Wang, Hanyi
Wo king Pape
Ca bon leakage h ough supply chain adjus men s
IDOS Discussion Pape , No. 10/2025
P o ided in Coope a ion wi h:
Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS), Bonn
Sugges ed Ci a ion: Wang, Hanyi (2025) : Ca bon leakage h ough supply chain adjus men s, IDOS
Discussion Pape , No. 10/2025, ISBN 978-3-96021-255-3, Ge man Ins i u e o De elopmen and
Sus ainabili y (IDOS), Bonn,
h ps://doi.o g/10.23661/idp10.2025
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/316405
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/4.0/
Ca bon Leakage h ough
Supply Chain Adjus men s
Hanyi Wang
IDOS DISCUSSION PAPER
10/2025
Ca bon leakage h ough supply chain
adjus men s
Hanyi Wang
Bonn 2025
Hanyi Wang is an economis who s udies he cos -e ec i eness and dis ibu ional impac s o clima e policies
on i ms and hei ade implica ions. She also s udies sus ainable in es ing’s impac on i ms and how egula o y
p essu es acili a e g een ansi ions. D Wang ecei ed he PhD in Economics om he Uni e si y o Cali o nia
San Diego in June 2024.
The Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS) is ins i u ionally inanced by he Fede al Minis y
o Economic Coope a ion and De elopmen (BMZ), based on a esolu ion o he Ge man Bundes ag, and he
s a e o No h Rhine-Wes phalia (NRW) as a membe o he Johannes-Rau-Fo schungsgemeinscha (JRF).
Sugges ed ci a ion:
Wang, H. (2025). Ca bon leakage h ough supply chain adjus men s (IDOS Discussion Pape 10/2025). Bonn:
Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS). h ps://doi.o g/10.23661/idp10.2025
Disclaime :
The analyses exp essed in his pape a e hose o he au ho (s) and do no necessa ily e lec he iews o
policies o he Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS).
Excep o he wise no ed, his publica ion is licensed unde C ea i e Commons A ibu ion (CC BY 4.0). You a e
ee o copy, communica e and adap his wo k, as long as you a ibu e he Ge man Ins i u e o De elopmen
and Sus ainabili y (IDOS) gGmbH and he au ho (s).
IDOS Discussion Pape / Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS) gGmbH
ISSN 2751-4439 (P in )
ISSN 2751-4447 (Online)
ISBN 978-3-96021-255-3 (P in )
DOI: h ps://doi.o g/10.23661/idp10.2025
© Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS) gGmbH
Tulpen eld 6, 53113 Bonn
Email: publica ions@idos- esea ch.de
h ps://www.idos- esea ch.de
P in ed on eco- iendly, ce i ied pape .
IDOS Discussion Pape 10/2025
III
Abs ac
This pape examines ca bon leakage h ough supply chain ecalib a ion in esponse o
Eu opean ca bon policies. Using inpu -ou pu da a and a high- equency iden i ica ion
app oach o ca bon policy shocks, his pape in es iga es whe he s ingen ca bon
egula ions in Eu ope a ec he impo s o ca bon-in ensi e inpu s om majo eme ging
economies lacking simila policies. The indings e eal a empo a y inc ease in he a e o
change o impo s om eme ging coun ies ela i e o all inpu s in he ca bon-in ensi e sec o s
ollowing ca bon policy shocks, wi h e ec s peaking a e wo yea s be o e dissipa ing. While
no di ec ly quan i ying emissions ans e , his s udy sugges s some e idence o sho - e m
inpu subs i u ion pa e ns consis en wi h ca bon leakage h ough in e na ional supply chains.
IDOS Discussion Pape 10/2025
IV
Con en s
Abs ac
Abb e ia ions
1 In oduc ion 1
2 The Eu opean ca bon ma ke backg ound 5
3 Da a 6
3.1 Eme ging economy impo sha es 7
3.2 Sec o al-le el ca bon p icing da a 7
3.3 En i onmen al policy s ingency (EPS) 8
3.4 O he a iables 8
4 Summa y s a is ics 8
5 Empi ical me hods 9
5.1 Iden i ica ion challenges 9
5.2 Eme ging economy impo sha es 10
5.3 Dynamic e ec s o ca bon p icing on impo pa e ns om eme ging economies 11
5.4 He e ogeneous e ec s 11
6 Resul s 12
6.1 Main esul s 12
6.2 He e ogenei y analysis 14
7 Robus ness 16
7.1 Responses o ca bon ax 16
7.2 Al e na i e en i onmen al policy measu e 17
7.3 Sec o composi ion 17
8 Conclusions and policy implica ions 18
Re e ences 19
Appendix 21
IDOS Discussion Pape 10/2025
V
Figu es
Figu e 1: Ca bon p ice and ca bon-in ensi e impo a io ends 2
Figu e 2: Ca bon policy shocks in he Eu opean ca bon ma ke 6
Figu e 3: Impulse esponses o impo pa e ns o ca bon policy shock (pc ) 13
Figu e 4: He e ogenei y by Eu opean (impo ing) coun ies’ GDP 14
Figu e 5: He e ogenei y by Eu opean (impo ing) coun ies’ ca bon ax s a us 15
Figu e 6: He e ogenei y by sec o al ca bon in ensi y 16
Tables
Table 1: Summa y s a is ics 9
Figu es in he Appendix
Figu e A1: C oss-sec ional and empo al a ia ion o EPS index 21
Figu e A2: Robus ness: impulse esponses o ca bon ax a e 22
Figu e A3: Robus ness: es ic ed sample excluding ene gy-in ensi e sec o s 23
Tables in he Appendix
Table A1: Va iable desc ip ion 21
Table A2: En i onmen al policy s ingency and impo sha es om eme ging coun ies 24
Table A3: En i onmen al policy s ingency and impo sha es om eme ging coun ies 24
IDOS Discussion Pape 10/2025
VI
Abb e ia ions
CBAM Ca bon Bo de Adjus men Mechanism
CGE compu a ional gene al equilib ium
CO2 ca bon dioxide
EEA Eu opean Economic A ea
EE MRIO ex ended mul i- egion inpu -ou pu
EPS en i onmen al policy s ingency
ETS Emissions T ading Sys em
EU Eu opean Union
EU ETS Eu opean Union Emissions T ading Sys em
GDP g oss domes ic p oduc
GMM gene alised me hod o momen s
ICIO in e -coun y inpu -ou pu
IEA In e na ional Ene gy Agency
IPPC In e go e nmen al Panel on Clima e Change
ISIC In e na ional S anda d Indus ial Classi ica ion
NACE S a is ical Classi ica ion o Economic Ac i i ies in he Eu opean Communi y
OLS o dina y leas squa es
PCT pe cen age change
PPI p oduce p ice index
R&D esea ch and de elopmen
VAR ec o au o eg ession
WDI Wo ld De elopmen Indica o s (Wo ld Bank)
WIOD Wo ld Inpu -Ou pu Tables
IDOS Discussion Pape 10/2025
1
1 In oduc ion
Clima e change ep esen s a global en i onmen al challenge ha equi es coo dina ed
in e na ional ac ion. While go e nmen s wo ldwide ha e in oduced policies o cu b ca bon
emissions, he e ec i eness o unila e al egula ions emains deba ed due o po en ial ca bon
leakage – whe e s ingen ca bon policies in one egion lead o inc eased emissions in less
egula ed a eas (Copeland, Shapi o, & Taylo , 2022).
This conce n e lec s he b oade pollu ion ha en hypo hesis om ade heo y, which sugges s
ha en i onmen al egula ions in de eloped coun ies can shi pollu ion-in ensi e p oduc ion
owa ds egions wi h lowe aba emen (Le inson & Taylo , 2008). Clima e policy is pa icula ly
ulne able o such e ec s gi en i s global na u e and he signi ican a ia ion in egula o y
s ingency ac oss coun ies. The po en ial o p oduc ion and emissions o shi o less egula ed
egions can signi ican ly unde mine he e ec i eness o unila e al clima e policies in educing
global emissions, posing a c i ical challenge o clima e policy design and implemen a ion.
Ca bon leakage can occu h ough h ee dis inc channels (Colme , Ma in, Muûls, & Wagne ,
2024). Fi s , i ms may shi hei supply chains o sou ce mo e in e media e p oduc s om
un egula ed supplie s, po en ially educing compliance cos s while sac i icing some alue added.
Second, ma ke o ces may edis ibu e p oduc ion o i ms in un egula ed sec o s, ei he
domes ically o ab oad, as i ms in egula ed sec o s ace highe cos s. Thi d, companies wi h
mul iple acili ies migh ealloca e p oduc ion wi hin hei ne wo k om egula ed o un egula ed
loca ions. Each channel ep esen s a pa hway h ough which en i onmen al egula ions could
lead o unin ended inc eases in emissions in less egula ed egions.
While p e ious s udies such as Dechezlep ê e, Gennaioli, Ma in, Muûls, and S oe k (2022) ha e
ocused p ima ily on mul ina ional i ms o p oduc ion eloca ion and ound limi ed e idence o
ca bon leakage, his s udy examines he po en ially mo e esponsi e supply chain channel.
Speci ically, his esea ch in es iga es whe he Eu opean ca bon policies lead o inc eased
sou cing o ca bon-in ensi e inpu s om eme ging economies, whe e en i onmen al egula ions
a e ypically less s ingen .
This ocus on inpu sou cing, a he han comple e acili y eloca ion, is mo i a ed by se e al key
insigh s om he ade and en i onmen al li e a u e. Ede ing on, Le inson, and Minie (2005)
a gue ha pollu ion-in ensi e indus ies o en ace signi ican ba ie s o eloca ion due o high
anspo a ion cos s, subs an ial plan ixed cos s, and bene i s om indus ial agglome a ion.
Ins ead o eloca ing en i ely, hese indus ies migh ind i mo e easible o adjus hei inpu
sou cing pa e ns. Using de ailed inpu -ou pu da a, his pape examines how sec o s adjus hei
in e na ional sou cing decisions in esponse o ca bon policy changes, ocusing pa icula ly on
shi s owa d supplie s in eme ging economies. This app oach allows us o iden i y a po en ially
impo an bu unde s udied channel o ca bon leakage h ough supply chain ecalib a ion.
Figu e 1 p o ides sugges i e e idence o a supply chain adjus men channel. The a io o ca bon-
in ensi e impo s om eme ging economies closely acks he e olu ion o Eu opean Union
Emissions T ading Sys em (EU ETS) ca bon p ices, a e aged ac oss Exiobase sec o s based on
In e go e nmen al Panel on Clima e Change (IPCC) classi ica ions o Emissions T ading Sys em
(ETS) co e age. Impo a ios inc eased om a ound 4.5 pe cen in 2000 o o e 6 pe cen by
2020, wi h no able accele a ion a e he EU ETS in oduc ion in 2005. Pa icula ly sha p
inc eases in impo a ios coincided wi h pe iods o high ETS p ices, such as 2005 o 2006 and
2019 o 2020. This pa allel mo emen be ween sec o al ca bon p ices and impo pa e ns, while
no es ablishing causali y, sugges s ha Eu opean indus ies may adjus hei sou cing decisions
in esponse o ca bon policy changes. While his co ela ion is consis en wi h po en ial ca bon
leakage h ough supply chains, es ablishing causali y equi es add essing impo an empi ical
challenges.
IDOS Discussion Pape 10/2025
8
ca bon policy shocks as he key independen a iable while con olling o bo h ETS p ices and
ca bon ax a es o cap u e he o e all ca bon p icing landscape.
3.3 En i onmen al policy s ingency (EPS)
The En i onmen al Policy S ingency (EPS) index is ob ained om he OECD. This coun y-
speci ic measu e co e s 40 coun ies om 1990 o 2020 and quan i ies he deg ee o which
en i onmen al policies pu an explici o implici p ice on pollu ion beha iou . The index
inco po a es 13 ins umen s ela ed o clima e and ai pollu ion.8 As a s anda dised measu e, he
EPS index enables c oss-coun y and in e empo al analyses o en i onmen al egula ion e ec s.
I uses a scale om 0 (no s ingen ) o 6 (highes deg ee o s ingency), e lec ing he ela i e
s ingency o a coun y’s en i onmen al policy ins umen s in a gi en yea .9 This EPS index is
employed he e in he obus ness es s o alida e he main indings and ensu e he consis ency
o he esul s ac oss di e en measu es o en i onmen al policy s ingency.
3.4 O he a iables
This analysis inco po a es se e al con ol a iables o accoun o a ious economic and
en i onmen al ac o s. EU ETS p ices and ca bon ax a es a e included om he Wo ld Ca bon
P icing Da abase (Dolphin & Xiahou, 2022) o con ol o he exis ing policy en i onmen .10
Ca bon in ensi y da a om Exiobase en i onmen al accoun s con ols o di e ences in
p oduc ion e iciency ac oss coun ies and sec o s. Coun y-le el GDP om he Wo ld Bank’s
Wo ld De elopmen Indica o s (WDI) accoun s o economic de elopmen and ma ke size.
P oduce p ice indices (PPI) con ol o b oade cos p essu es a ec ing indus ial p oduc ion
decisions. T ade openness ( ade- o-GDP a io) cap u es coun ies’ in eg a ion in o global
ma ke s and hei p opensi y o ade. Toge he , hese a iables help isola e he e ec s o ca bon
policy shocks om o he ac o s ha migh in luence impo pa e ns.
4 Summa y s a is ics
The analysis co e s 1999 o 2019, ending be o e po en ial con ounding e en s: he Co id-19
pandemic, he Uk aine con lic , and he EU ETS Ma ke S abili y Rese e in oduc ion. The inal
sample comp ises 76 Exiobase sec o s ac oss 21 Eu opean coun ies. Table 1 p esen s
comp ehensi e s a is ics o he key a iables.
Panel A shows he dis ibu ion o ca bon-in ensi e impo sha es measu ed h ee ways. The a io
o o al impo ed inpu s (Impo Sha e1) a e ages 14.62 pe cen , wi h conside able a ia ion
(SD=15.32 pe cen ). When measu ed agains o al inpu s (Impo Sha e2), including bo h
domes ic and impo ed, he a e age sha e is lowe a 5.97 pe cen (SD=10.71 pe cen ), e lec ing
he impo ance o domes ic inpu s in p oduc ion. The a io o o al domes ic supply (Impo Sha e3)
shows a highe mean o 47.92 pe cen bu wi h subs an ial he e ogenei y (SD=253.19 pe cen ),
indica ing signi ican c oss-sec ional a ia ion in impo dependence ela i e o domes ic
p oduc ion capaci y.
8 The EPS index does no include wa e and was e managemen policies since he da a a e no a ailable in a
la ge c oss-coun y panel and a e also ha d o u n in o a quan i a i e c oss-coun y indica o .
9 Sco es a e assigned acco ding o he dis ibu ion o he obse a ions wi h he espec i e policy implemen ed.
10 See in his connec ion h ps://gi hub.com/g-dolphin/Wo ldCa bonP icingDa abase.
IDOS Discussion Pape 10/2025
9
Panel B p esen s he ca bon policy shock measu e de i ed om su p ises in ca bon pe mi p ices.
These shocks, cons uc ed as di e ences be ween ac ual and expec ed p ice changes, ha e nea -
ze o means. This aligns wi h a ional expec a ions heo y, as sys ema ic bias in expec a ions should
be a bi aged away. Howe e , he subs an ial s anda d de ia ions e eal signi ican unexpec ed
policy a ia ions ha could in luence sou cing decisions. The shock dis ibu ion sugges s equen
policy su p ises in bo h di ec ions, p o iding a ia ion o iden i ying causal e ec s.
Panel C summa ises he con ol a iables cap u ing a ious economic and policy dimensions.
Ca bon in ensi y exhibi s subs an ial sec o al he e ogenei y, e lec ing di e ences in p oduc ion
echnologies and ene gy e iciency. Economic indica o s like GDP, PPI, and ade openness show
conside able a ia ion ac oss coun ies and ime, highligh ing he impo ance o con olling o
di e en s ages o economic de elopmen and ma ke in eg a ion. Al hough ca bon axes a y by
coun y due o na ional policy di e ences, ETS p ices a e uni o m ac oss all EU coun ies in any
gi en yea . In he summa y s a is ics, bo h measu es a e a e aged ac oss Exiobase sec o s o
e lec sec o al co e age o bo h ca bon policies.
Table 1: Summa y s a is ics
Obse a ions
Mean
SD
P1
P50
P99
Panel A:
Ou come a iables
Impo Sha e1(o e
o al impo ed inpu s)
28318 14.62 15.32 5.04 9.63 17.89
Impo Sha e2(o e
o al inpu s)
28319 5.97 10.71 1.02 2.51 5.72
Impo Sha e3(o e
o al domes ic supply)
28319 47.92 253.1 1.42 4.01 11.88
Panel B:
Ca bon policy shock
CPShock(pc )
33516
-1.66e-12
2.32
-1.59
-.51
1.30
Panel C:
Con ol a iables
Log(GDP)
20748
20.00
1.32
19.09
19.77
21.37
Log(ca bon in ensi y)
33516
.45
1.46
.012
.06
.27
ETS p ice
33516
6.70
11.62
0
0
9.45
Ca bon ax a e
33516
2.50
11.43
0
0
0
PPI
25384
85.51
10.14
79.80
87.34
93.14
T ade openness
33516
98.82
40.79
64.54
86.39
127.77
No es: This able p esen s he summa y s a is ics o ou come a iables o impo pa e ns, ca bon policy shock
a iables, and con ol a iables. The a iables a e de ined in Appendix Table A1.
5 Empi ical me hods
5.1 Iden i ica ion challenges
E alua ing he impac o ca bon p icing on po en ial leakage e ec s ideally equi es a s udy design
ha isola es he changes induced by hese policies. Howe e , se e al challenges complica e his
ideal scena io. Fi s , iden i ying and isola ing he impac o ca bon policies is di icul gi en a ious
ac o s simul aneously in luencing i m and indus y beha iou . Second, pa icipa ion in ca bon
p icing schemes, such as he EU ETS, is no andom, in oducing po en ial selec ion bias. In
p ac ice, a andomised con ol ial o ca bon p icing is poli ically un easible. Resea che s use
econome ic me hods, such as di e ence-in-di e ences, assuming ha ca bon policies do no
a ec un egula ed i ms o sec o s. Howe e , his assump ion may be iola ed as policy e ec s
IDOS Discussion Pape 10/2025
10
can be ansmi ed h ough supply chains, po en ially con amina ing con ol g oups. Mo eo e ,
when s udying po en ial leakage h ough impo pa e ns, addi ional challenges a ise. Va ious
ac o s beyond ca bon policies, including global economic ends, ade ag eemen s, and
echnological changes may in luence changes in impo a ios. The g adual implemen a ion o
ca bon policies also makes i di icul o iden i y clea “be o e” and “a e ” pe iods o analysis.
To add ess hese iden i ica ion challenges, he high- equency iden i ica ion app oach de eloped
by Känzig (2023) was u ilised o quan i y ca bon policy shocks. This me hod builds on echniques
o iginally de eloped o mone a y policy analysis (Gü kaynak e al., 2004; Ge le & Ka adi, 2015;
Nakamu a & S einsson, 2018), whe e esea che s measu e asse p ice mo emen s in na ow
windows a ound policy announcemen s o isola e policy impac s. The EU ca bon ma ke ’s
equen policy upda es and ac i e u u es ading make i ideal o high- equency iden i ica ion.
By measu ing p ice changes in igh windows a ound egula o y e en s in he ca bon ma ke , his
app oach can plausibly ule ou e e se causali y since b oade economic condi ions a e al eady
inco po a ed in p e-e en p ices and unlikely o change wi hin he na ow e en window.
The cons uc ed ca bon policy shocks a e conside ed pa allel o exogenous shocks such as
wea he e en s o mone a y policy changes. This me hod allows o he isola ion o he impac o
ca bon policies om o he con ounding ac o s and o e comes he challenges o non- andom policy
implemen a ion. While ini ially de eloped o mone a y policy, his iden i ica ion s a egy has p o en
e ec i e in a ious policy con ex s, including global oil ma ke s and emissions ading schemes.
5.2 Eme ging economy impo sha es
Th ee measu es o impo sha es om eme ging economies a e cons uc ed as ollows:
Impo Sha eik,
m= Ca bon−in ensi e inpu s sou ced om eme ging economies𝑖𝑖𝑖𝑖,𝑡𝑡
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶−𝑖𝑖𝐶𝐶𝑖𝑖𝑖𝑖𝐶𝐶𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑑𝑑𝑖𝑖𝐶𝐶𝐶𝐶𝑑𝑑𝑖𝑖𝐶𝐶𝐶𝐶𝑖𝑖𝐶𝐶𝐶𝐶𝑚𝑚,
whe e Impo Sha eik,
m deno es he ac ion o ca bon-in ensi e inpu s ha sec o i in coun y k
sou ces om eme ging coun ies wi hou ca bon policies in yea , ela i e o he o al ca bon-
in ensi e inpu s o supply om all sou ces. Ca bon-in ensi e sec o s a e de ined as hose in he
op 40 pe cen o ca bon in ensi y ac oss all sec o s. The supe sc ip 𝑚𝑚 ∈1,2,3 indica es h ee
di e en measu es, each using a di e en denomina o :
• 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑚𝑚𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑1= o al ca bon-in ensi e impo ed inpu s
• 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑚𝑚𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑2= all ca bon-in ensi e inpu s (impo ed and domes ic)
• 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑚𝑚𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑3= o al ca bon-in ensi e domes ic supply.
These measu es allow o he assessmen o he ela i e impo ance o ca bon-in ensi e impo s
om eme ging coun ies ha lack ca bon policies. The i s measu e ocuses on he composi ion
o he impo s. This is use ul o di ec ly obse ing changes in in e na ional sou cing pa e ns. I
shows whe he ca bon-in ensi e impo s om eme ging economies a e inc easing ela i e o
impo s om o he sou ces. The second measu e includes bo h impo ed and domes ic inpu s. I
p o ides a b oade pe spec i e on how he sha e o ca bon-in ensi e impo s om eme ging
economies changes ela i e o he o al inpu mix. This measu e is pa icula ly impo an as i
indica es a po en ial inpu subs i u ion e ec , showing whe he he e is a gene al shi owa ds
o eign sou cing o jus a ecomposi ion o exis ing impo s. As such, i is he p ima y ocus o his
pape , o e ing he mos di ec e idence o whe he i ms subs i u e domes ic inpu s wi h impo s
om eme ging economies in esponse o ca bon policies. The hi d measu e compa es ca bon-
in ensi e impo s o he o al domes ic economic ac i i y in he sec o . I can indica e whe he
impo s a e g owing ela i e o domes ic p oduc ion, which is pa icula ly ele an o assessing
po en ial domes ic indus y impac s.
IDOS Discussion Pape 10/2025
11
5.3 Dynamic e ec s o ca bon p icing on impo pa e ns om
eme ging economies
The dynamic causal e ec s on he h ee measu es o ca bon-in ensi e impo sha es a e es ima ed
using (panel) local p ojec ions à la Jo dà (2005).
Δℎy𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ
𝑑𝑑=α+βℎCPShock𝑖𝑖+∑θ𝑝𝑝
ℎ𝑥𝑥𝑖𝑖𝑖𝑖,𝑖𝑖−𝑝𝑝
′
𝑃𝑃
𝑝𝑝=1 +σ𝑖𝑖
ℎ+ϵ𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ,
whe e h deno es he ho izon a which he ela i e e ec is es ima ed. The dependen a iable
Δℎy𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ
𝑑𝑑≡y𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ
𝑑𝑑−y𝑖𝑖𝑖𝑖,𝑖𝑖−1
𝑑𝑑 is de ined as he cumula i e di e ence o he impo sha e ou come
a iables (m∈1,2,3) measu ing he ac ion o ca bon-in ensi e inpu s ha indus y sec o k in
Eu opean coun y i sou ces om eme ging economies in yea +h, as de ined abo e.
Fo panel da a local p ojec ions, one would no mally p ojec he ou come a iables on he shocks
and con ol a iables, including he lag o he ou come a iable. Howe e , including he lagged
ou come a iable wi h ixed e ec s c ea es biases in he es ima ion ha would equi e mo e complex
gene alised me hod o momen s (GMM) me hods o add ess. To a oid his issue while main aining
simple o dina y leas squa es (OLS) es ima ion, he app oach ollows Jo dà, Schula ick and Taylo
(2015) and p ojec s he cumula i e di e ence y𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ
𝑑𝑑−y𝑖𝑖𝑖𝑖,𝑖𝑖−1
𝑑𝑑$ on he igh hand side a iables,
excluding he lagged dependen a iable. The independen a iable o in e es , CPShock𝑖𝑖 deno es
ca bon policy shocks a yea $ $, ex ac ed om a p oxy-VAR model om Känzig (2023). Following
Känzig and Kon ad (2023), lagged shock a iables a e no included in he speci ica ion since he
shock se ies shows no signi ican se ial co ela ion (Ljung-Box es p- alue = 0.88).
Pe cen age changes a he han baseline (eu o) changes in ca bon p ices a e employed as he shock
measu e, as pe cen age changes be e e lec how sec o s e alua e ela i e cos s and make sou cing
decisions. The ec o 𝒙𝒙𝒊𝒊𝒊𝒊,𝒕𝒕−𝒑𝒑
′ includes lagged con ol a iables including ca bon in ensi y, GDP, ETS
p ice, and ca bon ax a e, wi h lags up o o de P, allowing o iche dynamics in economic ac o s
ha migh in luence impo sha es. Sec o (des ina ion coun y’s) ixed-e ec s 𝜎𝜎𝑖𝑖
ℎ is included o
accoun o ime-in a ian cha ac e is ics. The supe sc ip h on he coe icien s indica es ha sepa a e
eg essions a e es ima ed o each ho izon h, allowing o dynamic e ec s o e ime.
5.4 He e ogeneous e ec s
He e ogeneous e ec s o ca bon policies on impo pa e ns om eme ging economies a e
explo ed ac oss di e en sec o ypes, coun y sizes, and exis ing ca bon p icing egimes.
Fi s , he a ious e ec s ac oss impo ing sec o s wi h di e en le els o ca bon in ensi y a e
examined. Impo ing sec o s o he Eu opean coun ies a e ca ego ised in o low, middle, and high
ca bon in ensi y g oups ( e ciles) based on hei p e-shock ca bon in ensi y ( on CO2 pe
housand-dolla g oss ou pu ). By examining he e ogenei y ac oss ca bon in ensi y le els, he
assessmen can de e mine whe he ca bon policies disp opo iona ely a ec high-emission
sec o s, po en ially leading o g ea e ca bon leakage.
Nex , he e ogenei y ac oss Eu opean impo ing coun ies is in es iga ed based on hei economic
size. Coun ies a e g ouped in o e ciles acco ding o hei logged GDP le els (in cons an 2015
dolla s). In he analysis, hese a e e e ed o as low, middle, and high GDP coun ies,
espec i ely. This app oach helps one o unde s and i he impac o ca bon p icing on impo
pa e ns a ies wi h he economic size o he impo ing coun y, po en ially e ealing di e ences in
adap a ion s a egies o ulne abili ies o ca bon leakage ac oss economies o di e en scales.
Finally, he e ogenei y based on he p esence o na ional ca bon axes in Eu opean impo ing
coun ies is explo ed. A dis inc ion is made be ween coun ies wi hou addi ional na ional ca bon
axes and hose wi h na ional ca bon axes on op o he EU-wide ETS sys em. This dis inc ion
IDOS Discussion Pape 10/2025
12
allows o examina ion o whe he he p esence o addi ional na ional ca bon p icing mechanisms
in luences he e ec o b oade ca bon policy shocks on impo pa e ns om eme ging economies.
6 Resul s
6.1 Main esul s
How do ca bon policies a ec he impo pa e ns o he Eu opean coun ies om he eme ging
coun ies? The baseline ca bon policy shocks a e ex ac ed om su p ises in eu o-denomina ed
ca bon p ice changes ela i e o p e ailing wholesale elec ici y p ices, as cons uc ed by Känzig
(2023) The h ee ou come a iables measu ing he ac ion o ca bon-in ensi e inpu s sou ced
om eme ging economies a e examined ela i e o i) o al impo ed inpu s; ii) all inpu s (impo ed
and domes ic); and iii) o al domes ic supply.
The esul s sugges modes e idence o Eu opean sec o s adjus ing hei sou cing o ca bon-
in ensi e inpu s in esponse o ca bon policy shocks. This is examined h ough h ee measu es,
each cap u ing di e en aspec s o po en ial ca bon leakage. Figu e 3 p esen s impulse
esponses o an unexpec ed inc ease in ca bon policies o all h ee measu es.
The middle panel shows he p ima y measu e – he a io o eme ging economy impo s o all
inpu s (bo h impo ed and domes ic) – which di ec ly cap u es subs i u ion be ween domes ic
and o eign sou ces. This measu e e eals a small bu posi i e cumula i e esponse o
app oxima ely 0.2 pe cen age poin s ha peaks a ound wo yea s a e a one s anda d de ia ion
inc ease in ca bon policy shock (no malised o inc ease ene gy p ices by one pe cen on
impac ), be o e g adually e u ning o ze o. While Känzig (2023), who de eloped he ca bon
policy shocks used in his s udy, ound ha hese shocks lead o a 0.6% educ ion in domes ic
GHG emissions, ou indings e eal a signi ican leakage channel h ough supply chain
adjus men s. The inc ease in ca bon-in ensi e impo s indica es ha ca bon leakage h ough
supply chains may o se some o he domes ic emissions gains, hough p ecise quan i ica ion
o his o se would equi e addi ional analysis. This esul con as s wi h Colme e al. (2024),
who ind no s a is ically signi ican changes in F ench i ms’ impo ing beha iou unde he EU
ETS and conclude ha supply chain leakage is no a majo d i e o emissions educ ions. This
s udy's indings can be conside ed alongside ea lie wo k by Sa o and Dechezlep ê e (2015),
who examined a di e en ime pe iod (1996-2011) and ound ha a 10 pe cen inc ease in
ene gy p ice di e ences be ween coun ies leads o a 0.2 pe cen inc ease in o e all impo s.
The ansi o y na u e o his e ec sugges s ini ial adjus men h ough in e na ional sou cing,
ollowed by longe - e m adap a ion h ough echnology upg ades o e iciency imp o emen s. The
op panel p esen s he a io o eme ging economy impo s o o al impo ed inpu s, showing simila
pa e ns in how impo composi ion shi s. The bo om panel shows he a io o o al domes ic
supply, whe e e ec s become less signi ican , likely due o b oade economic luc ua ions
a ec ing his measu e o domes ic p oduc ion.
These indings indica e ha , while s ic e ca bon policies may lead Eu opean sec o s o
empo a ily inc ease hei eliance on ca bon-in ensi e inpu s om eme ging economies, he
e ec s dissipa e o e ime a he han esul ing in pe manen shi s in sou cing pa e ns. This
empo al pa e n e eals impo an dynamics in how i ms adap o en i onmen al egula ions.
Ini ially, i ms appea o espond h ough he mos lexible a ailable channel – adjus ing hei
in e na ional sou cing pa e ns o main ain compe i i eness. Howe e , as i ms adap o e he
longe e m, hey likely de elop mo e sus ainable solu ions such as echnology upg ades,
e iciency imp o emen s, o p ocess inno a ions ha educe ca bon in ensi y. This e olu ion
om sho - e m ade adjus men s o longe - e m echnological adap a ion sugges s ha , while
ca bon leakage h ough supply chains may occu ini ially, i ms ul ima ely ind ways o main ain
p oduc ion while complying wi h s ic e en i onmen al egula ions. The ansi o y na u e o he
e ec s also indica es ha conce ns abou pe manen p oduc ion eloca ion o las ing damage o
IDOS Discussion Pape 10/2025
13
domes ic indus y compe i i eness may be o e s a ed, hough sho - e m adjus men cos s
emain impo an conside a ions o policy design.
Figu e 3: Impulse esponses o impo pa e ns o ca bon policy shock (pc )
No es: This igu e plo s he impulse esponses o inpu sou cing om eme ging economies ollowing a ca bon policy
shock, es ima ed using local p ojec ion. The op, middle, and bo om panels show he ac ion o inpu s ha indus y
sec o k in coun y i sou ces om eme ging economies in yea +h, ela i e o o al impo ed inpu s, all inpu s, and o al
domes ic supply in ca bon-in ensi e sec o s, espec i ely. The x-axis ep esen s yea s a e he shock. Ca bon policy
shocks (pc ) a e ex ac ed om he ca bon policy su p ises measu ed as eu o change in ca bon p ice, ela i e o he
p e ailing wholesale elec ici y p ice (Känzig, 2023). Solid lines ep esen poin es ima es, while da k- and ligh -shaded
a eas indica e 90 pe cen and 95 pe cen con idence bands, espec i ely.
Sou ce: Au ho ’s calcula ions
IDOS Discussion Pape 10/2025
14
6.2 He e ogenei y analysis
The he e ogenei y o ou esul s is examined ac oss mul iple dimensions, ocusing on bo h sec o -
speci ic and coun y-speci ic cha ac e is ics.
Figu e 4 illus a es he he e ogeneous e ec s o ca bon policies ac oss Eu opean (impo ing)
coun ies wi h di e en GDP le els. Lowe -GDP coun ies (bo om e cile) expe ience a mo e
p onounced inc ease in he a io o ca bon-in ensi e impo s om eme ging economies o o al
impo ed inpu s. This cumula i e e ec peaks a o e 0.5 pe cen a ound yea wo pos -shock,
be o e g adually diminishing o ze o by yea ou , sugges ing a empo a y bu signi ican adjus -
men in sou cing pa e ns. In con as , highe GDP coun ies ( op e cile) show a sligh dec ease
in his a io, po en ially indica ing hei g ea e capaci y o main ain o e en s eng hen domes ic
p oduc ion despi e s ic e en i onmen al egula ions. The con as be ween hese g oups
sugges s ha coun ies wi h lowe GDPs a e mo e ulne able o ca bon leakage e ec s, possibly
due o he limi ed echnological and inancial esou ces equi ed o adap hei p oduc ion
p ocesses. This inding implies ha economic size may play a c ucial ole in a coun y’s abili y o
main ain domes ic p oduc ion in he ace o s ingen ca bon policies, wi h smalle economies
po en ially mo e suscep ible o ou sou cing ca bon-in ensi e p oduc ion o eme ging economies.
The di e gen esponses also highligh he impo ance o conside ing coun y-speci ic
cha ac e is ics when designing ca bon policies, as uni o m egula ions may ha e une en
dis ibu ional consequences ac oss coun ies wi h di e en economic capaci ies.
Figu e 4: He e ogenei y by Eu opean (impo ing) coun ies’ GDP
No es: This igu e plo s he impulse esponses o inpu sou cing om eme ging economies ollowing a ca bon policy
shock, es ima ed using local p ojec ion, o low GDP coun ies (o ange), middle GDP coun ies (g ay), and high GDP
coun ies (blue). Eu opean coun ies a e g ouped in o e ciles based on hei logged GDP (2015 cons an ) le els. The
x-axis ep esen s yea s a e he shock. Ca bon policy shocks (pc ) a e ex ac ed om he ca bon policy su p ises
measu ed as eu o change in ca bon p ice, ela i e o p e ailing wholesale elec ici y p ice (Känzig, 2023). Shaded a eas
indica e 95 pe cen con idence bands.
Sou ce: Au ho ’s calcula ions
IDOS Discussion Pape 10/2025
15
Figu e 5 demons a es he e ogeneous e ec s based on Eu opean coun ies’ ca bon ax s a us.
Coun ies wi hou na ional ca bon axes (in addi ion o he EU-wide ETS) show a la ge inc ease
in ca bon-in ensi e impo s om eme ging ma ke s compa ed o hose wi h addi ional na ional
ca bon axes. This di e gence becomes pa icula ly p onounced wo yea s a e he shock,
sugges ing a delayed bu signi ican adjus men in sou cing pa e ns. This pa e n mi o s he
esul s in Figu e 4, likely because ca bon axes ha e been p ima ily implemen ed in weal hie
Wes e n and No he n Eu opean coun ies, which ypically ha e mo e esou ces o in es in
cleane p oduc ion echnologies. These indings indica e ha he implemen a ion o na ional
ca bon axes, complemen ing he EU-wide Emissions T ading Sys em (EU ETS), may enhance
e o s o mi iga e ca bon leakage h ough mul iple policy ins umen s wo king in andem.
Figu e 5: He e ogenei y by Eu opean (impo ing) coun ies’ ca bon ax s a us
No es: This igu e plo s he impulse esponses o inpu sou cing om eme ging economies ollowing a ca bon policy
shock, es ima ed using local p ojec ion, o coun ies wi hou na ional ca bon axes (pu ple), and coun ies wi h na ional
ca bon axes (na y) on op o he EU-wide ETS sys em. The x-axis ep esen s yea s a e he shock. Ca bon policy
shocks (pc ) a e ex ac ed om he ca bon policy su p ises measu ed as eu o change in ca bon p ice, ela i e o
p e ailing wholesale elec ici y p ice (Känzig, 2023). Shaded a eas indica e 95 pe cen con idence bands.
Sou ce: Au ho ’s calcula ions
Figu e 6 examines whe he sec o s wi h di e en ca bon in ensi ies espond di e en ly o ca bon
policy shocks. This analysis inds ha sec o s ac oss all ca bon in ensi y le els show simila
esponses, wi h high-ca bon sec o s being sligh ly mo e sensi i e o ca bon p icing. This pa e n
sugges s ha Eu opean ca bon policies lead o b oad-based shi s in sou cing pa e ns owa ds
eme ging economies, ega dless o sec o s’ emission in ensi ies. Such homogeneous esponses
ac oss sec o s poin o he possibili y ha supply chain ne wo ks ansmi policy-induced
adjus men s h oughou he indus ial s uc u e, a he han being con ined o he mos ca bon-
in ensi e ac i i ies.
IDOS Discussion Pape 10/2025
16
Figu e 6: He e ogenei y by sec o al ca bon in ensi y
No es: This igu e plo s he impulse esponses o inpu sou cing om eme ging economies ollowing a ca bon policy
shock, es ima ed using local p ojec ion, o low ca bon in ensi y (g een), middle ca bon in ensi y (g ay), and high ca bon
in ensi y sec o s (b own). The ca bon in ensi y (p e-shock) is measu ed as CO2 emissions (in ons) pe 2015 housand-
dolla g oss ou pu . Sec o s a e g ouped in o e ciles based on hei ca bon in ensi y. The x-axis ep esen s yea s a e he
shock. Ca bon policy shocks (pc ) a e ex ac ed om he ca bon policy su p ises measu ed as eu o change in ca bon p ice,
ela i e o p e ailing wholesale elec ici y p ice (Känzig, 2023). Shaded a eas indica e 95 pe cen con idence bands.
Sou ce: Au ho ’s calcula ions
7 Robus ness
7.1 Responses o ca bon ax
The main iden i ica ion s a egy using ca bon policy shocks is complemen ed wi h a “con ol-
based” app oach ollowing Me cal and S ock (2020) and Känzig and Kon ad (2023). This
app oach iden i ies he e ec s o ca bon axes by con olling o a ious economic and sec o al
ac o s o isola e plausibly exogenous a ia ions in ca bon p ices. This complemen a y analysis
se es bo h as a obus ness check and p o ides insigh s in o how di e en ypes o ca bon p icing
policies migh a ec supply chain adjus men s. The ollowing (panel) local p ojec ion is es ima ed:
Δℎy𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ =α+βℎC ax𝑖𝑖𝑖𝑖,𝑖𝑖+� θ𝑝𝑝
ℎ𝑥𝑥𝑖𝑖𝑖𝑖,𝑖𝑖−𝑝𝑝
′
𝑃𝑃
𝑝𝑝=1 +σ𝑖𝑖
ℎ+ϵ𝑖𝑖𝑖𝑖,𝑖𝑖+ℎ,
whe e he independen a iable o in e es C ax𝑖𝑖𝑖𝑖,𝑖𝑖 is he ca bon ax11 imposed on coun y i and
sec o k. The ou come a iable is he ac ion o ca bon-in ensi e inpu s ha indus y sec o k in
coun y i sou ces om eme ging economies in yea +h, ela i e o all inpu s (m=2). The ec o
11 As in he Wo ld Ca bon P icing Da abase: ne ax a e (accoun ing o exemp ion) in cu en local cu ency
uni pe onne o CO2.
IDOS Discussion Pape 10/2025
17
𝒙𝒙𝒊𝒊𝒊𝒊,𝒕𝒕−𝒑𝒑
′ includes lagged con ol a iables including he coun y’s GDP, sec o al ca bon in ensi y,
p oduce p ices, and ade openness. These con ols help accoun o a ious economic ac o s
ha migh in luence sou cing decisions independen ly o ca bon axa ion. Simila o he main
eg ession, sec o (des ina ion coun y’s) ixed-e ec s 𝜎𝜎𝑖𝑖
ℎ is included o accoun o ime-in a ian
cha ac e is ics ha migh a ec he p opensi y o sou ce om eme ging economies.
Figu e A2 in he Appendix shows small bu posi i e esponses o ca bon ax changes, wi h e ec s
becoming signi ican in yea 3. While hese esul s align di ec ionally wi h he main indings om
he policy shock analysis, he e ec s a e no ably smalle in magni ude and ake longe o
ma e ialise. The smalle and less signi ican e ec s likely e lec he g adual, an icipa ed na u e
o ca bon ax changes compa ed o he unexpec ed policy shocks in ou main speci ica ion. This
di e ence in esponse pa e ns highligh s he impo ance o he high- equency iden i ica ion
s a egy in cap u ing ma ke esponses o ca bon policy changes, as i be e isola es he
immedia e supply chain adjus men s o policy inno a ions. The delayed esponse o ca bon axes
also sugges s ha i ms may ha e mo e ime o plan and implemen al e na i e adjus men
s a egies when acing g adual, p edic able policy changes compa ed o sudden policy shocks.
7.2 Al e na i e en i onmen al policy measu e
The en i onmen al policy s ingency (EPS) index is also employed as an al e na i e measu e o
en i onmen al egula ion. De eloped by he OECD, his comp ehensi e index p o ides a
quan i a i e assessmen o en i onmen al policy s ingency ac oss 40 coun ies om 1990 o
2020, cap u ing he ex en o which na ional policies impose explici o implici cos s on pollu ing
ac i i ies. The index inco po a es bo h ma ke -based ins umen s (such as en i onmen al axes
and ading schemes) and non-ma ke egula ions (including emission limi s and esea ch and
de elopmen (R&D) subsidies), p o iding a b oade pe spec i e on en i onmen al policy han
ca bon p icing alone. Figu e A1 in Appendix shows he c oss-sec ional and empo al a ia ions o
he EPS index in Eu opean coun ies, highligh ing subs an ial he e ogenei y in en i onmen al
policy s ingency ac oss bo h coun ies and ime. The ollowing OLS eg ession is es ima ed:
Impo Sha e𝑖𝑖𝑖𝑖,𝑖𝑖=α+βEPS𝑖𝑖,𝑖𝑖+θ𝑝𝑝𝑥𝑥𝑖𝑖𝑖𝑖,𝑖𝑖
′+σ𝑖𝑖+ϵ𝑖𝑖𝑖𝑖,𝑖𝑖,
whe e EPS𝑖𝑖,𝑖𝑖 is he en i onmen al policy s ingency index o Eu opean coun y i in yea , 𝒙𝒙𝒊𝒊𝒊𝒊,𝒕𝒕
′ is
a ec o o con ol a iables, 𝜎𝜎𝑖𝑖 ep esen s sec o ixed e ec s, and 𝜖𝜖𝑖𝑖𝑖𝑖,𝑖𝑖 is he e o e m.
Table A2 in he Appendix epo s he esul s om his eg ession. The es ima ed coe icien s on
EPS a e posi i e and highly signi ican o impo sha e a iables (m=1, 2), indica ing a obus
ela ionship be ween en i onmen al policy s ingency and ca bon-in ensi e impo s om
eme ging coun ies. This aligns wi h he main eg ession esul s and p o ides addi ional e idence
ha s ic e en i onmen al policies may lead o supply chain adjus men s h ough inc eased
sou cing om coun ies wi h weake en i onmen al egula ions. Mo eo e , addi ional es s a e
conduc ed using a 3-yea mo ing a e age o he EPS index o cap u e mo e pe sis en policy
e ec s. These esul s ( epo ed in Appendix Table A3) show simila pa e ns, u he suppo ing
he obus ness o he main esul s.
7.3 Sec o composi ion
While he main analysis compa es esponses ac oss sec o s o di e en ca bon in ensi ies, he
obus ness o he indings is u he es ed by excluding ene gy-in ensi e sec o s. These sec o s,
including mining and qua ying (NACE B05-B08), pe oleum p oduc s (C19), chemicals (C20),
non-me allic mine als (C23), and basic me als (C24), ypically show high sensi i i y o ca bon
policies and o en ecei e special egula o y ea men such as ee allowance alloca ions in he
IDOS Discussion Pape 10/2025
24
Table A2: En i onmen al policy s ingency and impo sha es om eme ging coun ies
(1)
(2)
(3)
VARIABLES
Impo Sha e
1
Impo Sha e
2
Impo Sha e
3
EPS
2.608***
1.299***
9.203
(0.361)
(0.251)
(7.417)
Log(GDP)
2.397***
1.179***
8.732
(-0.918)
(0.190)
(5.398)
Log(ca bon in ensi y)
-0.918
0.190
26.08
(1.181)
(0.900)
(25.33)
PPI
-0.331***
-0.187***
-0.512
(0.0388)
(0.0322)
(0.626)
T ade openness
0.239***
0.157***
1.601***
(0.0249)
(0.0201)
(0.433)
Cons an
-22.38***
-5.626
-64.21
(7.167)
(4.543)
(67.99)
Obse a ions
13950
13951
13950
R-squa ed
0.290
0.315
0.160
Adjus ed R-squa ed
0.286
0.311
0.155
No es: Impo Sha e1,2,3 ep esen he sha e o impo s o e (1) o al impo ed inpu s, (2) o al inpu s, and (3) o al
domes ic supply, espec i ely, in ca bon-in ensi e sec o s. Robus s anda d e o s a e clus e ed a he sec o le el. *, **,
and *** indica e signi icance a he 10 pe cen , 5 pe cen , and 1 pe cen le els, espec i ely.
Sou ce: Au ho ’s calcula ions
Table A3: En i onmen al policy s ingency and impo sha es om eme ging coun ies
(1)
(2)
(3)
VARIABLES
Impo Sha e
1
Impo Sha e
2
Impo Sha e
3
EPS_3MA
2.419***
1.284***
14.06*
(0.390)
(0.276)
(7.593)
Log(GDP)
2.497***
1.190***
8.647
(0.290)
(0.244)
(5.639)
Log(ca bon in ensi y)
-0.879
0.272
31.40
(1.201)
(0.939)
(26.42)
PPI
-0.387***
-0.215***
-0.708
(0.0438)
(0.0365)
(0.698)
T ade openness
0.236***
0.155***
1.605***
(0.0248)
(0.0201)
(0.444)
Cons an
-29.42***
-16.56***
-245.3*
(6.106)
(5.291)
(143.7)
Obse a ions
13147
13148
13147
R-squa ed
0.295
0.320
0.163
Adjus ed R-squa ed
0.291
0.316
0.158
No es: Impo Sha e1,2,3 ep esen he sha e o impo s o e (1) o al impo ed inpu s, (2) o al inpu s, and (3) o al
domes ic supply, espec i ely, in ca bon-in ensi e sec o s. Robus s anda d e o s a e clus e ed a he sec o le el. *, **,
and *** indica e signi icance a he 10 pe cen , 5 pe cen , and 1 pe cen le els, espec i ely.
Sou ce: Au ho ’s calcula ions
