Growth-enhancing taxes

de Padua, Da id; Kiocho, Mae Hyacin h; Pa k, Donghyun
Wo king Pape
G ow h-enhancing axes
ADB Economics Wo king Pape Se ies, No. 727
P o ided in Coope a ion wi h:
Asian De elopmen Bank (ADB), Manila
Sugges ed Ci a ion: de Padua, Da id; Kiocho, Mae Hyacin h; Pa k, Donghyun (2024) : G ow h-
enhancing axes, ADB Economics Wo king Pape Se ies, No. 727, Asian De elopmen Bank (ADB),
Manila,
h ps://doi.o g/10.22617/WPS240303-2
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/299299
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
GROWTH-ENHANCING
TAXES
Da id de Padua, Mae Hyacin h Kiocho, and Donghyun Pa k
ADB ECONOMICS
WORKING PAPER SERIES
NO. 727
May 2024
G ow h-Enhancing Taxes
Tax e enues ha e a pe sis en posi i e impac on g ow h, and he associa ion is especially p onounced
in eme ging economies. Mac oeconomic and ins i u ional ac o s such as low in la ion and s ong
go e nance ein o ce he g ow h-enhancing e ec o axes, bu hese esul s a e condi ional on he income
le el o he economy. The indings o his s udy imply ha he e ec o axes on g ow h should be e alua ed
wi hin mac oeconomic and s uc u al cons ain s.
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 68 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.
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
G ow h-Enhancing Taxes
Da id de Padua, Mae Hyacin h Kiocho,
and Donghyun Pa k
No. 727 | May 2024
Mae Hyacin h Kiocho ([email p o ec ed])
is a consul an a he Sou heas Asia Depa men ,
Asian De elopmen Bank (ADB). Da id de Padua
([email protected] g) is an economics o ice and
Donghyun Pa k (dpa [email protected] g) is an economic
ad iso a he Economic Resea ch and De elopmen
Impac Depa men , ADB.
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. WPS240303-2
DOI: h p://dx.doi.o g/10.22617/WPS240303-2
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
oano 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
oob 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
heADB logo.
Co igenda o ADB publica ions may be ound a h p://www.adb.o g/publica ions/co igenda.
No e:
ADB ecognizes “Tu key” as Tü kiye.
ABSTRACT
We in es iga e he condi ions unde which ax e enues can enhance economic g ow h.
Using a newly cons uc ed da ase consis ing o 135 economies and spanning he pe iod
1990–2019, we s udy how changes in ax e enues impac economic g ow h using a
panel ec o au o eg ession (PVAR) model. Tax e enues ha e a pe sis en posi i e
impac on g ow h, and he associa ion is especially p onounced in eme ging economies.
S ic in la ion a ge ing, low-in la ion, lexible exchange a es, a mo e de eloped inancial
sec o , highe in es men a es, and s ong go e nance ein o ce he g ow h-enhancing
e ec o axes, bu hese esul s a e condi ional on he income le el o he economy. Ou
indings imply ha he e ec o axes on g ow h should be e alua ed wi hin
mac oeconomic and s uc u al cons ain s.
Keywo ds: axes, g ow h
JEL code: H20

1. In oduc ion
Tax e enues a e essen ial o economic de elopmen . Taxes p o ide go e nmen s wi h
he esou ces o und heal hca e, educa ion, in as uc u e, and o he g ow h-p omo ing
public goods. Taxes can also be a means o achie e o he policy objec i es, such as
edis ibu ing income, imp o ing he e iciency o ma ke s, and in luencing he beha io o
socie y. Howe e , i is well-known ha axes c ea e dis o ions and p oduce deadweigh
losses. While highe public spending inanced by highe axes can enhance g ow h, he
dis o ions om highe axes can s i le g ow h. The ela ionship may depend on he
income le el o he economy gi en di e ing policy and s uc u al cons ain s.
This pape aims o con ibu e o he exis ing li e a u e by examining whe he
a ious mac oeconomic policies and s uc u al ac o s in luence he impac o axes on
g ow h. Taxes a ec ing economic g ow h and economic g ow h a ec ing axes
complica e he es ima ion o his ela ionship.1 We add ess his issue by employing a
panel ec o au o eg ession (PVAR) model, which allows o endogenei y among
a iables. In addi ion, we analyze subsamples o di e en economy g oups (e.g.,
ad anced and eme ging economies), ollowing Qu eshi and Liaqa (2020) and Lo and
Malinen (2014). Qu eshi and Liaqa (2020) examine he ela ionship be ween ex e nal
deb and economic g ow h o economy g oups a di e en income le els. Simila ly, Lo
and Malinen (2014) s udy he link be ween so e eign deb and economic g ow h o
economy g oups wi h di e en le els o deb – o–g oss domes ic p oduc (GDP) a io. In
his s udy, we in oduce mac oeconomic policies, s uc u al ac o s, and he
1 Acco ding o A nold (2008), “In an addi ional se o obus ness checks, an a emp is made o con ol o
he ac ha mos o he ax indica o s used in he analysis a e de i ed om Re enue S a is ics and om
Na ional Accoun s. This could lead o an endogenei y bias inso a as ax e enues inc ease in expansions
and decline in ecessions, e en hough sho - un dynamics a e accoun ed o in he eg essions.”
2
mac oeconomic en i onmen o analyze whe he hese ac o s in luence he ela ionship
be ween axes and g ow h. Such an analysis can help iden i y he condi ions in which
axes acili a e g ow h. This can help in o m go e nmen s a egies, which op imize he
impac o axes on g ow h.
Se e al s udies, which ind ha ax inc eases ha m g ow h (Gemmell e al. 2011,
A nold e al. 2011, Alesina and A dagna 2010), ocus on O ganisa ion o Economic Co-
ope a ion and De elopmen (OECD) coun ies. The empi ical indings o his pape
sugges ha he impac o ax e enues o economic g ow h is no s aigh o wa d. Mo e
speci ically, he ax–g ow h ela ionship can be in luenced by mac oeconomic policies,
s uc u al ac o s, and he mac oeconomic en i onmen . These ac o s help de e mine he
impac o ax e enues on g ow h. The baseline esul s show ha ax e enues ha e a
posi i e impac on g ow h. The ela ionship di e s o ad anced and eme ging
economies. Taxes ha e a nega i e e ec on g ow h in he o me bu po en ially posi i e
e ec in he la e .
The empi ical esul s con i m ha mac oeconomic policies, s uc u al ac o s, and
mac oeconomic ac o s in luence he ax–g ow h nexus. Fo ad anced economies, axes
become g ow h- iendly unde a s ic in la ion- a ge ing scheme and subs an ial le el o
in es men s. Ac oss he exchange a e egimes and he le els o in la ion, he ad e se
e ec o axes on g ow h is s ill obse ed. Howe e , he e a e some di e ences in he
pe sis ence and magni ude o he impac . The nega i e shock is less pe sis en unde a
lexible exchange a e egime ela i e o a ixed exchange a e egime, and he nega i e
impac is g ea e a high le els o in la ion. Fo eme ging economies, axes ha m g ow h
unde a s ic in la ion- a ge ing scheme and a high le els o in la ion. In he ull sample,
3
axes hu g ow h when he inancial sec o is less de eloped. This sugges s ha a well-
de eloped inancial sec o enhances he g ow h-p omo ing e ec s o axes. Finally, s ong
go e nance is i al o a signi ican posi i e impac o axes on g ow h.
The esul s o his pape imp o e ou unde s anding o he ela ionship be ween
axes and g ow h. We ind ha he ela ionship depends on mac oeconomic and s uc u al
ac o s, which can ei he ampli y o dilu e he g ow h e ec o axes. Tha is, axes on
g ow h should no be iewed alone bu wi h mac oeconomic and s uc u al ac o s.
The exis ing empi ical e idence is ambiguous. Knelle e al. (1999) ind ad e se
e ec s o some axes on g ow h. The s udy classi ies axes in o dis o iona y and non-
dis o iona y axes and ind ha dis o iona y axes lowe g ow h while non-dis o iona y
axes do no . Angelopoulos e al. (2007) and A nold (2008) show ha he ela ionship
depends on he ax s uc u e. Labo income ax is nega i ely associa ed wi h g ow h. On
he o he hand, p ope y ax, consump ion ax, and pe sonal income ax a e mo e g ow h-
iendly. Meanwhile, he link be ween co po a e income ax and g ow h di e s in he wo
s udies. Angelopoulos e al. (2007) ind co po a e income ax o be posi i ely ela ed o
g ow h, while A nold (2008) inds a la ge nega i e associa ion. O he s udies obse e a
non-linea ela ionship be ween axes and g ow h. Gaspa e al. (2016) es ima e a ipping
poin o he ax- o-GDP a io ha would speed up g ow h and de elopmen . Economies
whose ax- o-GDP a io is abo e 12.75% ha e a GDP pe capi a ha is 7.5% g ea e han
economies whose ax- o-GDP a io is below 12.75%. Fo Eu opean economies, Esen and
Aydin (2019) simila ly es ima e a h eshold le el o axes ha os e s economic g ow h.
4
O he ac o s can in luence he ela ionship be ween axes and g ow h. Fo
example, Phuc Canh (2018) shows ha he e ec o iscal policy on g ow h a ies among
eme ging economies because o di e ences in ins i u ions and ex e nal deb le els. The
s udy inds ha high deb weakens he posi i e impac o iscal policy on g ow h, while
be e ins i u ions magni y he posi i e impac . Go e nance- ela ed ac o s, such as
co up ion, accoun abili y, ule o law, and economic eedom, a e obse ed o in luence
ax e o s (Bi d e al., 2008) and a ec he main impac o axes on g ow h (Baldacci e
al., 2004). Fu he mo e, he bu den o axes on g ow h can be in luenced by inancial
sec o de elopmen s, such as be e c edi in o ma ion sha ing and g ea e inancial
access, which educe ax e asion (Beck e al., 2014).
The es o he pape is o ganized as ollows: Sec ion 2 explains he PVAR
app oach and desc ibes he panel da a. Sec ion 3 p esen s he baseline PVAR es ima ion
esul s o he ax–g ow h ela ionship. Sec ion 4 examines he impac o mac oeconomic
policies, s uc u al ac o s, and he mac oeconomic en i onmen . Sec ion 5 in es iga es
he e ec o axes on in es men , an impo an channel h ough which axes can a ec
g ow h. Sec ion 6 discusses h esholds and Sec ion 7 concludes he pape .
2. Me hodology and Da a
2.1. Panel Vec o Au o eg ession App oach
We used he PVAR model o acili a e ou analysis o he ela ionship be ween axes and
g ow h. An ad an age o he PVAR model is ha i assumes all a iables o be
endogenous and in e dependen . In addi ion, i has a c oss-sec ional dimension, which
akes he c oss-sec ional he e ogenei y in o accoun .
11
ha e an in luence on how axes a ec economic de elopmen . In employing such policies,
he go e nmen s seek o achie e c edibili y, lexibili y, o s abili y o imp o e
mac oeconomic pe o mance and esilience o shocks (Be ge e al. 2000, Kopi s 2001,
Ay es e al. 2014). Se e al s udies ha e shown ha in la ion and g ow h pe o mance
a y ac oss he ypes o in la ion- a ge ing schemes, exchange a e egime, and iscal
ules (Ghosh e al. 1997, Le y-Yeya i and S u zenegge 2003, Bleaney and F ancisco
2007, Gonçal es and Salles 2008, Mollick e al. 2011, A onso and Jalles 2013, Ay es e
al. 2014, G embi e al. 2016). We explo e whe he he choice o mac oeconomic policies
also ma e s o he ela ionship be ween axes and g ow h.
4.1.1. In la ion Ta ge ing
Findings om se e al s udies ha e shown ha he adop ion o an in la ion- a ge ing
scheme can lead o di e en ia ion o mac oeconomic pe o mance ac oss economies.
The economies ha adop ed in la ion a ge ing a e obse ed o ha e lowe in la ion,
educed ou pu a iabili y, and highe income pe capi a as in la ion a ge ing ins ills
s abili y, which allows economies o be e wi hs and c ises (Gonçal es and Salles 2008,
Mollick e al. 2011, Ay es e al., 2014). In addi ion, Ay es e al. (2014) ind ha e ec s o
an in la ion- a ge ing scheme on in la ion and g ow h a y ac oss egions. A g oup o
Middle Eas e n, No h A ican, Sou he n Eu opean, and Eas e n Eu opean economies
expe ienced lowe in la ion a es and sho - e m economic g ow h a e adop ing in la ion
a ge ing. On he o he hand, Asian, sub-Saha an A ican, and Oceanic economies ha
adop ed in la ion a ge ing expe ienced a ise in in la ion and no subs an ial changes in
economic g ow h. Fo ad anced economies, Ball and She idan (2004) ind no di e ence
in he economic pe o mance o a ge ing e sus non- a ge ing economies. Meanwhile,

12
Gonçal es and Salles (2008) show ha eme ging economies ha adop ed in la ion
a ge ing enjoyed lowe in la ion a es and smalle ou pu a iabili y. Since he adop ion
o in la ion a ge ing can a ec aspec s o economic pe o mance, he e is a possibili y
ha i can also in luence he ela ionship be ween axes and g ow h.
The agg ega e sample is g ouped based on whe he he implemen ed in la ion-
a ge ing scheme is s ic o no . Fo all economies, he impac o ax e enues on g ow h
is posi i e ega dless o he in la ion- a ge ing scheme (Figu e 2). The di e ence be ween
he wo in la ion- a ge ing schemes is obse ed in he du a ion o he e ec , wi h a mo e
pe sis en impac associa ed wi h a loose in la ion- a ge ing scheme. Fo ad anced
economies, a nega i e e ec o axes on g ow h is associa ed wi h a loose in la ion-
a ge ing scheme, while a posi i e impac is no ed wi h a s ic in la ion- a ge ing scheme.
The opposi e is obse ed o eme ging economies— he posi i e impac o ax e enues
on g ow h is seen wi h a loose in la ion- a ge ing scheme, while a nega i e impac on
g ow h is associa ed wi h a s ic in la ion- a ge ing scheme.
13
Figu e 2: In la ion-Ta ge ing Scheme
(a) Loose In la ion-Ta ge ing Scheme (b) S ic In la ion-Ta ge ing Scheme
These indings a e consis en wi h exis ing li e a u e. Luco e (2012) inds ha he
adop ion o in la ion- a ge ing scheme imp o es ax collec ion in eme ging economies.
The s udy uses a p opensi y sco e ma ching app oach and inds ha o al public e enues
a e signi ican ly posi i e and la ge in magni ude in eme ging economies ha adop ed
No e: Impulse esponse unc ions o pe capi a g oss domes ic p oduc (GDP) g ow h o a shock in he ax
as a pe cen age o GDP we e ob ained om he panel ec o au o eg ession (PVAR) o all coun ies,
ad anced economies, and eme ging economies, as well as ac oss he ypes o in la ion- a ge ing schemes
(i.e., s ic and loose). The shaded a ea ep esen s he 90% con idence in e als based on 200 Mon e Ca lo
simula ions.
Sou ce: Au ho s’ calcula ions.
14
in la ion a ge ing. A simila app oach is used by Kazemi e al. (2020) in in es iga ing he
impac o in la ion a ge ing on di ec axes and dis inguishing he impac be ween oil-
impo ing and oil-expo ing economies. They obse e ha he adop ion o in la ion
a ge ing inc eases di ec axes o oil impo e s, while he e is no impac on he di ec
axes among oil expo e s. Meanwhile, Gal is Ci o and Fe ei a de Mendonça (2016)
ela e he success in achie ing he in la ion a ge o he epu a ion o he cen al bank
and examine he e ec o he epu a ion on ax e o . Thei esul s show ha he
epu a ion o he mone a y au ho i y causes a highe ax e o in Colombia.
Pas s udies ela ing mone a y policy o iscal policy discuss ha he adop ion o
an in la ion- a ge ing amewo k can lead o mo e disciplined iscal policies (Luco e 2012)
and p omo e ins i u ional quali y (Minea e al. 2021), which would imp o e ax collec ions.
The adop ion o in la ion a ge ing cons ains he go e nmen in using seignio age ax o
sou ce e enues. Thus, he go e nmen is o ced o imp o e ax collec ion e o s om
o he sou ces o e enues o ecoup he losses om seignio age ax. Ou esul s indica e
ha s ic in la ion a ge ing is associa ed wi h a posi i e e ec o axes on g ow h in
ad anced economies while loose in la ion a ge ing is associa ed wi h a posi i e impac
in eme ging economies. This di e ence may be ela ed o he c edibili y and hus
e ec i eness o in la ion a ge ing, which, in u n, a ec s he impac o axes on g ow h.
S ic in la ion a ge ing may be mo e c edible in ad anced economies, while loose
in la ion a ge ing may be mo e c edible in eme ging economies. S ic a ge ing is iewed
as less c edible in eme ging economies, which depend mo e hea ily on seignio age as a
sou ce o go e nmen e enues (Cukie man e al. 1992). On he o he hand, loose
15
a ge ing may be seen as a lack o commi men o in la ion a ge ing in ad anced
economies.
4.1.2. Exchange Ra e Regime
Pas s udies obse ed ha mac oeconomic pe o mance o economies di e s ac oss
di e en exchange a e egimes. In de eloping economies, Bleaney and F ancisco (2007)
ind ha lowe in la ion and slowe g ow h a e associa ed wi h ha d pegs, and ha loa
and so pegs ha e simila g ow h a es.2 Howe e , loa pegs ha e sligh ly highe in la ion
compa ed o so pegs. Fo Ghosh e al. (1997), lowe in la ion and less a iabili y a e
obse ed alongside ixed exchange a e egimes. Though ixed exchange a e egimes
a e also associa ed wi h highe in es men s, hey a e seen o ha e slowe p oduc i i y
g ow h. Hence, ou pu g ow h is highe wi h lexible exchange a e egimes. On he o he
hand, he indings o Le y-Yeya i and S u zenegge (2003) p esen ha exchange a e
egimes do no a ec g ow h o indus ial economies. Bu o de eloping economies,
exchange a e egimes ha e a signi ican impac on g ow h, whe ein slowe g ow h and
g ea e ou pu ola ili y a e associa ed wi h less lexible exchange a e egimes. As such,
i is o in e es o examine i exchange a e egimes ha e an in luence on how axes may
impac economic g ow h.
The da a o Ilze zki e al. (2019, 2022) classi y he di e en exchange a e
egimes in o six ca ego ies, as shown in Table 1.
2 Bleaney and F ancisco (2007) g oup obse a ions in o h ee classi ica ion schemes: loa pegs, so pegs,
and ha d pegs. Floa pegs desc ibe hose ha a e ee, managed, o di y loa s. So pegs a e he egimes
ha a e ca ego ized in he in e media e ca ego y, such as c awling pegs o bands.
16
Table 1: Coa se Classi ica ion o Exchange Ra e Regimes
Ca ego y
Desc ip ion
1  No sepa a e legal ende
 P e-announced peg o cu ency boa d a angemen
 P e-announced ho izon al band ha is na owe han o equal o (+/-)
2.0%
 De ac o peg
2  P e-announced c awling peg
 P e-announced c awling band ha is na owe han o equal o (+/-)
2.0%
 De ac o c awling peg
 De ac o c awling band ha is na owe han o equal o (+/-) 2.0%
3  P e-announced c awling band ha is wide han o equal o (+/-) 2.0%
 De ac o c awling band ha is na owe han o equal o (+/-) 5.0%
 Mo ing band ha is na owe han o equal o (+/-) 2.0% (i.e., allows o
bo h app ecia ion and dep ecia ion o e ime)
 Managed loa ing
4  F eely loa ing
5  F eely alling
6  Dual ma ke in which pa allel ma ke da a is missing
Sou ces: Ilze zki, E han, Ca men M. Reinha , and Kenne h S. Rogo . 2019. “Exchange A angemen s
En e ing he Twen y-Fi s Cen u y: Which Ancho will Hold?” The Qua e ly Jou nal o Economics, 134 (2):
599–646; Ilze zki, E han, Ca men M. Reinha , and Kenne h S. Rogo . 2022. “Re hinking Exchange Ra e
Regimes.” In Handbook o In e na ional Economics 6, edi ed by Gi a Gopina h, Elhanan Helpman, and
Kenne h Rogo , 91–145. No h Holland: Else ie .
We sepa a e he sample in o wo g oups based on wo ypes o exchange a e
egime: ixed and lexible. The a angemen s ca ego ized as less han 3 based on he
classi ica ion abo e is g ouped in o he ixed exchange a e egime.3 On he o he hand,
he a angemen s wi h g ea e han o equal o 3 a e g ouped unde he lexible exchange
a e egime.
Fo all economies, he posi i e impac o ax e enues on g ow h ha we ha e
seen in he baseline esul s is only p esen among hose wi h a lexible exchange a e
3 Ilze zki e al. (2019) obse e ha a angemen s ha a e less lexible han managed loa ing ha e a low
deg ee o exchange a e a iabili y.

17
egime (Figu e 3). The e ec o ax e enues on g ow h o ad anced economies does
no di e be ween he wo ypes o exchange a e egime. A nega i e e ec o ax
e enues on g ow h is obse ed o bo h exchange a e egimes. The impac is ansi o y
o he lexible exchange a e egime, wi h he e ec pe sis ing o only 1 yea a e he
shock. The ad e se e ec o axes on g ow h associa ed wi h a ixed exchange a e
egime is mo e pe sis en , las ing o 4 yea s a e he shock. Fo he eme ging
economies, he e ec o ax e enues on g ow h a ies be ween he wo. A nega i e
impac on g ow h is associa ed wi h a ixed exchange a e egime, bu he impac is no
signi ican . The e ec o ax e enues on g ow h is posi i e o hose wi h a lexible
exchange a e egime.
18
Figu e 3: Exchange Ra e Regime
Fixed Exchange Ra e Regime Flexible Exchange Ra e Regime
A lexible exchange a e egime seems o acili a e axes o posi i ely a ec g ow h
o eme ging economies; while o ad anced economies, a lexible exchange a e egime
lessens he pe manence o he nega i e impac o axes on g ow h. The esul is aligned
wi h p e ious s udies, which ound a posi i e link be ween lexible exchange a e egime
and economic g ow h. Le y-Yeya i and S u zenegge (2003) ind ha lowe g ow h and
g ea e ou pu ola ili y a e associa ed wi h ixed exchange a e egime, pa icula ly o
No e: Impulse esponse unc ions o pe capi a g oss domes ic p oduc (GDP) g ow h o a shock in he ax
as a pe cen age o GDP we e ob ained om he panel ec o au o eg ession (PVAR) o all coun ies,
ad anced economies, and eme ging economies, as well as ac oss he ypes o exchange a e egimes (i.e.,
ixed and lexible). The shaded a ea ep esen s he 90% con idence in e als based on 200 Mon e Ca lo
simula ions.
Sou ce: Au ho s’ calcula ions.
19
de eloping economies. They explain ha , in he e en o shocks, he cons ained
adjus men s in exchange a es and p ices unde a ixed exchange a e egime may esul
in p ice dis o ions and misalloca ion o esou ces, which can cause highe ou pu
ola ili y. The in lexibili y o mac oeconomic adjus men s, especially in imes o
unce ain y, would ha e nega i e consequences on g ow h. Se e al s udies ela e he
deg ee o mac oeconomic adjus men s b ough by he di e en exchange a e egimes o
he go e nmen ’s iscal discipline. In he model o To nell and Velasco (2000), a lexible
exchange a e egime compels he go e nmen o be mo e disciplined, since
consequences o unsound iscal policies a e immedia ely ealized ia exchange a e
mo emen s. Compa ed o a ixed a e egime, he consequences o imp uden iscal
policies a e delayed and mani es ed when he si ua ion is al eady unsus ainable o when
he e en ual collapse o he peg is al eady i e e sible, which can be poli ically cos ly o
he go e nmen . Thei empi ical exe cise using sub-Saha an A ican economies ma ches
he implica ions o hei model whe ein he choice o he exchange a e egimes a ec s
he deg ee o iscal adjus men s. Jalles e al. (2016) also ind ha (i) ixed exchange a e
egime is associa ed wi h less iscal discipline and (ii) lexible exchange a e egime
in ensi ies he posi i e e ec s o s ong poli ics on iscal pe o mance. They sugges ha ,
in a way, he lexible exchange a e egime c ea es an en i onmen ha helps os e iscal
discipline.
The posi i e impac o axes on g ow h o eme ging economies and he less
pe manen nega i e ad e se e ec o axes on g ow h o ad anced economies obse ed
in he lexible exchange a e egime may be a esul o he iscal discipline ha he lexible
exchange a e egime acili a es. Fiscal discipline is ela i ely g ea e in eme ging
20
economies, which end o ace highe isks o iscal sus ainabili y. The associa ion o
lexible exchange a e egime wi h g ea e iscal discipline sugges s ha sound iscal
policies a e impo an ansmission mechanisms h ough which he impac o axes on
g ow h is enhanced in economies wi h lexible exchange a es.
Ano he impo an mechanism may be he well-known ole o lexible exchange
a es as shock abso be s. Tha is, lexible exchange a es cushion economies agains
ex e nal shocks, enabling he go e nmen s o p io i ize g ow h-p omo ing expendi u es
such as heal h, educa ion, and in as uc u e as he objec i e o iscal policy. This
endency is likely o be mo e p e alen in eme ging economies, whe e economic g ow h
is a ela i ely mo e signi ican o e all policy objec i e han in ad anced economies.
4.1.3. Fiscal Rules
Fiscal ules se a cons ain on budge a y agg ega es, o which axes a e a majo
componen . Since axes can be di ec ly a ec ed by he iscal ules, hese can in u n
in luence he impac o axes on g ow h. Some s udies ha e shown ha iscal ules a e
e ec i e in imp o ing iscal balances. G embi e al. (2016) conclude ha iscal ules a e
impo an in es aining public deb , and hey ind ha elaxa ion o iscal ules leads o
lowe ax a es and lowe ax e enues in uncons ained ci ies in I aly. Caselli and
Reynaud (2020) ind ha p esence o iscal ules is no necessa ily associa ed wi h lowe
de ici s, bu iscal ules should be well-designed in o de o ha e a posi i e and signi ican
impac on iscal balance. O he s udies ha e also p esen ed ha he e ec o iscal ules
on iscal balances ha e consequences o g ow h. The a e age g ow h o Eu opean
economies is seen o be s a is ically highe o he pe iod a e he Maas ich T ea y, in
which he 3% o GDP de ici ule was assessed (Cas o 2011). A onso and Jalles (2013)
27
cos o he socie y’s ac i i ies.4 No h (1990) s a es, “Thi d wo ld coun ies a e poo
because he ins i u ional cons ain s de ine a se o payo s o poli ical/economic ac i i y
ha do no encou age p oduc i e ac i i y.” A smalle a enue o esea ch ela ed o
ins i u ions and economic pe o mance ocuses on he link o go e nmen , ins i u ions,
and economic g ow h. The go e nmen is among socie y’s playe s, so i s ac ions and
hei consequences a e cons ained by ins i u ions. In he model o Chaudh y and Ga ne
(2007), hey p esen ha en -seeking ac i i ies o he go e nmen , which ha e an
ad e se e ec on g ow h, a e supp essed by a be e ins i u ional en i onmen
ep esen ed by a highe cos o adop ing g ow h- educing policies aced by he
go e nmen . A onso and Jalles (2016) a gue ha go e nmen s u ilize a subs an ial sha e
o he economy’s esou ces so ha ac ions o he go e nmen exe an impac on he
g ow h o he economies and ha he e ec on g ow h depends on he quali y o
ins i u ions. They ound ha la ge go e nmen s esul in nega i e g ow h and ha he
nega i e impac on g ow h is la ge among hose wi h low le els o ins i u ional quali y.
Meanwhile, Phuc Canh (2018) shows ha he e ec i eness o iscal policy in acili a ing
economic g ow h is a ec ed by he di e ences in ins i u ions in eme ging economies. The
s udy inds ha he in e ac ion e ms o go e nmen expendi u e wi h he ins i u ional
indica o s pos a posi i e impac on g ow h, which implies ha imp o emen in ins i u ions
enhances he e ec i eness o iscal policy in p omo ing g ow h.
The quali y o go e nance and ins i u ions is impo an o he p o-g ow h e ec o
axes and o he iscal policies (A in e al. 2021, I anyna and Sale no 2021). In line wi h
Phuc Canh (2018), who inds ha go e nance and ins i u ional quali y in luence he
4 No h (2008) p esen s ha o mal ules, in o mal cons ain s, and cha ac e is ics ela ed o he
en o cemen o he cons ain s cons i u e ins i u ions.

28
e ec i eness o iscal policy, ou s udy obse es ha he baseline impac o axes on
g ow h is isible only when he le el o go e nance is high. A a b oade le el, ou esul s
sugges ha good go e nance and ins i u ions a e i al o he ans o ma ion o ax
e enues in o p oduc i e and e icien go e nmen spending, which, in u n, ampli y hei
impac on g ow h. On he o he hand, bad go e nance and ins i u ions ha m he quali y
o go e nmen spending.
4.2.2. Financial Sec o De elopmen
To cha ac e ize he inancial sec o de elopmen , we use he inancial de elopmen
index. The index is no malized so ha alues ange om 0 o 1, whe e highe alues
indica e g ea e inancial de elopmen . The sample is ca ego ized in o wo g oups, hose
wi h a less de eloped inancial sec o and hose wi h a mo e ad anced inancial sec o .
We de ine he g oups based on he median alue o he inancial de elopmen index. The
obse a ions wi h alues less han he median a e g ouped in o he less de eloped
inancial sec o , while hose wi h alues g ea e han o equal o he median is classi ied
unde he mo e ad anced inancial sec o g oup.
The e ec o ax e enues on g ow h di e s be ween he wo g oups based on he
inancial sec o de elopmen (Figu e 7). Fo all economies, a less de eloped inancial
sec o is associa ed wi h ax e enues ha ing a nega i e impac on g ow h. Meanwhile,
he e ec o ax e enues on g ow h is posi i e when he inancial sec o is mo e
de eloped. The exe cise canno be done o ad anced economies since all o hem ha e
ad anced inancial sec o s. Fo eme ging economies, hose wi h less-de eloped inancial
sec o s show a nega i e impac o axes on g ow h.
29
Figu e 7: Financial De elopmen Index
Less De eloped Financial Sec o Mo e Ad anced Financial Sec o
A posi i e (nega i e) impac o axes on g ow h when he inancial de elopmen
index is ela i ely high (low) may sugges ha a well-de eloped inancial sec o mi iga es
he cos s o axes. While axes c ea e dis o ions in alloca i e decisions, he inancial
sec o os e s g ow h by imp o ing he alloca ion and mobiliza ion o esou ces. The
No es:
1.Impulse esponse unc ions o pe capi a g oss domes ic p oduc (GDP) g ow h o a shock in he ax as a pe cen age
o GDP we e ob ained om he panel ec o au o eg ession (PVAR) o all coun ies, ad anced economies, and
eme ging economies, as well as g oupings on he de elopmen o he inancial sec o using he inancial
de elopmen index. The shaded a ea ep esen s he 90% con idence in e als based on 200 Mon e Ca lo
simula ions.
2.The PVAR was no es ima ed o he ad anced economies wi h less de eloped inancial sec o since he e a e oo
ew obse a ions.
Sou ce: Au ho s’ calcula ions.
30
unc ions o he inancial sec o such as alloca ing esou ces, mobilizing sa ings, and
easing he ading o goods and se ices a e signi ican ly connec ed o g ow h. Le ine
(1997) poin s ha he inancial sec o lowe s in o ma ion and ansac ion cos s in sa ings
and in es men decisions. De G ego io and Guido i (1995) ind ha inancial
in e media ion imp o es he e iciency o in es men , which would impac g ow h.
Ano he way o in e p e ing he esul is ha a de eloped inancial sec o imp o es
he e iciency o ax collec ion so ha he bene i s o ax ou weigh i s cos s as cos s a e
minimized. Ilie ski (2015) shows ha he highe s ock ma ke o al alue added is
posi i ely signi ican o ax e enues, while Gilbe and Ilie ski (2016) show ha expec ed
axes ise wi h banking ac i i y. The s udies ind ha he inancial sec o boos s ax
e enues. They conjec u e ha a weak inancial sys em imposes a cons ain in he
collec ion o axes, esul ing in ine iciency, a lowe ax base, and a high occu ence o
ax a oidance. Tsau ai (2021) inds ha inancial de elopmen enhances he e ec o
axes on g ow h, whe ein a complemen a i y be ween axa ion and inancial de elopmen
leads o a posi i e impac on g ow h. A mo e ad anced inancial sec o p o ides be e
means o iscal policies; hence, he aim o ax collec ion, alloca ion, and dis ibu ion o
os e g ow h is made e ec i e only when he inancial sys em is de eloped (O and
Ta om 2006, Gnangnon 2019, Tsau ai 2021).
A well-de eloped inancial sec o augmen s he g ow h impac o axes o ye
ano he eason. Financial sec o de elopmen lowe s he go e nmen ’s cos o bo owing.
Fo example, a la ge and smoo hly unc ioning ma ke o go e nmen bonds enables he
go e nmen o bo ow mo e a a lowe cos . Financial de elopmen hus enables he
go e nmen o mo e easily bo ow and complemen ax e enues wi h bo owed unds
31
when necessa y— o example, when he e is an unexpec ed sho all in ax e enue
collec ion. This p e en s he dis up ion o public spending, including g ow h-p omo ing
public spending such as in as uc u e in es men in eme ging economies.
4.3. Mac oeconomic En i onmen
4.3.1. Sa ings
The sample is g ouped in e ms o le els o sa ings cha ac e ized by he a iable na ional
sa ings as pe cen age o GDP. The median o he a iable is u ilized o ca ego ize low
and high le els o sa ings. The alues less han he median a e classi ied as low le els
o sa ings, while he alues g ea e han o equal o he median a e conside ed as high
le els o sa ings.
The baseline e ec o axes on g ow h appea s a low le els o sa ings, hough
only signi ican when using all economies and o eme ging economies (Figu e 8). A high
le els o sa ings, axes seem o ha e an immedia e ad e se e ec on g ow h in all
samples, bu his nega i e e ec is insigni ican .
The posi i e e ec o axes on g ow h in eme ging economies is obse ed a low
le els o sa ings. This may be because, o low-income economies, he sa ings channel
is no as powe ul as he ac o p oduc i i y channel o axes o a ec g ow h (Baldacci
e al. 2004). Taxes di ec ly a ec he a e o e u n o sa ings, esul ing o changes in
capi al accumula ion o sa e s (Summe s 1982). Wi h high le els o sa ings, he impac
o axes on g ow h is no longe signi ican . I may be he case ha , a high le els o
sa ings, he ad e se e ec o axes on accumula ion o capi al may be s onge . As he
sha e o sa ings o na ional income is high, he e ec can lead o nega i e consequences
on g ow h, which can cancel ou he g ow h- iendly e ec s o axes.
32
Mo e undamen ally, axes a e a ela i ely mo e impo an channel o mobilizing
domes ic esou ces o in es men and o he p oduc i e spending when he le el o
sa ings is low. Tha is, he sca ci y o p i a e domes ic esou ces means ha public
domes ic esou ces—i.e., ax e enues—mus play a g ea e ole in g ow h-p omo ing
expendi u es. As sa ings ise, he ela i e impo ance o ax e enues in domes ic
esou ce mobiliza ion declines.
Figu e 8: Sa ings
Low High
No e: Impulse esponse unc ions o pe capi a g oss domes ic p oduc (GDP) g ow h o a shock in he ax as a
pe cen age o GDP we e ob ained om he panel ec o au o eg ession (PVAR) o all coun ies, ad anced
economies, and eme ging economies, as well as g oupings on he le els o sa ings. The shaded a ea ep esen s
he 90% con idence in e als based on 200 Mon e Ca lo simula ions.
Sou ce: Au ho s’ calcula ions.

33
4.3.2. In es men s
The le els o in es men s a e desc ibed by he in es men s as a pe cen age o GDP
a iable. We use he median o he a iable o sepa a e obse a ions—low le els o
in es men s a e cha ac e ized by alues less han he median, while high le els o
in es men s a e alues g ea e han o equal o he median.
Sa ings and in es men s a e among he channels in which axes can in luence
g ow h. Taxes a ec p i a e decisions, including he indi idual’s decision o sa e and
accumula e capi al (Johansson e al. 2008), since axes al e he income s eam o he
indi idual and he eal a e o e u n o sa ings and in es men s (Boadway and Wildasin
1994) . O he s udies poin ou ha i is no he le el o sa ings and in es men s ha is
being a ec ed by axes, bu a he he composi ion o sa ings and in es men s
(Bo enbe g 1989, Johansson e al. 2008).5 Ba o (1991) inds ha axes can be g ow h-
enhancing h ough in es men s in public se ices ha ing posi i e ex e nali ies in he
p i a e sec o . Howe e , axes can also ad e sely a ec g ow h by educing sa ings and
capi al. Aside om impac ing g ow h h ough changes in p i a e sa ings and decisions,
axes can di ec ly a ec public sa ings. K ieckhaus (2002) inds ha public sa ings a e
associa ed wi h highe g ow h o de eloping s a es when hey a e used o p oduc i e
in es men s.
Fo he ull sample o economies and eme ging economies, he posi i e impac o
axes on g ow h is g ea e a low le els o in es men s han a high le els o in es men s
(Figu e 9). This is in ui i ely plausible since public spending inanced by ax e enues
5 Bo enbe g (1989) and Johansson e al. (2008) poin ou ha he e ec o axes on he le el o sa ings
and in es men s is small and unce ain as indings de e mining he ela ionship o sa ings and he eal
a e o e u n emain inconclusi e (e.g., Hall, 1988; Summe s, 1982).
34
such as heal h, educa ion, and in as uc u e will ha e a ela i ely mo e signi ican impac
on g ow h when o he in es men s a e low. Fo ad anced economies, he posi i e impac
o ax e enues on g ow h is obse ed a only high le els o in es men , p obably because
o he syne gy be ween public and p i a e in es men s in p oduc i i y-enhancing
in es men s such as R&D. This is in line wi h he s udy o Baldacci e al. (2004), which
inds ha , in high-income economies, in es men s a e he p ima y channel h ough which
iscal policies a ec g ow h.
Figu e 9: In es men s
Low High
No e: Impulse esponse unc ions o pe capi a g oss domes ic p oduc (GDP) g ow h o a shock in he ax as a
pe cen age o GDP we e ob ained om he panel ec o au o eg ession (PVAR) o all coun ies, ad anced
economies, and eme ging economies, as well as g oupings on he le els o in es men s. The shaded a ea
ep esen s he 90% con idence in e als based on 200 Mon e Ca lo simula ions.
Sou ce: Au ho s’ calcula ions.
35
4.3.3. In la ion
To cha ac e ize obse a ions wi h ela i ely low and high in la ion, we sepa a e he
samples by using he median. Low in la ion is desc ibed by alues less han he median,
while high in la ion is cha ac e ized by alues g ea e han o equal o he median.
Fo all economies, he baseline esul whe ein axes posi i ely a ec he g ow h is
obse ed only when in la ion is ela i ely low (Figu e 10). Rega dless o he le el o
in la ion, axes ha e an ad e se e ec on g ow h o he ad anced economies, hough he
magni ude o he nega i e impac seems o be la ge when in la ion is ela i ely high. Fo
eme ging economies, he impac o axes on g ow h a ies be ween hose wi h ela i ely
low and high in la ion. The posi i e e ec o axes on g ow h is associa ed wi h a ela i ely
low in la ion. When in la ion is ela i ely high, o eme ging economies, axes ha e an
ad e se e ec on g ow h.
Pe sis en high in la ion is a signal o ine iciencies in he ax sys em (De G ego io
1993). This can be explained by he model o Cukie man e al. (1992) in which economies
wi h g ea e poli ical ins abili y and a mo e pola ized poli ical sys em will exhibi ine icien
ax sys ems and, he e o e, will ely mo e hea ily on seignio age o in la iona y ax. The
link be ween in la ion and he ax s uc u e is also shown by Roubini and Sala-i-Ma in
(1992), whe ein a go e nmen subjec ed o a high ax e asion in income chooses o
inc ease e enue h ough seignio age by ep essing he inancial sec o and inc easing
in la ion a es.
The posi i e impac o axes on g ow h becoming non-exis en a high le els o
in la ion using all economies and u ning nega i e a high le els o in la ion o eme ging
economies may be explained h ough he p esence o ine icien ax sys ems. Ine icien
36
ax sys em is associa ed wi h subs an ial cos s in he adminis a ion and en o cemen o
collec ing ax e enues. Wi h an ine icien ax sys em, he dis o iona y e ec s o axes
may be mo e subs an ial so ha he cos s o axes ou weigh he bene i s o axes h ough
he inancing o p oduc i e spending.
O e all, he esul s indica e ha low in la ion ampli ies he posi i e impac o ax
e enues on g ow h. Low in la ion e lec s mac oeconomic s abili y, which, in u n, e lec s
sound mone a y and iscal policies. A cen al componen o sound iscal policy is he
e icien use o ax e enues, which inc eases he posi i e e ec o axes on g ow h. The
posi i e g ow h impac o axes ul ima ely depends on he ex en o which ax e enues
a e used p oduc i ely, o ins ance ia public in es men in in as uc u e.
43
Figu e 13: Tax Incen i es Ac oss Income G oups
Sou ce: P esen a ion by K on ol, Hania, and Sebas ian James. 2021. “Taxing Times: The Role o
In es men Incen i es in Economic Reco e y and G ow h” [Webina ]. Wo ld Bank. 27 May.
h ps://www.wo ldbank.o g/en/e en s/2021/05/05/ axing- imes- he- ole-o -incen i es-in-economic-
eco e y-and-g ow h
6. Commen s on Th eshold
Se e al heo e ical models and empi ical s udies ha e p esen ed he non-linea
ela ionship o axes and g ow h. Ba o (1991) and A mey (1995) pos ula e an in e se U-
shaped ela ionship be ween go e nmen size and g ow h. Using an endogenous g ow h
model, Ba o (1991) shows ha , ini ially, g ow h a e ises wi h he ax a e as p i a e
p oduc i i y is enhanced by public se ices inanced h ough axes. Bu as he ax a e
inc eases, p i a e in es men s a e discou aged, which can be de imen al o g ow h.
E en ually, his la e e ec domina es he e ec on p i a e p oduc i i y so ha wi h a e y
la ge go e nmen , ax a e is nega i ely ela ed o g ow h. Esen and Aydin (2019),
employing a dynamic panel h eshold model, ind his in e se U-shaped ela ionship o
ax e enues and g ow h o 11 cen al and sou heas e n Eu opean and Bal ic economies.
Thei indings p esen ha axes, as a sha e o GDP below he h eshold, ha e a bene icial
impac on g ow h. In addi ion, abo e he h eshold, he ax a io o GDP ad e sely a ec s
0% 10% 20% 30% 40% 50% 60%
To al
High income
Uppe -Middle-Income
Lowe -Middle-Income
Low income
Sha e o coun ies ha made incen i es mo e gene ou s in a leas one sec o
Sha e o coun ies ha made incen i es less gene ous in a leas one sec o

44
g ow h. They also obse e di e en h esholds o di e en economy g oups: 18.0% o
ull ansi ion economies, 18.5% o de eloping economies, and 23.0% o de eloped
economies. Simila ly, he esul s o o he s udies show he in e se U-shaped ela ionship
o go e nmen size and g ow h, bu use go e nmen expendi u es o de ine go e nmen
size (Chen and Lee 2005, Al unc and Aydın 2013, Asimakopoulos and Ka a ias 2016).
Fo Gaspa e al. (2016), hey also ind a h eshold in he e ec s o axes on g ow h;
howe e , ins ead o a h eshold ha maximizes g ow h, he h eshold ep esen s a
minimum le el ha would accele a e g ow h. Using wo la ge unbalanced panel da ase s,
he es ima ed ax- o-GDP h esholds a e (i) 12.9% om a con empo a y da abase o 139
economies om 1965 o 2011 and (ii) 12.7% om a his o ical da abase o 30 ad anced
economies om 1800 o 1980.
We also in es iga e whe he he non-linea i y o he e ec s o axes on g ow h is
p esen using ou ax da abase.7 The ixed-e ec panel h eshold model by Hansen
(1999, 2000) is employed o es ima e he h eshold. We un he es ima ion model using
ou baseline speci ica ion wi h lag o GDP g ow h as an added con ol.8
Ou exe cise inds a ax– o–GDP h eshold o 12.3%, which is close o he
h eshold es ima ed by Gaspa e al. (2016). We also obse e ha h esholds a e di e en
o ad anced and eme ging economies when es ima ed sepa a ely o he wo g oups.
The esul s p esen he ax- o-GDP h eshold le els o 16.7% o ad anced economies
and 13.2% o eme ging economies.
7 The ax da abase used in calcula ing he h eshold co e s he pe iod 1990–2020, which inds consis en
es ima es o he h esholds. The es ima ed h esholds using his pe iod a e no sensi i e o he sample o
coun ies ha a e co e ed in he es ima ion.
8 Knelle e al. (1999), Lee and Go don (2005), and Angelopoulos e al. (2007) include ini ial GDP as one
o hei con ols in es ima ing he impac o axes on g ow h using a ixed-e ec panel eg ession model.
45
7. Conclusion
The cen al objec i e o ou pape is o examine how mac oeconomic and s uc u al
ac o s a ec he impac o ax e enues on g ow h. In ui i ely, such ac o s help
de e mine whe he and o wha ex en axes con ibu e o g ow h. Fo ins ance, good
go e nance imp o es he quali y o public spending inanced by axes and ampli ies he
posi i e e ec o axes on g ow h. Simila ly, low in la ion—an indica o o sound mone a y
and iscal policies—inc eases he g ow h e ec o axes. In ou empi ical analysis, we
employ a PVAR model using panel da a o 135 economies and co e ing he pe iod 1990–
2019. We i s es he baseline e ec o axes on g ow h. We hen in es iga e how
mac oeconomic policies, s uc u al ac o s, and he mac oeconomic en i onmen
in luence he ela ionship be ween axes and g ow h. We do so by examining how he
baseline e ec o axes on g ow h changes wi h changes in mac oeconomic policies,
s uc u al ac o s, and he mac oeconomic en i onmen .
The baseline esul s show ha axes ha e a posi i e e ec on economic g ow h in
he ull sample o economies and eme ging economies. Con e sely, in he ad anced
economies, axes ad e sely impac g ow h. Among mac oeconomic policies, he in la ion-
a ge ing scheme and exchange a e egime appea o in luence he e ec o axes on
g ow h. We ind ha a lexible exchange a e egime is associa ed wi h axes ha ing a
mo e benign e ec on g ow h. On he o he hand, he in luence o in la ion- a ge ing
scheme on he g ow h e ec o axes a ies be ween ad anced and eme ging economies.
Fo s uc u al ac o s, bo h he quali y o go e nance and he inancial sec o a e
impo an in de e mining he ela ionship be ween axes and g ow h. We ind ha bo h
magni y he bene icial impac o axes on g ow h. In e ms o he mac oeconomic
46
en i onmen , he ela ionship be ween axes and g ow h is mo e appa en a low le els
o sa ings and in es men s. The impac o sa ings and in es men s on he ax–g ow h
nexus becomes mo e complex a highe le els o sa ings and in es men s. Howe e , in
ad anced economies, axes can acili a e g ow h a high in es men le els. Finally, high
in la ion is associa ed wi h a mo e ad e se impac o axes on g ow h. Table 3
summa izes ou empi ical esul s.
Table 3: Summa y o Resul s
Fac o s A ec ing he
Tax
–
G ow h Nexus
Ca ego y All
Economies
Ad anced
Economies
Eme ging
Economies
Baseline Posi i e a Nega i e a Posi i e a
Mac oeconomic Policies
In la ion Ta ge ing Loose Posi i e a Nega i e a Posi i e a
S ic Posi i e a Posi i e a Nega i e a
Exchange Ra e Regime Fixed None Nega i e a Nega i e
Flexible Posi i e a Nega i e a Posi i e a
Fiscal Rule—Expendi u e
Rule Wi hou Nega i e Posi i e Posi i e
Wi h Nega i e Nega i e None
Fiscal Rule—Re enue
Rule Wi hou Posi i e Nega i e Posi i e a
Wi h Posi i e a Too ew obs None
Fiscal Rule—Budge
Balance Wi hou Posi i e,
hen
Nega i e
Posi i e, hen
Nega i e Nega i e
Wi h Nega i e a Nega i e Nega i e
Fiscal Rule—Deb Rule Wi hou Nega i e,
hen Posi i e Nega i e Nega i e
Wi h Nega i e a Nega i e Nega i e a
Fiscal Rule—Any Rule Wi hou Posi i e,
hen
Nega i e
None Nega i e
Wi h Nega i e a Nega i e Nega i e
S uc u al Fac o s
WGI—Go e nmen
E ec i eness Weak None No obs None
S ong Posi i e a Nega i e a Posi i e a
Con inued on he nex page
47
Fac o s A ec ing he
Tax
–
G ow h Nexus
Ca ego y All
Economies
Ad anced
Economies
Eme ging
Economies
WGI—Con ol o
Co up ion Weak None Too ew obs None
S ong Posi i e a Nega i e Posi i e a
WGI—Poli ical S abili y Weak None Too ew obs None
S ong Posi i e a Nega i e Posi i e a
WGI—Regula o y Quali y Weak None No obs None
S ong Posi i e a Nega i e a Posi i e a
WGI—Rule o Law Weak None No obs None
S ong Posi i e a Nega i e a Posi i e a
WGI—Voice and
Accoun abili y Weak None Too ew obs None
S ong Posi i e a Nega i e a Posi i e a
Financial Sec o
De elopmen
Less
De eloped
Nega i e a No obs Nega i e a
Mo e
Ad anced Posi i e a Nega i e Posi i e
Mac oeconomic Fac o s
Sa ings Low Posi i e a Nega i e Posi i e a
High Nega i e Nega i e Nega i e
In es men s Low Posi i e a Nega i e a Posi i e a
High Posi i e Posi i e a Posi i e
In la ion Low Posi i e a Nega i e a Posi i e a
High None Nega i e a Nega i e a
WGI = Wo ldwide Go e nance Indica o s.
a Deno es signi icance a he 90% con idence in e al.
No es:
1. Calcula ed based on he es ima ion o he panel ec o au o eg ession (PVAR) equa ion (1) and he co esponding
impulse esponse unc ions o g oss domes ic p oduc (GDP) g ow h o a shock in axes as pe cen age o GDP.
2. “None” ep esen s a la impulse esponse unc ion.
3. “No obs” and “Too ew obs” mean ha he subsample con ains no obse a ions o oo ew obse a ions, espec i ely,
o es ima e he PVAR (Equa ion (1)).
Sou ce: Au ho s’ calcula ions.
Ou indings sugges ha axes a e no necessa ily de imen al o economic
de elopmen and can e en be g ow h-enhancing. Al hough axes a e inhe en ly
dis o iona y and hus a sou ce o economic ine iciency, hey inance g ow h-p omo ing
public expendi u es on heal h, educa ion, and in as uc u e. The la e a e especially
impo an in eme ging economies, which ha e ela i ely smalle go e nmen s and place
48
a highe p io i y on g ow h as compa ed wi h ad anced economies, which place a highe
p io i y on income edis ibu ion. This di e ence be ween ad anced economies and
eme ging economies helps o explain why he o e all balance o ou e idence poin s o a
g ea e con ibu ion o ax e enues o g ow h in eme ging economies. A a b oade le el,
ou e idence con i ms ha he e ec o ax e enues on economic g ow h should no be
e alua ed in isola ion bu in conjunc ion wi h mac oeconomic and s uc u al ac o s ha
can in luence he ax–g ow h nexus. Ou esul s a e gene ally consis en wi h economic
in ui ion. Fo ins ance, we ind ha low in la ion, which e lec s sound iscal policies and,
hence, e icien use o ax e enues, ampli ies he posi i e e ec o ax e enues on
g ow h.

49
APPENDIX
Da ase Sho Desc ip ion Sou ce
OECD
Re enue
S a is ics
Da abase
The da ase p o ides de ailed
in o ma ion on ax and go e nmen
e enues, ca ego ized in o he ypes
o ax e enue and le els o
go e nmen .
OECD
IMF
Go e nmen
Finance
S a is ics
The da ase p esen s he iscal da a
o all epo ing coun ies in he
amewo k o he Go e nmen
Finance S a is ics Manual 2014,
which includes he da a on e enues,
expendi u es, ansac ions in inancial
asse s and liabili ies, and i s
subsec o s.
IMF
ADB Key
Indica o s
Da abase
The da abase con ains
mac oeconomic and socioeconomic
s a is ics o ADB's membe
economies, including da a on
na ional accoun s, p ices,
go e nmen inance, ade, balance
o paymen s, money and in e es
a es, ex e nal deb , popula ion, labo
o ce, and social indica o s.
ADB
Wo ld
De elopmen
Indica o s
The da abase compiles se e al
compa able s a is ics abou global
de elopmen , co e ing hemes such
as po e y and inequali y, people,
en i onmen , economy, s a es,
ma ke s, and global links.
Wo ld Bank
Exchange Ra e
Regime
Classi ica ion
The da ase indica es he de ac o
exchange a e a angemen o he
coun ies, classi ying he exchange
a e egimes in o wo ypes o
classi ica ions—one wi h 15
ca ego ies ( ine classi ica ion), and
he o he wi h 6 ca ego ies (coa se
classi ica ion).
Ilze zki, E han, Ca men M.
Reinha , and Kenne h S.
Rogo . 2019. “Exchange
A angemen s En e ing he
Twen y-Fi s Cen u y: Which
Ancho will Hold?” The
Qua e ly Jou nal o
Economics, 134 (2): 599–646.
Ilze zki, E han, Ca men M.
Reinha , and Kenne h S.
Con inued on he nex page
50
Da ase Sho Desc ip ion Sou ce
Rogo . 2022. “Re hinking
Exchange Ra e Regimes.” In
Handbook o In e na ional
Economics 6 edi ed by Gi a
Gopina h, Elhanan Helpman,
and Kenne h Rogo , 91–145.
No h Holland: Else ie .
Reinha , Ca men M., and Kenne h
S. Rogo . 2004. “The Mode n
His o y o Exchange Ra e
A angemen s: A
Rein e p e a ion.” Qua e ly
Jou nal o Economics, CXIX
(1): 1–48.
IMF Fiscal
Rules
Da ase
The da ase con ains in o ma ion on
he use and design o iscal ules,
co e ing na ional and sup ana ional
iscal ules. The ou ypes o iscal
ules a e budge balance ules, deb
ules, expendi u e ules, and e enue
ules.
Da oodi, Hamid R., Paul Elge ,
Alexand a Fo iou, Daniel
Ga cia-Macia, And esa
Lage bo g, Raphael Lam, and
Sha anya Pillai. 2022. Fiscal
Rules: 1985–2021 [Da ase ].
In e na ional Mone a y Fund.
Wo ldwide
Go e nance
Indica o s
The da abase epo s go e nance
indica o s o six dimensions o
go e nance: (i) oice and
accoun abili y, (ii) poli ical s abili y
and absence o iolence o e o ism,
(iii) go e nmen e ec i eness, (i )
egula o y quali y, ( ) ule o law, and
( i) con ol o co up ion.
Kau mann, Daniel, Aa K aay, and
Massimo Mas uzzi. 2010. The
Wo ldwide Go e nance
Indica o s: Me hodology and
Analy ical Issues. Wo ld Bank
Policy Resea ch Wo king
Pape No. 5430. Wo ld Bank.
Financial
De elopmen
Index
The index da abase con ains nine
indexes, which summa izes how
de eloped inancial ins i u ions and
inancial ma ke s a e in e ms o
dep h, access, and e iciency.
IMF
ADB = Asian De elopmen Bank, IMF = In e na ional Mone a y Fund, OECD = O ganisa ion o Economic Co-
ope a ion and De elopmen .
Sou ce: Au ho s’ compila ion.
51
REFERENCES
Ab igo, Michael R. M., and Inessa Lo e. 2016. "Es ima ion o Panel Vec o Au o eg ession in
S a a." The S a a Jou nal, 16 (3): 778–804.
A onso, An ónio, and João To a Jalles. 2013. "Do Fiscal Rules Ma e o G ow h?" Applied
Economics Le e s, 20 (1): 34–40.
A onso, An ónio, and João To a Jalles. 2016. "Economic Pe o mance, Go e nmen Size, and
Ins i u ional Quali y." Empi ica, 43 (1): 83–109.
Alesina, Albe o, and Sil ia A dagna. 2010. "La ge Changes in Fiscal Policy: Taxes e sus
Spending." Tax Policy and he Economy, 24: 35–68.
Al unc, O. Fa uk, and Celil Aydın. 2013. "The Rela ionship be ween Op imal Size o
Go e nmen and Economic G ow h: Empi ical E idence om Tu key, Romania and
Bulga ia." P ocedia - Social and Beha io al Sciences, 92, 66–75.
Angelopoulos, Kons an inos, Geo ge Economides, and Pan elis Kammas. 2007. "Tax-Spending
Policies and Economic G ow h: Theo e ical P edic ions and E idence om he OECD."
Eu opean Jou nal o Poli ical Economy, 23 (4): 885–902.
Apos olakis, Geo ge, and A hanasios Papadopoulos. 2019. "Financial S abili y, Mone a y
S abili y and G ow h: A PVAR Analysis." Open Economies Re iew, 30 (1): 157–78.
A mey, Richa d Kei h. 1995. The F eedom Re olu ion. Regne y Publishing.
A nold, Jens Ma hias. 2008. "Do Tax S uc u es A ec Agg ega e Economic G ow h? Empi ical
E idence om a Panel o OECD Coun ies." OECD Economics Depa men Wo king
Pape , No. 643. OECD Publishing.
A nold, Jens Ma hias, Be B ys, Ch is ophe Heady, Åsa Johansson, Cy ille Schwellnus, and
Lau a Va ia. 2011. "Tax Policy o Economic Reco e y and G ow h." The Economic
Jou nal, 121 (550): F59-F80.
A in, Mak B., P adhan, Rud a P., and Mahindhi an S. Nai . 2021. "A e The e Links be ween
Ins i u ional Quali y, Go e nmen Expendi u e, Tax Re enue and Economic G ow h?
E idence om Low-Income and Lowe Middle-Income Coun ies." Economic Analysis and
Policy, 70 (6–7): 468–89.
Asimakopoulos, S ylianos, and Yiannis Ka a ias. 2016. "The Impac o Go e nmen Size on
Economic G ow h: A Th eshold Analysis." Economics Le e s, 139 (2): 65–8.
Ay es, Kelly, A iel R. Belasen, and Ali M. Ku an. 2014. "Does In la ion Ta ge ing Lowe In la ion
and Spu G ow h?" Jou nal o Policy Modeling, 36 (2): 373–88.
Badinge , Ha ald, and Wol Hein ich Reu e . 2017. "The Case o Fiscal Rules." Economic
Modelling, 60: 334–43.
52
Baldacci, Emanuele, A ye L. Hillman, and Naoko C. Kojo. 2004. "G ow h, Go e nance, and
Fiscal Policy T ansmission Channels in Low-Income Coun ies." Eu opean Jou nal o
Poli ical Economy, 20 (3): 517–49.
Ball, Lau ence M., and Niamh She idan. 2004. "Does In la ion Ta ge ing Ma e ?" In The
In la ion-Ta ge ing Deba e edi ed by Ben S. Be nanke and Michael Wood o d, 249–76.
Uni e si y o Chicago P ess.
Ba o, Robe J. 1991. "A C oss-Coun y S udy o G ow h, Sa ing, and Go e nmen ." In
Na ional Sa ing and Economic Pe o mance, edi ed by B. Douglas Be nheim and John B.
Sho en, 271–304. Uni e si y o Chicago P ess.
Beck, Tho s en, Chen Lin, and Yue Ma. 2014. Why Do Fi ms E ade Taxes? The Role o
In o ma ion Sha ing and Financial Sec o Ou each." Jou nal o Finance, 69 (2): 763–817.
Be ge , Helge, Jan-Egbe S u m, and Jakob de Haan. 2000. “An Empi ical In es iga ion in o
Exchange Ra e Regime Choice and Exchange Ra e Vola ili y.” CESi o Wo king Pape No.
263. Munich: Cen e o Economic S udies and i o Ins i u e (CESi o).
Be gh, And eas, and Magnus Hen ekson. 2011. "Go e nmen Size and G ow h: A Su ey and
In e p e a ion o he E idence." IFN Wo king Pape No. 858. Resea ch Ins i u e o Indus ial
Economics (IFN).
Bi d, Richa d M., Jo ge Ma inez-Vazquez, and Benno To gle . 2008. "Tax E o in De eloping
Coun ies and High Income Coun ies: The Impac o Co up ion, Voice and
Accoun abili y." Economic Analysis and Policy, 38 (1): 55–71.
Bleaney, Michael, and Manuela F ancisco. 2007. "Exchange Ra e Regimes, In la ion and
G ow h in De eloping Coun ies—An Assessmen ." The B.E. Jou nal o Mac oeconomics,
7 (1): 1–20.
Boadway, Robin, and Da id Wildasin. 1994. Taxa ion and Sa ings: A Su ey." Fiscal S udies,
15 (3): 19–63.
Bo enbe g, A. Lans. 1989. "Tax Policy and Na ional Sa ings in he Uni ed S a es: A Su ey."
Na ional Tax Jou nal, 42 (2): 123–38.
Cano a, Fabio, and Ma eo Cicca elli. 2013. "Panel Vec o Au o eg essi e Models: A Su ey."
VAR Models in Mac oeconomics – New De elopmen s and Applica ions: Essays in Hono
o Ch is ophe A. Sims, 32: 205–246. Eme ald G oup Publishing Limi ed.
Caselli, F ancesca, and Julien Reynaud. 2020. "Do Fiscal Rules Cause Be e Fiscal Balances?
A New Ins umen al Va iable S a egy." Eu opean Jou nal o Poli ical Economy, 63:
101873.
Cas o, Vi o . 2011. "The Impac o he Eu opean Union Fiscal Rules on Economic G ow h."
Jou nal o Mac oeconomics, 33 (2): 313–26.
Chaudh y, Azam, and Phillip Ga ne . 2007. "Do Go e nmen s Supp ess G ow h? Ins i u ions,
Ren -Seeking, and Inno a ion Blocking in a Model o Schumpe e ian G ow h." Economics
& Poli ics, 19 (1): 35–52.