A guey olles, Robin; Heimann, Tobias; Delzei , Ru h
A icle — Published Ve sion
Impac o Gasoline and Diesel Subsidy Re o ms on Global
Bio uel Manda es
Global Change Biology Bioene gy
P o ided in Coope a ion wi h:
Kiel Ins i u e o he Wo ld Economy – Leibniz Cen e o Resea ch on Global Economic Challenges
Sugges ed Ci a ion: A guey olles, Robin; Heimann, Tobias; Delzei , Ru h (2025) : Impac o Gasoline
and Diesel Subsidy Re o ms on Global Bio uel Manda es, Global Change Biology Bioene gy, ISSN
1757-1707, Wiley, Hoboken, NJ, Vol. 17, Iss. 2, pp. 1-14,
h ps://doi.o g/10.1111/gcbb.70019
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/323486
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/
1 o 14
GCB Bioene gy, 2025; 17:e70019
h ps://doi.o g/10.1111/gcbb.70019
GCB Bioene gy
RESEARCH ARTICLE OPEN ACCESS
Impac o Gasoline and Diesel Subsidy Re o ms on Global
Bio uel Manda es
RobinA guey olles1 | TobiasHeimann2 | Ru hDelzei 1
1Depa men o En i onmen al Sciences, Uni e si y o Basel, Basel, Swi ze land | 2Kiel Ins i u e o he Wo ld Economy, Kiel, Ge many
Co espondence: Robin A guey olles ( obin.a guey [email protected])
Recei ed: 21 Oc obe 2024 | Re ised: 10 Decembe 2024 | Accep ed: 14 Decembe 2024
Funding: The au ho s ecei ed no speci ic unding o his wo k.
Keywo ds: bio- economy| bio uel| compu able gene al equilib ium (CGE)| ene gy ansi ion| ossil uel subsidy e o m| leakages
ABSTRACT
Fossil uel subsidy e o m(s) suppo he deploymen o low- ca bon echnologies, ye ossil uel subsidies emain s ubbo nly high,
while money alloca ed by go e nmen s o enewable ene gy con inues o g ow. In he anspo sec o , his ension is obse ed
be ween bio uels ha s ill ely on na ional policies and gasoline/diesel subsidies. Using a global Compu able Gene al Equilib ium
(CGE) model, we s udy how phasing ou gasoline and diesel subsidies would impac global bio uel manda es. We ind ha whe e
hey a e implemen ed, Fossil Fuel Subsidy Re o ms inc ease bio uel compe i i eness and lowe he cos o achie ing he man-
da es. The iscal bene i is he e o e wo old wi h sa ings on ossil and bio- based ene gy subsidies. In a mul ila e al e o m sce-
na io, we simula e he ise in iscal e enue om phasing ou he ossil uel subsidies o be 25% highe when he a oided spending
on bio uels' suppo is accoun ed o . In he es o he wo ld, howe e , he bio uel a ge s become cos lie o achie e as he p ice
o ossil uels d ops. Conside ing ha global bio uel 2030 a ge s a e achie ed, go e nmen s' suppo o bio uel alls by $6 billion
in egions phasing gasoline and diesel subsidies bu inc eases by $600 million in he es o he wo ld.
1 | In oduc ion
Fossil uel subsidies b oke an all- ime high in 2022 (In e na ional
Ene gy Agency2023b), while money alloca ed by go e nmen s
o enewable ene gy con inued o g ow (In e na ional Ene gy
Agency 2023a). Ye , Fossil Fuel Subsidy Re o ms (FFSR) a e
known o suppo he deploymen o low- ca bon echnologies
(B idle and Ki son 2014; In e na ional Ene gy Agency 2014).
This ension be ween suppo /subsidy p og ams is e iden in
he anspo sec o . I is he la ges single- sec o ecipien o os-
sil uel subsidies (Black e al.2023), while bio uels can be one o
he mos success ul ways o deca bonizing he anspo sec o
(Ebadian e al.2020) s ill elying on na ional suppo policies o
pene a e he ma ke (OECD/FAO2021).
Se e al coun ies e e o a ossil uel subsidy phase- ou in
hei Na ionally De e mined Con ibu ions, and 50 unde ook
a ious deg ees o e o ms be ween 2015 and 2018 (Global
Subsidies Ini ia i e 2019). The esul s o pas e o ms we e
mixed (UNDP 2021). In Indonesia, o example, subsidies o
ossil uels used in he anspo sec o d opped om $15 bil-
lion o almos ze o ollowing a success ul uel- p icing e o m in
2015–2017 (OECD2019b). These, howe e , bounced back wi h
gasoline and diesel subsidies in 2022 es ima ed he e a $6.7
billion and $12.1 billion, espec i ely (In e na ional Mone a y
Fund 2023). Simul aneously, he Indonesian go e nmen
con inues o p omo e biodiesel o a clean ene gy ansi ion
(De mawan e al.2022), while low oil p ices impede bio uels'
compe i i eness (Winches e and Led ina2017).
A la ge body o li e a u e exis s on he impac o FFSRs and
bio uel manda es on he en i onmen and wel a e. Food se-
cu i y has been a conce n o bio uels (Mi chell 2008; Zhang
e al. 2013; Zilbe man e al. 2013), while hei g eenhouse
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is
p ope ly ci ed.
© 2025 The Au ho (s). GCB Bioene gy published by John Wiley & Sons L d.
2 o 14 GCB Bioene gy, 2025
gas (GHG) emission balance is ambiguous and depends on,
among o he hings, land- use changes (Sea chinge e al.2008;
Timilsina and Me el 2013) and eeds ock composi ion
(Labo de and Valin2012; B i z and Delzei 2013), whe e hey
a e p oduced (Ma eike 2011), and by- p oduc conside a ions
(Tahe ipou e al.2010). FFSRs on he o he hand we e ound o
lowe GHG emissions (Bu niaux and Cha eau2014; Chepelie
and Mensb ugghe2020; Black e al.2023) bu o be eg essi e
i no complemen ed by e enue ecycling mechanisms (Siddig
e al.2014; Wesseh, Lin, and A sagli2016).
To he bes o ou knowledge, ossil uel subsidy e o ms' im-
pac on bio uel ma ke s and manda es emains unde e mined.
This s udy does no aim o p o ide an e alua ion o ei he policy
bu a he in es iga e he in e ac ion be ween gasoline and die-
sel FFSR and exis ing global bio uel manda es. The ocus is no
on he combined en i onmen al o dis ibu ional impac o he
policies. Ins ead, we p o ide a de ailed e alua ion o he impac
o gasoline and diesel FFSRs on ossil and bio uel sec o s in a
wo ld whe e exis ing global bio uel manda es a e achie ed. The
goal is o explo e how he consump ion and compe i i eness o
anspo uel changes and how his, in u n, impac s he go -
e nmen s' iscal e enue and he o e all economy o p o ide an
e alua ion o FFSRs ha accoun o in e ac ions wi h bio uel
manda es.
Conside ing ha ossil and bio uel commodi ies a e pa o a
complex global alue chain and ou ocus on how hese in e ac ,
we employ he DART- BIO model, a global ecu si e dynamic
Compu able Gene al Equilib ium (CGE) model (Calzadilla,
Delzei , and Kleppe 2017; Delzei , Winkle , and Söde 2018).
CGE models ha accoun o he linkage and eedback e ec s
be ween ma ke s a e pa icula ly well sui ed o he s udy o
bio uel policies (K e schme and Pe e son 2010). Simila ly,
hey cap u e he signi ican gene al equilib ium e ec o FFSR
(Saunde s and Schneide 2000) and linkages be ween all ma -
ke s including hose ha equi e ene gy as an inpu (Ellis2010).
We in eg a e o he DART- BIO model p e- and pos - ax es ima es
om he IMF o gasoline and diesel subsidies. P e- ax subsi-
dies measu e suppo mechanisms ha di ec ly impac end- use
p ices. These a e ela i ely less complex han o he o ms o
go e nmen suppo and a e gene ally well- accep ed o i he
de ini ion o a subsidy by di e en s akeholde s (UNEP, OECD,
IISD2019). This is help ul in iew o he ac ha go e nmen s
like he UK easu y a e epo ed o ha e denied p o iding sub-
sidies a all “on he g ounds ha any suppo i o e s doesn'
a i icially lowe p ices paid by consume s” (Meh a2022). Pos -
ax subsidies on he o he hand accoun o he cos o ex e -
nali ies indi ec ly paid by socie y and he e o e all ou side he
de ini ion o subsidies om he Ag eemen on Subsidies and
Coun e ailing Measu es (ASCM). S ill, no paying he cos o
damages o he en i onmen om p oduc ion can legi ima ely
be unde s ood as a subsidy (S igli z2006). Pos - ax subsidies also
p o ide an oppo uni y o in es iga e e icien p icing policies
which is key o in o ming he discussion on uel p icing e o ms
(Coady e al.2019).
We ind ha he subsidy e o ms ein o ce bio uel manda es.
Whe e hey a e implemen ed, FFSRs inc ease bio uel compe i-
i eness and lowe he cos o achie ing he manda es. The iscal
bene i o he e o ms is he e o e wo old wi h sa ings on bo h
ossil uel and bio uel subsidies. Accoun ing o he dec ease in
bio uel suppo om a mul ila e al p e- ax gasoline and diesel
subsidies e o m, we simula e he o e all ise in iscal e enue o
be 25% highe . In he es o he wo ld, howe e , he bio uel a -
ge s become cos lie o achie e as he in e na ional p ice o ossil
uels d ops. Conside ing ha bio uel 2030 a ge s a e achie ed
in all egions, we ind ha bio uel subsidies all by $6 billion
in egions phasing ou hei subsidies bu ha his is pa ially
o se by a $600 million inc ease in he es o he wo ld. We
conclude ha aligning policy objec i es is a low- hanging ui
gi en he cumula i e bene i s i gene a es a a ela i ely low eco-
nomic cos .
The es o he a icle is s uc u ed as ollows: Sec ion2 in o-
duces he DART- BIO model, he da a used, and he de ini ion
o scena ios; Sec ion3 p esen s he esul s o he simula ions;
Sec ion4 discusses he esul s; Sec ion5 p esen s some o he
limi a ions o he s udy. Finally, Sec ion 6 concludes and ex-
plo es policy implica ions.
2 | Ma e ials and Me hods
Fossil Fuel Subsidy Re o ms (FFSRs) and bio uel policies p o-
oke simul aneous adjus men in he ossil uel and bio uel
ma ke . A he same ime, bio uels a e embedded in a highly
in eg a ed ag icul u al ma ke ha elies on ene gy inpu s o
sus ain p oduc ion. Cap u ing hese in e ac ions is essen ial o
unde s anding he impac o each policy on one ano he . Pa ial
equilib ium models ha s uggle o ep esen he link be ween
sec o s a e no well sui ed o add ess ques ions ela ing o in-
e na ional compe i i eness e ec s (Ellis 2010). Ins ead, CGE
models ha e been ecognized o be pa icula ly well- equipped
o cap u e hese complex in e ac ions because hei sec o s ep-
esen he ull economy (K e schme and Pe e son2010).
2.1 | Model
We use he DART- BIO model, which is a e sion o he Dynamic
Applied Regional T ade (DART) model designed o ep esen
in de ail he ag icul u al sec o , land use, and con en ional bio-
uels. The DART model is a global mul isec o al, mul i egional
ecu si e- dynamic CGE model ha was de eloped in he 1990s
a he Kiel Ins i u e o he Wo ld Economy (Sp inge 1998).
I has since hen been widely applied o analyze in e na ional
clima e policies (e.g., Kleppe and Pe e son 2006a), en i-
onmen al policies (e.g., Wei zel e al. 2012), ene gy policies
(e.g., Kleppe and Pe e son 2006b), and bio uel policies (e.g.,
K e schme , Na i a, and Pe e son2009; Calzadilla, Delzei , and
Kleppe 2017; Delzei , Winkle , and Söde 2018; Schuenemann
and Delzei 2022).
The DART- BIO is he bio- economy and land- use e sion o he
DART model. Fundamen ally, bo h sha e he same s uc u e.
The bio- economy ex ension howe e includes land he e ogene-
i y using he ag o- ecological zones classi ica ion as well as he
complex p oduc ion p ocess chains o bio uels. The model is
la gely calib a ed on he GTAP9 da abase (Aguia , Na ayanan,
and McDougall 2016). I s cu en agg ega ion includes 23
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
3 o 14
egions, 51 sec o s, and 21 ac o s o p oduc ion. No iceably, as
pe Table1, he OPEC coun ies and Qa a ha e been agg ega ed
based on hei simila i ies wi h espec o ene gy subsidies. Fo
he same eason, coun ies ha ha e signed he Decla a ion
o Coope a ion (DoC) ha e been agg ega ed in o he PPEC e-
gion. Russia and Malaysia a e no included in he PPEC egion
al hough hey ha e signed he decla a ion. Ins ead, Russia is
modeled as a single egion while Malaysia is agg ega ed wi h
Indonesia due o i s impo an ole in wo ldwide bio uel p oduc-
ion and consump ion.
As de ailed by Delzei e al.(2021), o enable he modeling o
he ela ionship be ween bio- based and ossil uel- based uel
o anspo , se e al key sec o s we e disagg ega ed om he
o iginal GTAP da abase. Conce ning bio uels, biodiesel and
bioe hanol om di e en sou ces (i.e., oilseeds and g ains)
we e included as well as hei co esponding p ocessing sec o s
and by- p oduc s. The adap ed da abase also includes dedica ed
sec o s o mo o gasoline and mo o diesel. Sec o s we e spli
using da a om he meó Consul ing Team, a company p o iding
consul ing se ices wi h a special ocus on enewables' sus ain-
abili y and clima e change (F.O. Lich 2015), and FAOSTAT on
p oduc ion, p ice, and bila e al ade da a. P oduc ion olumes
we e con e ed in dolla s be o e he ade sha es we e used o
dis ibu e p oduc ion assuming ade cos s, a i s, and expo
axes/subsidies emained unchanged om he o iginal GTAP
sec o (Schuenemann and Delzei 2019).
The DART- BIO model is a classical Wal asian gene al equilib-
ium model whe e he economy in each egion is modeled as
a compe i i e economy wi h lexible p ice and ma ke clea ing
condi ions. A single agen simul aneously ep esen s u ili y-
maximizing consume s, p o i - maximizing p oduce s, and he
egional go e nmen . The economies e ol e h ough a sequence
o single- pe iod s a ic equilib ium connec ed h ough capi al ac-
cumula ion and changes in labo supply o pe iods om 2011
un il 2030. Changes in he labo o ce, he a e o labo p o-
duc i i y g ow h, and human capi al accumula ion de e mine
labo supply and p oduc i i y changes. Labo p oduc i i y and
human capi al g ow h a e assumed o be cons an bu egionally
di e en ia ed.
P oduc ion exhibi s a cons an e u n o scale and is modeled
using mul ile el nes ed cons an elas ici y o subs i u ion (CES)
p oduc ion unc ions. C ucially, in he case o mo o gasoline
and diesel, c ude oil en e s eed oil in p oduc ion wi h no subs i-
u ion possibili ies wi h labo and capi al. Biodiesel on he o he
hand can be p oduced using a ious oil seeds ha can be subs i-
u ed o one ano he , while bioe hanol is a composi e o a ious
g ain- speci ic bioe hanol p oduc s, each o which wi h i s own
TABLE 1 | Lis o egions in DART- BIO.
Cen al and Sou h Ame ica Eu ope
BRA B azil FSU Res o he o me So ie Union
PAC Pa aguay, A gen ina,
U uguay, Chile
CEU Cen al Eu opean Union wi h Belgium,
F ance, Luxembou g, Ne he lands
LAM Res o La in Ame ica DEU Ge many
Middle Eas and No he n A ica MED Medi e anean wi h Cyp us, G eece,
I aly, Mal a, Po ugal, Spain
MEA Res o he Middle
Eas and A ica
MEE Eas e n EU wi h Aus ia, Czech Republic,
Es onia, Hunga y, La ia, Li huania, Poland,
Slo akia, Slo enia, Romania, Bulga ia, C oa ia
Asia NEW No h- Wes e n EU wi h Denma k, Finland,
I eland, Sweden, Uni ed Kingdom
CHN China, Hong Kong NOR No way
IND India RNE Res o No he n Eu ope. Swi ze land,
Iceland, Liech ens ein
EAS Eas e n Asia. Japan, Sou h
Ko ea, Taiwan, Singapo e
No h Ame ica
MAI Malaysia, Indonesia CAN Canada
ROA Res o Asia USA Uni ed S a es o Ame ica
RUS Russia OPEC and pa ne
Oceania OPEC Alge ia, Angola, Congo, Ecuado , I an, I aq,
Kuwai , Libya, Nige ia, Gabon, Saudi A abia,
Uni ed A ab Emi a es, Venezuela, Qa a
ANZ Aus alia, New Zealand,
Res o Oceania
PPEC Aze baijan, Bah ain, B unei Da ussalam,
Kazakhs an, Mexico, Oman, Sou h Sudan
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
4 o 14 GCB Bioene gy, 2025
nes ing and p oduc ion unc ion. Final consump ion in u n is
modeled wi h nonuni a y income elas ici ies using he linea
expendi u e sys em (LES) app oach and is composed o bo h a
ixed subsis ence le el and a supe nume a y consump ion. Only
he la e eac s o changes in p ice/income and he e o e allows
subs i u ion be ween biodiesel and diesel o bioe hanol and gas-
oline, as well as, o a lesse ex en , be ween each o he . Fo a ull
echnical desc ip ion o he model, see Delzei e al.(2021).
2.2 | Subsidies and Calib a ion
The cu en s udy complemen s he model wi h he 2022
IMF's publicly a ailable da a on ossil uel subsidies published
in 2015 and 2021. The es ima es co e 150 and 192 coun ies
in he espec i e da abases and he 2003–2025 pe iod. Speci ic
gasoline and diesel es ima es a e included, allowing he in e-
g a ion o he subsidies in o he model wi h no addi ional com-
modi y disagg ega ion. The IMF de i es p e- ax ossil uel
subsidies using he p ice- gap app oach by compa ing a e age
end- use p ices wi h ee- ma ke e e ence p ices e lec ing
he ull cos o supply (Coady e al.2019). Suppo measu es
ha do no in luence end- use p ices a e no cap u ed, bu
he me hod bene i s om being ela i ely low da a- in ensi e
and measu ing a well- accep ed ca ego y o subsidies (UNEP,
OECD, IISD2019).
The da abase also con ains pos - ax subsidies, which a e de ined
as he exemp ion om he co ec i e Pigou ian ax e lec ing
he cos o ex e nali ies (Coady e al.2019). These include he
cos o local ai pollu ion mo ali y, b oade cos s associa ed
wi h he use o uels in oad ehicles, and global wa ming (Pa y,
Black, and Ve non2021). Pos - ax subsidies s ay away om he
WTO 1994 de ini ion o a subsidy a i ied in he Ag eemen on
Subsidies and Coun e ailing Measu es (ASCM). Conside ing
he duali y o he opic and he link o clima e mi iga ion, pos -
ax subsidies, howe e , o e an oppo uni y o in es iga e e i-
cien p icing policies.
To ma ch he model speci ica ion, he ossil uels subsidy and
consump ion alues p o ided by he IMF a e summed be ween
coun ies o ma ch he egions in Table1. Di ing he esul ing
subsidies and consump ion alues by each o he allows us o de-
i e he ad- alo em subsidy a e co esponding o he model e-
gional agg ega ion. To in eg a e hese ad- alo em a es in o he
model, a new (nega i e) sales ax a iable is applied o he mod-
el's A ming on agg ega e o impo s and domes ic consump ion.
This means ha all agen s ace he same a es, as well as indus-
ies consuming he same in e media es.
The model's base yea alues a e ecalcula ed ollowing he
in eg a ion o he subsidies o minimize he de ia ion om
he benchma k alues which implici ly e lec s he dis o ions
o he ma ke . The model is hen calib a ed using labo p o-
duc i i y o ma ch he egional GDP g ow h p ojec ions o he
OECD (2019a). To simula e a obus baseline, we simul ane-
ously calib a ed he p oduc ion o mo o gasoline, diesel, and
c ude oil. Be ween 2011 and 2020, he global cumula i e g ow h
o gasoline and diesel p oduc ion as well as he sha e o each
egion in global p oduc ion is calib a ed o eplica e igu es e-
po ed by he Uni ed Na ions S a is ics Di ision. A e 2020, we
calib a ed he egional p oduc ion sha e o c ude oil based on
he Join Resea ch Cen e (JRC) da a (Woj owicz e al.2021).
2.2.1 | Baseline Simula ion (“Re ”)
In he baseline, ossil uel subsidies a e kep cons an a e 2021.
Bio uel manda es' a ge s o biodiesel and bioe hanol a e g ad-
ually and linea ly eached in all egions by 2030. Bio uel a ge s
a e exp essed as pe cen o consump ion and based on he p ojec-
ions used in he FAO/OECD 2022 ag icul u al ou look (OECD/
FAO2021). The EU manda e se s maximum consump ion sha es
ins ead o a minimum blending a e o consump ion a ge .
Conside ing ha he EU 2030 a ge o 14% sha e o enewables
in o al anspo uel consump ion canno be me wi hou a
su icien ly la ge amoun o bio uels, we assume ha membe
s a es mee he enewable ene gy in anspo a ge wi h he
Di ec i e(2018) maximum allowable sha e o bio uels. As pe he
EU Renewable Ene gy Di ec i e RED, palm oil- based biodiesel
is phased ou in he EU due o i s “high- iluc isk” classi ica ion.
2.2.2 | P e- Tax Fossil Fuel Subsidy Re o m (“P eT_
FFSR”)
In he i s policy scena io (“P eT_FFSR”), a e o m s a ing in
2023 is in oduced unde which gasoline and diesel p e- ax sub-
sidies a e linea ly phased ou in he egions wi h bio uel man-
da es ( egions in i alic in Table2). This e lec s a si ua ion whe e
coun ies mo i a e a e o m wi h he ambi ion o suppo hei
enewable anspo uel manda es/ a ge s o mo e b oadly a
enewable ene gy ansi ion. We dis ega d Uk aine, pa o he
FSU egion, and RUS who despi e ha ing a bio uel manda e can-
no be expec ed o implemen a ossil uel subsidy e o m gi en
he cu en geopoli ical con ex o he wa in Uk aine. As in he
baseline, global bio uel manda es a e included in he simula ion
o unde s and he in e ac ion be ween he wo policies. Regions
no lis ed in he able did no ha e p e- ax subsidies in 2021
based on he IMF da abase.
TABLE 2 | Mo o diesel and gasoline p e- ax subsidies in 2021.
Regions Mo o gasoline Mo o diesel
OPEC 35.2% 85.7%
Fo me So ie Union 0.3%
Russia 1.5%
PPEC (Pa ne s o
OPEC)
2.4% 9.3%
Oceania 0.03%
Res o he Middle Eas
and A ica
1.9% 8.4%
Res o Asia 0.02% 0.2%
Malaysia and
Indonesia
3.8% 23.9%
La in Ame ica 0.6% 2.1%
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
5 o 14
2.2.3 | Pos - Tax Subsidy Phase- Ou in he EU
(“PosT_FFSR”)
Bio uels a e no ee om he ex e nali ies included in he
IMF pos - ax subsidies es ima es. Only applying a policy in-
e nalizing he cos s o hese ex e nali ies o ossil uels could
he e o e be conside ed inconsis en . The di e ence in local
ai pollu ion and global wa ming (la ges sou ces o subsi-
dies in he da abase) be ween ossil- and bio- based anspo
uels is howe e subjec o unce ain y. GHG emissions no
being explici ly depic ed in he model, ins ead o only applying
he pos - ax subsidies o ossil uels we ely on he EU RED's
sus ainabili y c i e ia o explo e a “single egion” phase- ou
scena io.
The EU RED's sus ainabili y c i e ia s a e ha ood and eed
c op bio uels' GHG in ensi y should be a leas 50% lowe han
hei ossil uel al e na i e o coun owa d he manda e's a ge
(Di ec i e2018). We acco dingly assume ha hal o he pos -
ax subsidy a es on gasoline and diesel apply o bioe hanol and
biodiesel p oduced using hese eeds ocks in he egion and in
he second policy scena io (“PosT_FFSR”), a phase ou o pos -
ax subsidies on all ou uels is implemen ed in he EU s a -
ing in 2023. This complemen s he i s mul ila e al phase- ou
scena io wi h one whe e he coun ies implemen ing he e o m
ace a common bio uel manda e a ge (EU le el) bu di e en-
ia ed ossil uel subsidies (coun y le el). Double coun ing is
a oided as mo o gasoline and diesel do no en e he p oduc ion
o bio uels as in e media e.
3 | Resul s
3.1 | Go e nmen al Spending on FFS
P e- ax ossil uel subsidies end o be concen a ed in ne c ude
oil expo ing coun ies, and in 2020 we e made up o abou 31%
o gasoline and diesel subsidies. We simula e, in 2020, ha global
p e- ax gasoline and diesel subsidies amoun o 120 billion dol-
la s and each $135 billion in 2030. Solely conside ing egions
wi h bio uel manda es, he o e all alue o gasoline and diesel
p e- ax subsidies in he baseline scena io alls o $24.7 billion in
2030. The ela i e size o he subsidies in alue being consis en
wi h he ad- alo em a es (see Table2), Malaysia and Indonesia
(MAI) u n ou o be he egions wi h he la ges p e- ax sub-
sidies and a bio uel manda e, ollowed by he Middle Eas and
No he n A ica and PPEC. Diesel wi h i s highe ad- alo em
a e is ound o make up a ound 82% o o al p e- ax gasoline
and diesel subsidies be ween 2023 and 2030.
Pos - ax subsidies on gasoline and diesel a e la ge and mo e
widesp ead han p e- ax subsidies. In he EU whe e hey a e
phased ou in he second policy scena io (“PosT_FFSR”), we
simula e ha hey amoun o $92 billion in 2030. The egional
and uel a ia ions, depic ed in Table3, a e he esul o bo h he
ela i e size o he ad- alo em subsidies (see Da a S1) and he
alue o consump ion. As o p e- ax subsidies, diesel pos - ax
subsidies a e ound o be highe han hose on gasoline, and he
same can be said when compa ing biodiesel o bioe hanol. In
he whole EU egion in 2030, diesel makes up abou 77% o o al
pos - ax subsidies. This sha e is ela i ely cons an ac oss all EU
egions excep o Ge many whe e he sha e eaches 87% and
No h- Wes e n EU whe e i d ops o 56%.
3.2 | Fossil Fuels o T anspo
The simula ion esul s show ha , in egions ha emo e p e
o pos - ax ossil uel subsidies, p ices o gasoline and diesel
ise ela i e o he e e ence scena io. This leads o a educ ion
in consump ion la gely p opo ional o he alue o he subsi-
dies ha we e phased ou . In he i s policy scena io (“P eT_
FFSR”), egions phasing ou p e- ax subsidies expe ience a $24
billion (5%) decline in he o e all consump ion o ossil uels o
anspo by 2030 compa ed o he baseline (summing up alues
o gasoline and diesel, see Figu e1). In he second policy sce-
na io (“PosT_FFSR”), ollowing he emo al o pos - ax subsidy,
he EU consump ion o ossil uel o anspo d ops by $109
billion (18%).
In he es o he wo ld, howe e , he o al consump ion o gas-
oline and diesel inc eases by $3.8 billion and $6.6 billion in
scena ios 1 and 2, espec i ely. Fo diesel, his ype o consump-
ion leakage, caused by wha is some imes e e ed o as he
“ ossil- uel- p ice” leakage channel (Böh inge , Rosendahl, and
Schneide 2014), is bes explained by conside ing bila e al ade.
As displayed in Table4, egions phasing ou ossil uel subsidies
expo s and impo s o diesel o/ om he es o he wo ld sh ink
ollowing he p e- ax FFSR. This sugges s ha he g owing die-
sel consump ion ou side o he e o m a ea is no he esul o a
di e sion o domes ic consump ion o he expo ma ke om
he egions phasing ou hei subsidies. Ra he i is igge ed by
hese egions' dec easing impo demand. In o he wo ds, he
es o he wo ld abso bs on i s home ma ke some o he expo s
i no longe makes o egions implemen ing he e o m.
The same ade e ec s canno be he only sou ce o he g owing
gasoline consump ion since, as pe Table4, he ade o his uel
inc eases be ween egions inside and ou side he e o m a ea.
TABLE 3 | EU baseline pos - ax subsidies in 2030, in $ billion.
Cen al EU Ge many Eas e n EU Medi e anean No h- Wes e n EU To al
Bioe hanol 0.2 0.1 0.1 0.2 0.2 0.7
Biodiesel 1.0 1.1 0.2 1.1 0.3 3.7
Gasoline 4.4 1.8 1.7 4.6 4.2 16.7
Diesel 19.1 20.2 4.9 21.2 5.9 71.3
To al 24.6 23.2 6.9 27.1 10.5 92.3
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
6 o 14 GCB Bioene gy, 2025
Ra he han ade, we ind ha an expansion in gasoline p oduc-
ion o $2.5 billion in he es o he wo ld is la gely esponsible
o he g owing consump ion he e. This is especially e iden
in he US and China, he la ges p oduce s and consume s o
gasoline who oge he make up abou hal o his change. As
e lec ed in dec easing p oduc ion cos s o gasoline, his is he
esul o he dec ease in demand o c ude oil o p oduce die-
sel. The p oduc ion o diesel, which is mo e hea ily subsidized
han gasoline, declines e e ywhe e ollowing he e o m, ei he
di ec ly because o he domes ic e o m o due o he lowe de-
mand caused by e o ms ab oad. This pu s downwa d p essu e
on he p ice o c ude oil, he la ges inpu in he p oduc ion o
bo h diesel and gasoline. In u n, his leads gasoline o become
ela i ely mo e a ac i e, as i s p oduc ion cos s d op, and con-
sump ion g ows ou side o he e o m a ea.
The same mechanics apply o he consump ion leakage o gas-
oline and diesel in scena io 2. Diesel consump ion ou side he
EU inc eases mos ly om ade be ween egions ou side he EU
while gasoline consump ion g ows due o an expansion in p o-
duc ion. Again, we ind ha he US and China oge he make up
hal o he change in gasoline consump ion ou side o he e o m
a ea and ha he sha e o gasoline in hei o e all anspo uel
mix expands. In e es ingly, gasoline expo s om he es o
he wo ld o egions phasing ou hei subsidies inc ease in sce-
na io 1 while hey dec ease in scena io 2. This helps o explain
he $0.8 billion ise in gasoline consump ion whe e he p e- ax
subsidies a e phased ou in scena io 1 (see Figu e1). No only
is gasoline bene i ing om lowe p oduc ion cos s, bu gasoline
is also becoming ela i ely cheape han diesel as i s subsidies
a e no as la ge. In he MEA, o example, ollowing he p e- ax
FFSR, gasoline's own p ices ise by 1.8%, while ela i e o diesel,
i d ops by 7%. This oge he wi h he lowe p oduc ion cos s a
home and ab oad leads o an inc ease in p oduc ion, impo s,
and consump ion whe e he p e- ax subsidies a e emo ed. In
scena io 2, he same e ec is obse ed in Ge many whe e diesel
subsidies a e ela i ely la ge han gasoline. Gasoline p oduc-
ion, impo s, and consump ion ise he e, bu his is o se a he
o e all EU le el by a d op in he es o he EU egions.
3.3 | Bio uels
Bio uels a e di ec ly impac ed by he changes in he ossil ans-
po uel sec o . Since bio uel manda es' a ge s a e calib a ed as
a sha e o anspo uel consump ion, an inc ease/dec ease in
he demand o gasoline o diesel implies an inc ease/dec ease
in he amoun o bio uels needed o go e nmen s o each hei
a ge s. In egions implemen ing he e o m, his means ha
ypically while he p ice o bio uels ises, hei consump ion
d ops oge he wi h he go e nmen suppo needed o achie e
he manda e. Ou side o hese egions, bio uel p ices con ac ,
bu inc eased ossil uel consump ion leads o mo e bio uel con-
sump ion o mee he manda es and go e nmen spending on
bio uel subsidies o inc ease.
In he i s policy scena io (“P eT_FFSR”), we ind ha by
2030, he suppo om egions implemen ing a p e- ax FFSR o
bio uels dec ease by abou $6.2 billion (27%) compa ed o he
baseline, while a $0.6 billion inc ease is obse ed in he es o
he wo ld (Table5). In egions implemen ing he p e- ax FFSR,
his ansla es in o a 25- cen sa ing on bio uel suppo o each
dolla no spen on gasoline and diesel subsidy. The Malaysia
and Indonesia egion wi h he la ges subsidies on bo h ossil
and bio uels expe iences he mos change in consump ion and
subsidy le el. In he Res o he Middle Eas and A ica he need
o bio uel suppo ollowing he e o m comple ely disappea s.
While g owing i s sha e o bio uels in o al anspo uel by a
FIGURE 1 | Gasoline and diesel 2030 consump ion changes ela i e o he baseline, in $ billion.
TABLE 4 | Scena io 1 (“P e_FFSR”) 2030 changes in ade ela i e
o he baseline, in $ billion.
Mo o
gasoline Mo o diesel
FFSR RoW FFSR RoW
FFSR Impo s 0.1 0.3 (−2.9) (−10.7)
Expo s 0.0 (−0.9)
Row Impo s 0.0 0.1 (−0.9) 3.2
Expo s 0.3 (−10.7)
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
7 o 14
modes 0.2%, his ep esen s a 90% expansion in bio uel con-
sump ion and is enough o he egion o each i s a ge by 2030
wi hou he need o go e nmen in e en ions.
In he second policy scena io (“PosT_FFSR”), by 2030, he sup-
po om EU go e nmen s o bio uels dec ease by abou $6.8 bil-
lion (35%) compa ed o he baseline, while a $0.6 billion inc ease
is obse ed ou side he EU. In egions implemen ing he pos -
ax FFSR, his ansla es in o a 7- cen sa ing on bio uel suppo
o each dolla o gasoline and diesel subsidy phased ou . We
ind ha he pe cen age change in bio uel suppo depic ed in
Figu e2, is la ge han he pos - ax ad- alo em subsidy phased
ou on he co esponding uel. This is he esul o he combined
d op in bio uel consump ion needed o achie e he a ge and a
compe i i e e ec making bio uels ela i ely mo e a ac i e. In
he cen al EU egion, o e all bio uel suppo declined by 45%,
which is mo e han wice he size o he subsidies on diesel and
gasoline phased ou in he egion. This is explained by he ac
ha while consump ion sh inks a a a e compa able o he pos -
ax subsidy emo ed, he p ice o bio uels like biodiesel also ell
by 1.2% ela i e o hei ossil uel al e na i e (i.e., diesel).
We ind ha his is also wha explains he dec ease in bioe hanol
subsidy in Ge many. In scena io 2 (“PosT_FFSR”), he Ge man
consump ion o bioe hanol inc eases by 5% because he subsidy
on gasoline is many imes lowe han he one on diesel, and
also lowe han in any o he EU egion. Ye , he go e nmen 's
suppo o bioe hanol alls by 4% ollowing he phase- ou , sug-
ges ing ha he compe i i e e ec be ween ossil and bio uels
domina es o e he induced consump ion one. Indeed, no only
does he ela i e p ice o bioe hanol d ops by 0.3%, bu he sub-
sidy pe uni o bioe hanol consumed dec eases as well. Mo e
b oadly he same is ue o all egions ha phase ou he subsi-
dies on ossil uels. Bio uels become ela i ely mo e compe i i e
as hei ela i e p ice dec eases, leading o a lowe subsidy pe
uni o bio uel o each he manda es.
Ou side o he e o m a ea, he e ec is e e sed. While in abso-
lu e e ms bio uel p ices con ac , hey do so by less han ossil
uel p ices. As he p ice gap be ween bio uels and ossil uels
inc eases, he go e nmen mus expand suppo pe uni o bio-
uel consumed o each he manda es' a ge s. As a esul , while
he EU bio uel ma ke bene i s when he egion implemen s a
pos - ax FFSR, i su e s om ossil uel subsidy e o ms in o he
egions o he wo ld. Simila ly, coun ies like China, ha do no
implemen a subsidy e o m in ei he scena io, expe ience an in-
c ease in bio uel subsidies and ela i e p ice in bo h simula ions.
3.4 | Res o he Economy
Phasing- ou subsidies on gasoline and diesel igge s a educ ion
in he demand and p ice o c ude oil, he main inpu o hei
p oduc ion. Summing o e all egions phasing ou subsidies, we
ind ha in 2030 c ude oil consump ion sh inks by $15 billion
(3%) in he i s scena io and $32 billion (8%) in he second sce-
na io. Some o his is o se by a ise in consump ion in he es
o he wo ld o $0.7 billion and $7 billion in he wo scena ios,
espec i ely, leading o a global educ ion in c ude oil consump-
ion o $14 billion (0.4%) and $25 billion (0.7%), espec i ely. In
TABLE 5 | 2030 Change in bio uel subsidies ela i e o baseline, in $ billion.
Policy scena io 1 (P eT_FFSR) Policy scena io 2 (PosT_FFSR)
FFSR eg RoW Wo ld EU Non- EU Wo ld
Biodiesel (−6.2) 0.2 (−6.1) (−5.4) 0.2 (−5.2)
Bioe hanol 0.01 0.4 0.4 (−1.3) 0.4 (−1.0)
To al (−6.2) 0.6 (−5.6) (−6.8) 0.6 (−6.2)
FIGURE 2 | Scena io 2 (“PosT_FFSR”) 2030 changes in EU bio uels subsidies ela i e o baseline, in $. *Biodiesel om all sou ces.
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
8 o 14 GCB Bioene gy, 2025
u n, consump ion o pe oleum and coke p oduc s, which a e
p oduced using he now cheape c ude oil, inc eases in and ou
o he e o m a ea. This o wa d linkage is simula ed in 2030 a
$20 billion (0.8%) globally in he i s scena io and a $54 billion
(2.3%) in he second. In compa ison, d i en by COVID- 19 es ic-
ions, he inal consump ion o oil and pe oleum p oduc s o
ene gy pu poses in he EU ell by 10% in 2020 (Eu os a 2022).
Backwa d and o wa d linkages a e also obse ed in he ag icul-
u al sec o whe e he p ice and demand o bio uel eeds ocks
end o decline oge he wi h he consump ion o he bio uels. In
scena io 1, his is especially e iden o he palm oil sec o whose
consump ion d ops by $1.6 billion in he Malaysia and Indonesia
egion. In scena io 2, he EU egion no ably dec eases i s de-
mand o apeseed oil and whea /co n which in he egion a e
he main eeds ocks in he p oduc ion o biodiesel and bioe h-
anol. As consump ion o oil seeds and g ain o bio uel p o-
duc ion sh inks, he supply o by- p oduc s con ac s, and hei
p ices inc ease by up o 8% in 2030 in he EU egions which pu s
upwa d p essu e on he p ice o he li es ock sec o s (Table6).
These linkages ha e a bea ing on he e o ms' o e all impac on
iscal e enue. No only does he dec easing consump ion o gas-
oline and diesel e ode iscal e enue om he p e- exis ing axes
bu changes in he consump ion o o he ene gy commodi ies
as well as o commodi ies in o he sec o s impac go e nmen
e enues. Conside ing all changes in p e- exis ing ax e enues
and a oided spending on subsidies, we simula e o e all iscal
e enue o inc ease in bo h scena ios whe e he subsidies a e
phased ou as well as in he es o he wo ld. Whe e he FFSRs
a e implemen ed we ind ha go e nmen e enue inc eases by
1.9% and 1.2% in scena ios 1 and 2, espec i ely. Figu e3 b eaks
down he sou ce o he inc ease in iscal e enue whe e he sub-
sidies a e phased ou .
Mos o he inc ease in iscal e enue in egions implemen ing
he e o ms is ound o be he esul o a oided spending on
ossil uel subsidies. O e all, howe e , go e nmen e enue in
hese egions also inc eases by 0.37% and 0.14% om a oided
spending on bio uel suppo in scena ios 1 and 2, espec i ely.
This co esponds o he inc ease in go e nmen e enue being
25% and 17% highe when he impac o ossil uel subsidy e-
o m on bio uel suppo s is conside ed as pa o he changing
go e nmen budge . We no e ha while e enue om p e-
exis ing ax sh inks in scena io 2, i does no do so in a isible
manne in scena io 1. This is he esul o p e- exis ing axes
on gasoline and diesel in egions wi h p e- ax ossil uel sub-
sidies being ela i ely small, especially compa ed o he ones
in he EU.
Change o 2030 GDP emains below 1% in all scena ios and e-
gions (Table7). Tha being said, egions phasing ou subsidies
end o expe ience a d op in hei GDP. The e ec is especially
p onounced in scena io 2 o he EU which phases ou ela i ely
TABLE 6 | Scena io 2 (“PosT_FFSR”) 2030 apeseed (oil) and by- p oduc s consump ion change in he EU ela i e o he baseline.
Cen al EU Ge many Eas e n EU Medi e anean No h- Wes e n EU
Rapeseed P ice (−0.1%) (−0.3%) (−0.3%) (−0.1%) (−0.2%)
Quan i y (−9.8%) (−7.8%) (−5.7%) (−6.0%) (−6.5%)
Rapeseed oil P ice (−1.7%) (−3.4%) (−1.1%) (−2.2%) (−1.6%)
Quan i y (−9.0%) (−23.0%) (−1.9%) (−5.1%) (−1.1%)
Rapeseed meal P ice 8.4% 6.7% 4.8% 4.9% 4.8%
Quan i y (−13.0%) (−10.9%) (−8.1%) (−7.8%) (−7.3%)
FIGURE 3 | Composi ion o he change in iscal e enue whe e he subsidies a e phased ou , in 2030.
17571707, 2025, 2, Downloaded om h ps://onlinelib a y.wiley.com/doi/10.1111/gcbb.70019, Wiley Online Lib a y on [07/08/2025]. See he Te ms and Condi ions (h ps://onlinelib a y.wiley.com/ e ms-and-condi ions) on Wiley Online Lib a y o ules o use; OA a icles a e go e ned by he applicable C ea i e Commons License
