abo e Q1 o ul100. The shock is a uni -shock on he
ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
14
Figu e 6: Panel Local P ojec ions o Impac o Vulne abili y on So e eign Ra ings
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Vulne abili y is de ined as below/abo e Q1 o ul100. The shock is a uni -shock on he
ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough boo s apping.
Sou ce: Au ho s’ calcula ions.
4.2. In e ac ions wi h Ins i u ional Va iables
In his sec ion, we explo e he ole o ins i u ional ac o s as a bu e o clima e- ela ed
p emia on so e eign bond yields and a ings—i.e., he ex en o which ins i u ional ac o s
can alle ia e he ex en o clima e isk impac s on he iscal space. Ou ins i u ional
a iable o in e es is a measu e o ex e nal con lic ( om he In e na ional Coun y Risk
Guide [ICRG] da abase, see www.p sg oup.com). Ou panel LP impulse esponses in
Figu e 7 e eal s iking esul s.
We ind ha clima e- ela ed isk p emia on so e eign bonds a e signi ican ly g ea e
o economies ha sco e less a o ably in e ms o ex e nal con lic . We hus in e ha
high poli ical isk (a lowe sco e in he ICRG da abase) ampli ies he comp ession on
iscal space owing o clima e ulne abili y by a ac o o a ound 2 a he 1-yea ho izon
compa ed wi h he ull sample esul s o hose wi h ex e nal con lic . A consis en na a i e
is ound in he case o so e eign a ings (Figu e 8).3
3 In Figu es C1 and C2 in Appendix C, we p esen he esul s using in e nal as opposed o ex e nal con lic s.
The esul s a e b oadly consis en , al hough in he case o so e eign a ings he e is a lack o s a is ical
signi icance. This may be ela ed o he measu e used, which is o eign cu ency so e eign a ings.
Vulne abili y – (shock on ul100)
15
4.3. In e ac ions wi h Financial Ins i u ion and Financial Condi ions
In Figu e 9, we use S i ydzenka (2016)’s inancial ins i u ion de elopmen index o
in es iga e he in luence o inancial ins i u ion de elopmen on he impac o ulne abili y
shocks on iscal space. Fo economies wi h ma u e inancial ins i u ions, clima e
ulne abili y shocks do no igge any inc ease in bond yields.
In Figu e 10, we can see ha clima e ulne abili y shocks do no ha e any signi ican
impac on so e eign a ings o economies wi h ele a ed le els o inancial ins i u ion
de elopmen . On he o he hand, clima e ulne abili y shocks p o oke a pe sis en
de e io a ion in so e eign a ings o economies wi h low inancial ins i u ions and o he
ull sample, unde sco ing he impo ance o sound inancial ins i u ions. The mi iga ing
impac o enhanced inancial de elopmen on he clima e– iscal nexus ollows in ui ion,
whe eby he e is g ea e dep h and liquidi y in local inancial ma ke s and insu ance
ma ke s a e be e de eloped.
We can make some obse a ions on his new esul in he li e a u e. Rela i e o he
baseline esul s bond p emia ha we ind in Figu e 5, he uppe limi o he impulse
esponse unc ions o a low le el o inancial ins i u ion de elopmen is simila , sligh ly
abo e 1% a a ho izon o 2 yea s. So an inc ease o 1 in he ulne abili y sco e implies
an inc ease in bond yields 2 yea s la e . Whe eas o economies wi h a highe le el o
inancial ins i u ion de elopmen , ulne abili y shocks do no imply any inc ease in bond
yields. This is an impo an esul .
The de elopmen o sound inancial ins i u ions makes i possible o p ese e iscal
space in he wake o ulne abili y o clima e isk shocks. In a di e en con ex , Aizenman
e al. (2024) ind ha sound inancial ins i u ions bu e exchange a e ins abili y. These
ela ed esul s may be iewed as complemen a y since acqui ing ha d cu ency (i.e., US
dolla , eu o, yen, e c.) o highly ulne able economies is a ques ion o u mos impo ance.
16
Figu e 7: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(Ex e nal Con lic s)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Ex e nal Con lic s is de ined as below/abo e Q2 o ex con . The shock is a uni -shock on
he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
17
Figu e 8: Panel Local P ojec ions o Impac o Vulne abili y on So e eign Ra ings
(Ex e nal Con lic s)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Ex e nal Con lic s is de ined as below/abo e Q2 o ex con . The shock is a uni -shock on
he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
18
Figu e 9: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(Financial Ins i u ions)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Financial Ins i u ions is de ined as below/abo e Q3 o FI. The shock is a uni -shock on he
ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough boo s apping.
Sou ce: Au ho s’ calcula ions.
19
Figu e 10: Panel Local P ojec ions o Impac o Vulne abili y on So e eign
Ra ings (Financial Ins i u ions)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Financial Ins i u ions is de ined as below/abo e Q3 o FI. The shock is a uni -shock on he
ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough boo s apping.
Sou ce: Au ho s’ calcula ions.
4.4. Quan ile Local P ojec ions
In his sec ion, we explo e he e ec o clima e isk ac oss he dis ibu ion o iscal
space. To do so, we use he app oach in Jo dà e al. (2022) o es ima e quan ile panel
local p ojec ions. We i s adjus he da ase o ob ain a balanced panel da ase . Fo he
bond yields, we ha e 32 economies obse ed om 2000 o 2019 wi h comple e
obse a ions. Fo he so e eign a ings, we ha e 71 economies obse ed om 2000 o
2019 wi h comple e obse a ions. We conside he ollowing adi ional panel LP
es ima ion unc ion:
Δ𝑓𝑓𝑓𝑓𝑖𝑖,𝑡𝑡+ℎ = 𝛼𝛼𝑖𝑖,ℎ+𝛽𝛽ℎ𝑉𝑉𝑉𝑉𝑉𝑉𝑖𝑖,𝑡𝑡+𝛿𝛿ℎ𝑥𝑥𝑖𝑖,𝑡𝑡𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 + 𝜖𝜖𝑖𝑖,𝑡𝑡,
= 𝜔𝜔𝑖𝑖,𝑡𝑡𝜃𝜃ℎ,𝜏𝜏, (2)
whe e 𝑓𝑓𝑓𝑓𝑖𝑖,𝑡𝑡 is he iscal space a iable, 𝛼𝛼𝑖𝑖,ℎ is he coun y- ixed e ec s, 𝑉𝑉𝑉𝑉𝑉𝑉𝑖𝑖,𝑡𝑡 he
ulne abili y a iable, and 𝑥𝑥𝑖𝑖,𝑡𝑡 he ec o o con ol a iables. Le 𝜔𝜔𝑖𝑖,𝑡𝑡 collec he shock,
con ol a iables, and ixed e ec s. Quan ile local p ojec ions a e calcula ed based on:
Vulne abili y – (shock on ul100)
20
𝜃𝜃
ˆ𝜏𝜏=a g min
𝜃𝜃𝜏𝜏
∑
𝑇𝑇−ℎ
𝑡𝑡=1 �𝜏𝜏𝜏𝜏�Δ𝑓𝑓𝑓𝑓𝑖𝑖,𝑡𝑡+ℎ ≥ 𝜔𝜔𝑖𝑖,𝑡𝑡𝜃𝜃𝜏𝜏��Δ𝑓𝑓𝑓𝑓𝑖𝑖,𝑡𝑡+ℎ − 𝜔𝜔𝑖𝑖,𝑡𝑡𝜃𝜃𝜏𝜏�
+(1 − 𝜏𝜏)𝜏𝜏�Δ𝑓𝑓𝑓𝑓𝑖𝑖,𝑡𝑡+ℎ <𝜔𝜔𝑖𝑖,𝑡𝑡𝜃𝜃𝜏𝜏��Δ𝑓𝑓𝑓𝑓𝑖𝑖,𝑡𝑡+ℎ − 𝜔𝜔𝑖𝑖,𝑡𝑡𝜃𝜃𝜏𝜏��.
(3)
whe e 𝜏𝜏 (.) deno es he indica o unc ion and 𝜏𝜏 ∈ (0,1) he 𝜏𝜏 h quan ile. The quan ile o
Δℎ𝑓𝑓𝑓𝑓𝑖𝑖𝑡𝑡(𝑝𝑝)+ℎ condi ional on 𝑋𝑋𝑖𝑖𝑡𝑡(𝑝𝑝) is hen ob ained as ollows:
𝑄𝑄�Δℎ𝑓𝑓𝑓𝑓 ∣ 𝑋𝑋𝑖𝑖𝑡𝑡(𝑝𝑝)�=𝑋𝑋𝑖𝑖𝑡𝑡(𝑝𝑝)𝜃𝜃ℎ,𝜏𝜏≡ 𝑞𝑞𝜏𝜏,𝑡𝑡
ℎ. (4)
The esul s a e p esen ed in Figu es 11 and 12, which con i m ha he impac o
ulne abili y isk shocks is mos e iden o highe quan iles o bond yields. A simila
pa e n eme ges o he case o so e eign a ings, he eby implying nonlinea e ec s and
la ge ela i e e ec s when iscal space is mos cons ained—i.e., a he uppe quan iles
in he dis ibu ion o bond yields and he lowe quan iles o so e eign a ings.
These esul s con i m he p elimina y e idence ha we p o ide in Table 2. In ac , abili y
o main ain low yields on go e nmen bonds and, hus, iscal space is ela ed o he le el
o inancial ins i u ion de elopmen . Abili y o access ha d cu ency du ing episodes o
inancial s ess is also key in p ese ing iscal space and emaining a ac i e and c edible
on in e na ional inancial ma ke s. Ou esul s, in Figu e 11, a e in line wi h he impulse
esponse unc ions p esen ed in Figu e 9 o low le el o inancial ins i u ion de elopmen .
In addi ion, o he lowe quan ile o he dis ibu ion o bond yields, he bond p emium
is no obse ed a e ulne abili y shocks a he 2-yea ho izon. Howe e , a s ong change
in bond yields is obse ed a e 3 yea s, indica ing ha policymake s should emain
igilan in he wake o ulne abili y shocks, e en when s a ing om a sound iscal
posi ion.
27
Appendix D. Baseline Impulse Response Func ions wi h an Ex ensi e Se o
Con ols
We add o ou a iables he o iginal se o con ols—namely, capi al accoun openness
index, exchange a e s abili y index, and sha es in o al ade o uel impo s and
expo s, as desc ibed in Sec ion 2.
Figu e D1: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(Ex ensi e Se o Con ols)
No e: The shock is a uni -shock on he ulne abili y a iable. Fixed e ec s a e included, and s anda d
e o s a e ob ained h ough boo s apping. 90% and 95% con idence in e als in da k blue and ligh blue,
espec i ely.
Sou ce: Au ho s’ calcula ions.
28
Figu e D2: Panel Local P ojec ions o Impac o Vulne abili y on So e eign
Ra ings (Ex ensi e Se o Con ols)
No e: The shock is a uni -shock on he ulne abili y a iable. Fixed e ec s a e included, and s anda d
e o s a e ob ained h ough boo s apping. 90% and 95% con idence in e als in da k blue and ligh blue,
espec i ely.
Sou ce: Au ho s’ calcula ions.
29
Appendix E. Impulse Response Func ions wi h Financial Ma ke De elopmen
Indica o
Figu e E1: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(Financial Ma ke s)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Financial Ma ke s is de ined as below/abo e Q3 o FM. The shock is a uni -shock on he
ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
30
Figu e E2: Panel Local P ojec ions o Impac o Vulne abili y on So e eign
Ra ings (Financial Ma ke s)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Financial Ma ke s is de ined as below/abo e Q3 o FM. The shock is a uni -shock on he
ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
31
Appendix F. Nonlinea i ies in Impulse Responses Acco ding o Go e nmen
S abili y
Figu e F1: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(Go e nmen S abili y)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Go e nmen S abili y is de ined as below/abo e Q3 o go s ab. The shock is a uni -shock
on he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
32
Figu e F2: Panel Local P ojec ions o Impac o Vulne abili y on So e eign
Ra ings (Go e nmen S abili y)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Go e nmen S abili y is de ined as below/abo e Q3 o go s ab. The shock is a uni -shock
on he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
33
Appendix G. Nonlinea i ies in Impulse Responses Acco ding o In ol emen o
Mili a y in Poli ics
Figu e G1: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(Mili a y in Poli ics)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Mili a y in Poli ics is de ined as below/abo e Q3 o milpol. The shock is a uni -shock on
he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
34
Figu e G2: Panel Local P ojec ions o Impac o Vulne abili y on So e eign
Ra ings (Mili a y in Poli ics)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high Mili a y in Poli ics is de ined as below/abo e Q3 o milpol. The shock is a uni -shock on
he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
35
Appendix H. Nonlinea i ies in Impulse Responses Acco ding o Risk o E hnic
Tensions
Figu e H1: Panel Local P ojec ions o Impac o Vulne abili y on Bond Yields
(E hnic Tensions)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high E hnic Tensions is de ined as below/abo e Q3 o e hnic ens. The shock is a uni -shock
on he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
36
Figu e H2: Panel Local P ojec ions o Impac o Vulne abili y on So e eign Ra es
(E hnic Tensions)
CI = con idence in e al, IRF = impulse esponse unc ion.
No e: Low/high E hnic Tensions is de ined as below/abo e Q3 o e hnic ens. The shock is a uni -shock
on he ulne abili y a iable. Fixed e ec s a e included, and s anda d e o s a e ob ained h ough
boo s apping.
Sou ce: Au ho s’ calcula ions.
Vulne abili y – (shock on ul100)
