Bambe, Bao-We-Wal
Wo king Pape
Mac op uden ial policies and p i a e domes ic in es men
in de eloping coun ies: An ins umen al a iables
app oach
IDOS Discussion Pape , No. 3/2025
P o ided in Coope a ion wi h:
Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS), Bonn
Sugges ed Ci a ion: Bambe, Bao-We-Wal (2025) : Mac op uden ial policies and p i a e domes ic
in es men in de eloping coun ies: An ins umen al a iables app oach, IDOS Discussion Pape ,
No. 3/2025, ISBN 978-3-96021-246-1, Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS),
Bonn,
h ps://doi.o g/10.23661/idp3.2025
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/313611
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Mac op uden ial Policies and P i a e
Domes ic In es men in De eloping
Coun ies
An Ins umen al Va iables App oach
Bao-We-Wal Bambe
IDOS DISCUSSION PAPER
3/2025
Mac op uden ial policies and p i a e
domes ic in es men in de eloping
coun ies
An ins umen al a iables app oach
Bao-We-Wal Bambe
Bonn 2025
Bao-We-Wal Bambe is a esea che in he depa men “T ans o ma ion o Economic and Social Sys ems” a
he Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS) in Bonn.
Email: Bao-We-Wal.Bambe@idos- esea ch.de
Published wi h inancial suppo om he Fede al Minis y o Economic Coope a ion and De elopmen (BMZ),
based on a esolu ion o he Ge man Bundes ag. The ins i u es o he Johannes-Rau-Fo schungsgemeinscha
a e ins i u ionally unded by he s a e o NRW.
Sugges ed ci a ion:
Bambe, B.-W.-W. (2025). Mac op uden ial policies and p i a e domes ic in es men in de eloping coun ies:
E idence om an ins umen al a iables s a egy (IDOS Discussion Pape 3/2025). Bonn: Ge man Ins i u e o
De elopmen and Sus ainabili y (IDOS). h ps://doi.o g/10.23661/idp3.2025
Disclaime :
The iews exp essed in his pape a e hose o he au ho (s) and do no necessa ily e lec he iews o policies
o he Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS).
Excep o he wise no ed, his publica ion is licensed unde C ea i e Commons A ibu ion (CC BY 4.0). You a e
ee o copy, communica e and adap his wo k, as long as you a ibu e he Ge man Ins i u e o De elopmen
and Sus ainabili y (IDOS) gGmbH and he au ho (s).
IDOS Discussion Pape / Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS) gGmbH
ISSN 2751-4439 (P in )
ISSN 2751-4447 (Online)
ISBN 978-3-96021-246-1 (P in )
DOI: h ps://doi.o g/10.23661/idp3.2025
© Ge man Ins i u e o De elopmen and Sus ainabili y (IDOS) gGmbH
Tulpen eld 6, 53113 Bonn
Email: publica ions@idos- esea ch.de
h ps://www.idos- esea ch.de
P in ed on eco- iendly, ce i ied pape .
IDOS Discussion Pape 3/2025
III
Abs ac
This pape examines he e ec o mac op uden ial policies on p i a e domes ic in es men using
a panel o 87 de eloping coun ies om 2000 o 2017. Ou ins umen al a iables s a egy
exploi s he geog aphic di usion o mac op uden ial policies ac oss coun ies, wi h he idea ha
e o ms in neighbou ing coun ies can a ec he adop ion o s eng hening o domes ic e o ms
h ough pee p essu e o imi a ion e ec s. The indings indica e ha he igh ening o mac o-
p uden ial policies signi ican ly educes p i a e domes ic in es men . This e ec holds o bo h
ins umen s a ge ing bo owe s and hose a ge ing inancial ins i u ions, and is subjec o
he e ogenei y depending on se e al economic and ins i u ional ac o s. The ansmission
channel analysis highligh s ha he nega i e impac o mac op uden ial policies on in es men
is p ima ily d i en by a educ ion in c edi supply and inancial inclusion.
Keywo ds: Mac op uden ial policies; p i a e domes ic in es men ; de eloping coun ies;
ins umen al a iables
JEL Classi ica ion: E22; E44; G28
Acknowledgmen s
I hank Ka h in Be ensmann, Cla a B andi, Jean-Louis Combes, Tim Roe hel, Ch is oph Somme
and Yabibal Walle o hei aluable commen s and sugges ions o imp o e he pape . Usual
disclaime s apply.
IDOS Discussion Pape 3/2025
IV
Con en s
Abs ac
Acknowledgmen s
Abb e ia ions
1 In oduc ion 1
2 Backg ound and heo e ical p edic ions 2
3 Da a and s ylised ac s 3
3.1 Da a 3
3.2 S ylised ac s 4
4 Me hodology and main indings 8
4.1 Ins umen al a iables s a egy 8
4.2 Econome ic speci ica ion and main esul s 8
5 Robus ness checks 9
5.1 Al e na i e speci ica ions and addi ional con ols 9
5.2 Al e na i e subsamples and measu es 13
6 He e ogenei y 14
7 Mechanisms 17
8 Conclusions and policy ecommenda ions 18
Re e ences 20
Appendix A: Fu he obus ness 25
GMM es ima es 25
Th ee-yea window 25
Table A1: Mac op uden ial policies (MPI) and p i a e domes ic in es men : al e na i e
subsamples and measu es 26
Table A2: Al e na i e subsamples and measu es: i s s age esul s 27
Table A3: Mac op uden ial policies and p i a e domes ic in es men : Sys em-GMM
and IV- h ee-yea window es ima es 28
Table A4: He e ogenei y: i s s age esul s 29
Appendix B: Sample and desc ip i e s a is ics 30
Table B1: Summa y s a is ics o he baseline model a iables 30
Table B2: Sample 31
Table B3: Sou ces o a iables 32
IDOS Discussion Pape 3/2025
V
Figu es
Figu e 1: T ends in p i a e domes ic in es men by income le el 5
Figu e 2: T ends in MPI by income le el 6
Figu e 3: T ends in MPI by ca ego y 6
Tables
Table 1: Global mac op uden ial policy ins umen s su ey 7
Table 2: The e ec o mac op uden ial policies on p i a e domes ic in es men 11
Table 3: The e ec o mac op uden ial policies on p i a e domes ic in es men :
i s s age esul s 12
Table 4: The e ec o mac op uden ial policies on p i a e domes ic in es men :
he e ogenei y 16
Table 5: Mac op uden ial policies and p i a e domes ic in es men : channels 18
IDOS Discussion Pape 3/2025
VI
Abb e ia ions
CG c edi g ow h
CONC concen a ion limi s
DTI deb - o-income
FC o eign cu ency loans
FX o eign exchange
GDP g oss domes ic p oduc
GMM gene alised me hod o momen s
IMF In e na ional Mone a y Fund
ICRG In e na ional Coun y Risk Guide
INTER in e bank exposu es
IV ins umen al a iables
LEV le e age
LTV loan- o- alue
MPI mac op uden ial policy index
SIFI sys emically impo an inancial ins i u ions
WDI Wo ld Bank’s Wo ld De elopmen Indica o s
IDOS Discussion Pape 3/2025
1
1 In oduc ion
Mac op uden ial policies a e inc easingly used in ad anced and de eloping economies,
especially since he 2008–2009 global inancial c isis and ollowing Basel III, which in oduced
a comp ehensi e amewo k aimed a e o ming banking o e sigh , egula ion and isk con ol
– such as enhanced capi al adequacy equi emen s, imp o ed le e age managemen , new
capi al and liquidi y bu e s, and le e age a io limi s (Rubio & Ca asco-Gallego, 2016).1 The
ole o mac op uden ial policies in p omo ing inancial s abili y by egula ing c edi cycles is all
he mo e c ucial, as s udies show ha inancial c ises a e mo e likely o occu when hey a e
p eceded by p i a e c edi booms (Ge le e al., 2020; Schula ick & Taylo , 2012). Empi ical
e idence om a g owing body o li e a u e sugges s ha mac op uden ial policies end o educe
c edi p ocyclicali y, c edi g ow h, and house p ices (e.g. see Alam e al., 2019; Ce u i e al.,
2017; De Sch yde & Opi z, 2021; Gómez e al., 2020; Ku ne & Shim, 2016; Lim e al., 2011;
Teixei a & Ven e , 2023). Ano he s and o he li e a u e shows ha by mode a ing c edi and
asse p ice cycles, mac op uden ial ools help cons ain inancial cycles and sys emic isks ( o
ins ance, see Al unbas e al., 2018; Belkhi e al., 2022; Bianchi & Mendoza, 2018; Claessens
e al., 2013; Ge le e al., 2020; Fe nandez-Galla do, 2023).
While he li e a u e has widely examined he e ec s o mac op uden ial policies on c edi g ow h
o p ocyclicali y and inancial s abili y, ou s udy complemen s exis ing s udies by assessing he
side e ec s o hese policies on p i a e domes ic in es men in de eloping coun ies.2 The e ec
o mac op uden ial policies on p i a e domes ic in es men is no so clea -cu . I success ul, hey
should imp o e inancial s abili y, hus con ibu ing o educing economic unce ain y and
imp o ing p i a e sec o in es men o e ime – especially as inancial c ises a e ound o hampe
economic g ow h, employmen and in es men (e.g. see Ba o, 2001; Reinha & Reinha ,
2015). On he o he hand, ce ain mac op uden ial policies such as coun e cyclical capi al
bu e s, liquidi y ools, o hose a ge ing bo owe s (such as loan- o- alue o deb - o-income
a ios) ha e es ic i e e ec s on c edi , he eby wo sening inancial inclusion (Aiya e al., 2014;
Ayyaga i e al., 2018; Delécha e al., 2021). Lowe c edi supply esul ing om mac op uden ial
policies may penalise p i a e sec o in es men , especially in de eloping coun ies, whe e
nume ous i ms al eady ace signi ican challenges in accessing adequa e inancing (Beck e
al., 2005; Chau e & Jacolin, 2017; Ha ison e al., 2004).
This pape examines he e ec o mac op uden ial policies on p i a e domes ic in es men in
de eloping coun ies. Mac op uden ial policies a e likely o be co ela ed wi h (in)obse able
ac o s ha could also a ec he economy’s o e all pe o mance, including domes ic in es men .
The e o e, o mi iga e endogenei y, we d aw on exis ing s udies o exploi an exogenous sou ce
o a ia ion, ins umen ing mac op uden ial policies by he a e age egional mac op uden ial
policy index. The esul s om ou ins umen al- a iables (IV) s a egy sugges ha mac o-
p uden ial policies educe p i a e domes ic in es men . The e ec s a e s a is ically and eco-
nomically signi ican , and obus o a ious es s. He e ogenei y analyses show ha he nega i e
e ec o mac op uden ial policies on p i a e in es men is obse ed o bo h policies a ge ing
1 The In e na ional Mone a y Fund (IMF) de ines mac op uden ial policy as “ he use o p ima ily
p uden ial ools o limi sys emic isk o sys em-wide inancial isk” (IMF, 2011). See Clemen (2010)
o he o igins and e olu ion o he e m.
2 We ocus on de eloping coun ies, i.e., eme ging and low-income economies, o wo main easons.
Fi s , he la e ha e expe ienced a su ge in mac op uden ial ools o imp o e inancial s abili y in ecen
yea s, and in con as o hei ad anced coun e pa s, he p i a e sec o in hese egions is se e ely
penalized by low le els o in es men , hus c ea ing a u he challenge o achie ing de elopmen goals.
Second, ocusing on de eloping coun ies allows us o ha e a ela i ely homogeneous sample o
coun ies, o ins ance in e ms o se e al economic, s uc u al, and ins i u ional ac o s (e.g. pe capi a
income, ulne abili y o ex e nal shocks, quali y o ins i u ions, access o inancial ma ke s, e c.).
IDOS Discussion Pape 3/2025
8
4 Me hodology and main indings
4.1 Ins umen al a iables s a egy
Mac op uden ial policies may be endogenous, as hei adop ion o igh ening may be associa ed
wi h o he al e na i e measu es, c ea ing an iden i ica ion bias due o unobse ables. In o he
wo ds, es ima ing a causal e ec is challenging, as i is di icul o de e mine whe he he
obse ed e ec is genuinely due o mac op uden ial policies o a he o al e na i e policies.
Hence, pu ging mac op uden ial policy ac ions o esol e po en ial endogenei y p oblems is
c ucial. A he same ime, inding a ele an and alid ins umen is widely acknowledged as a
challenging ask. Many s udies examining he e ec o e o ms explo e he egional di usion o
hese e o ms as an ins umen al a iable, in he idea ha s uc u al e o ms o en occu as
egional wa es. The unde lying in ui ion is ha e o ms in neighbou ing coun ies can ha e a
s ong spillo e e ec in he adop ion o s eng hening o domes ic e o ms – ia simple imi a ion
mechanisms, pee p essu e, lea ning o compe i ion (see Bue a e al., 2011; Dobbin e al., 2007;
Hun ing on, 1991; Shipan & Volden, 2008). S udies explo ing he in luence o egional
democ a ic e o ms as a d i ing o ce o ins umen o ini ia ing na ional democ a ic e o ms
include, among o he s, Acemoglu e al., 2019; Giuliano e al., 2013; Kalenbo n & Lessmann,
2013; Pe sson & Tabellini, 2009. Simila ly, o he wo ks exploi iscal ules in neighbou ing
coun ies, wi h he idea ha coun ies d aw on he expe ience o hei neighbou s when
in oducing such e o ms (e.g. see Al unbaş & Tho n on, 2017; Ape i e al., 2024a; Bal i , 2024;
Caselli & Reynaud, 2020). In he case o mone a y e o ms, o ins ance, Balima e al. (2017)
use he p opo ion o neighbou ing coun ies ha ha e adop ed in la ion a ge ing as an
ins umen al a iable o a coun y’s decision o adop he mone a y amewo k.
Based on he li e a u e discussed abo e, we use he a e age o mac op uden ial policies in
egional coun ies as an ins umen al a iable o domes ic mac op uden ial policies, o p o ide
a sou ce o exogenous a ia ion. The e a e se e al easons why a coun y may adop o
s eng hen i s mac op uden ial policies, d awing on he expe ience o i s neighbou ing coun ies.
Fo example, economic in eg a ion may lead a coun y o s eng hen i s mac op uden ial egu-
la ions due o p essu e om i s pee s, o educe sys emic isks. The sp ead o mac op uden ial
policies in he Eu ozone, unde he ECB’s in luence, is a s iking example. The same applies in
A ica, whe e mac op uden ial policies ha e sp ead among he membe coun ies o mone a y
unions o s eng hen he esilience o banking sys ems in esponse o economic shocks. In Asia,
many coun ies adop ed o s eng hened hei mac op uden ial ools a e he 1997 Asian
inancial c isis o p e en u u e inancial ins abili y. Ou iden i ying hypo hesis is he e o e ha
mac op uden ial egula ions in egional pee coun ies can play an impo an ole in
s eng hening domes ic mac op uden ial policies, wi hou di ec ly a ec ing p i a e sec o
in es men in he domes ic coun y – condi ional on he ec o o con ols. Howe e , we ecog-
nise ha i mac op uden ial policies success ully educe sys emic isks, hey may (indi ec ly)
a ec egional economic pe o mance. To add ess his limi a ion o ou IV me hodology, we
include egional banking c ises and egional economic g ow h as addi ional con ols in ou main
eg ession o obus ness.
4.2 Econome ic speci ica ion and main esul s
We es ima e he e ec o mac op uden ial policies on p i a e domes ic in es men based on he
ollowing econome ic speci ica ion
Yi = αi + βXi + ηZi + µi + ψ + εi (1)
whe e Yi ep esen s p i a e domes ic in es men (as a pe cen age o GDP) o a coun y i in
yea . Xi, is he mac op uden ial policy index, and Zi is he se o con ol a iables o he baseline
model. µi and ψ deno e coun y and ime- ixed e ec s, espec i ely. Coun y- ixed e ec s
IDOS Discussion Pape 3/2025
9
cap u e unobse ed coun y-speci ic and ime-in a ian ac o s; and ime- ixed e ec s accoun
o common ime- a ying shocks co ela ed wi h mac op uden ial policies and p i a e domes ic
in es men . εi is he usual esidual e o e m. Column [1] o Table 2 epo s he main esul s,
including he baseline model con ols and coun y and yea ixed e ec s. The coe icien on he
mac op uden ial policy index is nega i e and signi ican a he 1% h eshold, sugges ing ha
mac op uden ial policies a e associa ed wi h a signi ican d op in p i a e sec o in es men in
de eloping coun ies. Speci ically, a one-uni inc ease in he Ce u i e al. (2017) index is
associa ed wi h a d op o oughly 1.4 pe cen age poin s in p i a e domes ic in es men . Mo e
impo an ly, s a is ical es s show ha his esul is no due o a lack o ele ance o he
ins umen . Indeed, he Kleibe gen-Paap F-s a is ic o he baseline model gi es a alue well
abo e he alue o 10 o he ule o humb o S aige and S ock (1997), sugges ing ha , in ou
exac ly iden i ied model, he weak ins umen bias is a he low (Ang is & Pischke, 2009).
Ano he way o check he ins umen ’s ele ance is o e e o he esul s o he i s -s age
equa ion epo ed in Table 3. These esul s show a posi i e and signi ican e ec o egional
mac op uden ial policies on domes ic mac op uden ial policies, ein o cing he ele ance o he
ins umen . Las ly, wi h ega d o he baseline model con ols, he esul s show ha employmen ,
co up ion con ol and pe capi a income enhance in es men , while go e nmen du abili y is
nega i ely associa ed. Ano he impo an ques ion ela es o he economic size o he main
es ima es. In ou sample and o e ou s udy pe iod, we epo an a e age p i a e domes ic
in es men o 13% o GDP. Consequen ly, he main esul s sugges ha o an a e age coun y
in he sample, a one-uni inc ease in he Ce u i e al. (2017) index — o MPI — educes p i a e
in es men by a ound 11%, indica ing an economically signi ican e ec .
5 Robus ness checks
In he p e ious sec ion, we ha e es ablished a s a is ically and economically signi ican nega i e
e ec o mac op uden ial policies on p i a e domes ic in es men in de eloping coun ies. In his
sec ion, we conduc a se ies o obus ness es s. Speci ically, we e-es ima e Equa ion 1 by
including addi ional con ols, conside ing al e na i e subsamples and measu es o mac o-
p uden ial policies, and using he Sys em-GMM me hod, espec i ely.
5.1 Al e na i e speci ica ions and addi ional con ols
Fi s , we conside al e na i e speci ica ions by lagging he ins umen al a iable by one and wo
yea s espec i ely. This allows us o accoun o po en ial lags in con e gence dynamics
be ween egional and domes ic mac op uden ial policies. In o he wo ds, we conside ha he
egional di usion o e o ms could inc ease o e ime, wi h po en ial in luence on he baseline
model es ima es. Al hough he new coe icien s epo ed in Columns [2] and [3] o Table 2
inc ease sligh ly compa ed o ha o he baseline model, hei magni ude emains e y simila .
The same applies o he coe icien s o he i s -s age eg ession epo ed in Columns [2] and
[3] o Table 3. Second, ins ead o he KOF Globalisa ion Index used in he main model, in
Column [4], we conside he ade openness a iable om he Wo ld Bank’s WDI da abase,
measu ed as he sum o expo s and impo s as a pe cen age o GDP. The esul s hold.
We u he augmen he baseline model using addi ional con ols, which migh a ec ou
dependen a iable. Fi s , in addi ion o he ade globalisa ion a iable included in he main
model, we also con ol o inancial openness. Nex , we include he e ms o ade o cap u e he
po en ial in luence o cos ly shocks on p i a e domes ic in es men . We u he conside
mone a y ac o s such as he in la ion a e, he exchange a e egime and he eal e ec i e
exchange a e. We expec in la ion o educe domes ic in es men , ia he mac oeconomic
unce ain y and ola ili y i gene a es (e.g. see Bambe e al., 2024; Bloom e al., 2007; Dixi &
IDOS Discussion Pape 3/2025
10
Pindyck, 1994; Lucas J , 1967; Nickell, 1974). The exchange a e egime may equally be a key
de e minan o in la ion pe o mance and mac oeconomic ola ili y, wi h po en ially impo an
side e ec s on p i a e sec o in es men decisions (e.g. see Edwa ds, 1993; Ghosh e al., 1996).
App ecia ions in he eal e ec i e exchange a e may dampen in es men by exace ba ing
compe i ion challenges in he expo sec o . Fou h, we conside public in es men and
emi ances, since hey can exe signi ican upwa d o downwa d e ec s on p i a e domes ic
in es men (Bo ensz ein e al., 1998; Bui e , 1977; Chau e & Jacolin, 2017; Dash, 2023; F y,
1993). Fi h, we include na u al esou ces, which a e p obably a key de e minan o p i a e-
sec o in es men , pa icula ly in he manu ac u ing sec o . Indeed, he li e a u e shows ha
na u al esou ce booms in highly dependen coun ies end o gene a e exchange a e
app ecia ions, wi h ad e se e ec s on he compe i i eness o he non-ex ac i e sec o (see
Co den, 1984 and Sachs & Wa ne , 2001 o pionee ing wo k on he li e a u e dealing wi h Du ch
disease). Six h, we complemen he ins i u ional a iables o he baseline model by including he
V-Dem (Va ie ies o Democ acy) “p ope y igh s” index and he QOG (Quali y o Go e nmen )
“poli ical p essu es and con ols on he media index”. Se en h, we conside o he e o ms ha
could also impac domes ic in es men , including IMF p og ammes and an index o economic
eedom ha includes 12 quan i a i e and quali a i e a iables – om p ope y igh s o inancial
eedom – cap u ing go e nmen e o ms.8 Las ly, o educe he isk ha he e ec obse ed in
he baseline model is biased by he po en ial in luence o he ins umen on egional
pe o mance, we accoun o egional GDP g ow h and egional banking c ises.
Columns [5]-[18] o Table 2 include he new con ols independen ly, and he las column
conside s hem in he same eg ession. The new coe icien s o he a iable o in e es emain
s ongly obus . The same holds o he baseline model con ols, excep o ins i u ional
a iables, p obably due o he high co ela ion wi h he new ins i u ional indices. Rega ding he
new con ols, we ind ha capi al openness inc eases p i a e domes ic in es men , while na u al
esou ces a e nega i ely associa ed. Mo e impo an ly, accoun ing o IMF p og ammes and
economic eedom does no al e ou esul s, sugges ing ha he e ec s ob ained a e likely due
o mac op uden ial policies and no o al e na i e economic e o ms o policies. Simila ly,
accoun ing o egional economic g ow h and egional banking c ises does no a ec ou esul s,
which is eassu ing, as i ein o ces he exclusion condi ion s ipula ed abo e.
8 The e ms o ade, in la ion, na u al esou ces, and emi ances a e om he Wo ld Bank’s Wo ld
De elopmen Indica o s (WDI) da abase. Capi al openness is om Chinn and I o (2008) and a ies
app oxima ely be ween -2 and 2 (highe alues indica e g ea e openness). The exchange a e egime
is compu ed based on Ilze zki e al. (2019)’s classi ica ion. We cons uc a dummy equal o 1 i coun y
i is classi ied as ha ing a ixed exchange a e egime in he yea , and o 0 o he wise. The eal e ec i e
exchange a e a iable is om Da as (2012), 2007 is he e e ence yea wi h a base o 100. Public
in es men is p oxied by go e nmen g oss ixed capi al o ma ion (as a pe cen age o GDP) and is
om he IMF’s In es men and Capi al S ock da ase . IMF p og ammes a e cap u ed by a dummy
a iable equal o 1 i a coun y i has bene i ed om any ype o IMF-suppo ed p og amme in he yea
, and o 0 o he wise. The a iable is om D ehe (2006b). The V-Dem p ope y igh s index anges
be ween 0.1 and 0.9 in ou sample, whe e highe alues indica e be e pe o mance. The QOG index
o poli ical p essu es and con ols on he media index anges be ween 2 and 40 in ou sample, whe e
2 indica es be e pe o mance. The economic eedom/ e o m index can ange om 0 o 100 and is
om he He i age Founda ion. Following Fu ce i and Loungani (2018), we iden i y majo e o m
episodes using a dummy equal o 1 when, o a gi en coun y a a gi en ime, he annual change in
he index exceeds he o e all a e age annual change ac oss all obse a ions by wo s anda d
de ia ions, and 0 o he wise. Banking c ises a e cap u ed by a dummy equal o 1 du ing imes o c isis,
and 0 o he wise, and a e om Lae en and Valencia (2020).
Table 2: The e ec o mac op uden ial policies on p i a e domes ic in es men
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
MPI
-1.430
***
-2.071
***
-3.104
***
-1.327
***
-1.494
***
-1.600
***
-1.397
***
-1.701
***
-1.560
***
-1.407
***
-1.499
***
-1.526
***
-1.384
***
-1.675
***
-1.462
***
-1.416
***
-1.025
**
-1.410
***
-1.597
***
(0.446)
(0.566)
(0.819)
(0.416)
(0.485)
(0.477)
(0.451)
(0.493)
(0.438)
(0.432)
(0.438)
(0.451)
(0.434)
(0.490)
(0.466)
(0.442)
(0.445)
(0.449)
(0.526)
Log.
Employmen
10.406*** 11.298*** 14.588*** 11.861*** 12.777*** 11.006*** 10.825*** 9.604*** 11.528*** 10.329*** 11.023*** 10.778*** 11.580*** 9.825*** 13.016*** 10.446*** 10.591*** 10.316*** 14.964***
(2.248)
(2.593)
(3.547)
(2.257)
(2.294)
(2.402)
(2.342)
(2.441)
(2.335)
(2.213)
(2.322)
(2.247)
(2.252)
(2.399)
(2.268)
(2.259)
(2.097)
(2.248)
(2.509)
T ade
globalisa ion
-0.007 -0.011 -0.026 -0.018 -0.012 -0.008 0.003 0.001 -0.006 -0.002 0.008 -0.012 0.003 -0.013 -0.006 -0.005 -0.006 0.027
(0.017)
(0.018)
(0.023)
(0.018)
(0.017)
(0.017)
(0.017)
(0.018)
(0.017)
(0.017)
(0.017)
(0.017)
(0.017)
(0.017)
(0.017)
(0.016)
(0.017)
(0.020)
Co up ion
con ol
0.515*** 0.591*** 0.457* 0.408** 0.387** 0.515*** 0.434** 0.623*** 0.508*** 0.516*** 0.189 0.534*** 0.348* 0.604*** 0.431** 0.516*** 0.487*** 0.517*** 0.054
(0.182)
(0.202)
(0.240)
(0.175)
(0.190)
(0.188)
(0.183)
(0.193)
(0.184)
(0.182)
(0.190)
(0.183)
(0.191)
(0.200)
(0.182)
(0.182)
(0.175)
(0.182)
(0.206)
Log.
Go e nmen
du abili y
-0.297*** -0.304** -0.285* -0.237** -0.343*** -0.285** -0.367*** -0.249** -0.151 -0.298*** -0.181 -0.285*** -0.314*** -0.257** -0.352*** -0.303*** -0.270** -0.295*** -0.077
(0.109)
(0.119)
(0.147)
(0.119)
(0.116)
(0.113)
(0.110)
(0.118)
(0.114)
(0.108)
(0.116)
(0.110)
(0.116)
(0.118)
(0.110)
(0.109)
(0.105)
(0.109)
(0.143)
Log. GDP pe
capi a
5.970*** 6.626*** 8.006*** 5.252*** 5.967*** 6.153*** 5.971*** 6.174*** 5.170*** 5.927*** 7.579*** 5.855*** 6.254*** 6.160*** 5.933*** 5.968*** 5.800*** 5.972*** 6.916***
(1.183)
(1.293)
(1.663)
(1.393)
(1.209)
(1.199)
(1.221)
(1.201)
(1.331)
(1.189)
(1.267)
(1.192)
(1.242)
(1.193)
(1.186)
(1.182)
(1.124)
(1.176)
(1.419)
Obse a ions
1412
1412
1329
1316
1350
1398
1373
1330
1395
1412
1380
1412
1361
1330
1378
1412
1412
1412
1178
R-squa ed 0.785 0.752 0.681 0.793 0.783 0.777 0.783 0.782 0.780 0.786 0.791 0.783 0.787 0.783 0.785 0.786 0.802 0.786 0.806
Kleibe gen-
Paap LM s a
(p- alue)
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Kleibe gen-
Paap F-s a
58.51 44.35 26.01 68.13 50.28 54.53 57.21 51.13 64.19 63.88 58.79 58.28 62.74 51.32 54.09 59.29 54.81 58.29 43.84
This able epo s es ima es o he e ec o mac op uden ial policies on p i a e domes ic in es men , using he ins umen al a iables (IV). The ins umen (Con igui y) is he a e age mac op uden ial
policy index in egional coun ies. Column [1] displays he main esul s. In Columns [2] and [3], we conside al e na i e speci ica ions by lagging he ins umen al a iable by one and wo yea s, espec i ely.
In Column [4], ins ead o he KOF Globalisa ion Index used in he main model, we conside he ade openness a iable om he Wo ld Bank’s WDI da abase, measu ed as he sum o expo s and
impo s as a pe cen age o GDP. Columns [5]-[18] include he ollowing addi ional con ols independen ly, and he las column conside s hem in he same eg ession: capi al openness, log. e ms o
ade, in la ion, exchange a e egime, eal e ec i e exchange a e, public in es men , emi ances, log. na u al esou ces, p ope y igh s, poli ical p essu es and con ols on media, IMF p og ammes, economic
eedom/ e o ms, egional economic g ow h and egional banking c ises. Rega ding he new con ols, we ind ha capi al openness inc eases p i a e domes ic in es men , while na u al esou ces a e nega i ely
associa ed. Robus s anda d e o s a e in pa en heses. All eg essions include he cons an , no epo ed in he able. * p < 0.1, ** p < 0.05, *** p < 0.01
Table 3: The e ec o mac op uden ial policies on p i a e domes ic in es men : i s s age esul s
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
Con igui y
0.851
***
0.941
***
0.803
***
0.828
***
0.852
***
0.851
***
0.893
***
0.884
***
0.859
***
0.839
***
0.893
***
0.856
***
0.821
***
0.854
***
0.859
***
0.852
***
0.861
***
(0.107)
(0.110)
(0.109)
(0.108)
(0.108)
(0.114)
(0.107)
(0.106)
(0.108)
(0.106)
(0.108)
(0.115)
(0.107)
(0.107)
(0.112)
(0.107)
(0.124)
Log. Employmen
0.874
0.945
1.586
**
0.664
0.989
0.961
0.958
0.508
0.492
0.772
0.871
0.918
0.853
0.367
1.175
*
0.889
0.888
0.869
0.003
(0.659)
(0.673)
(0.711)
(0.680)
(0.641)
(0.667)
(0.678)
(0.668)
(0.667)
(0.644)
(0.675)
(0.653)
(0.674)
(0.655)
(0.637)
(0.667)
(0.663)
(0.662)
(0.662)
T ade globalisa ion
-0.004
-0.006
-0.009
**
-0.007
*
-0.005
-0.003
-0.001
-0.006
-0.004
-0.003
-0.002
-0.002
-0.001
-0.006
-0.004
-0.004
-0.004
0.000
(0.004)
(0.004)
(0.005)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.004)
(0.005)
Co up ion con ol
0.051
0.066
0.045
0.009
0.037
0.053
0.037
0.054
0.059
0.054
0.039
0.053
0.060
0.076
*
0.041
0.052
0.051
0.051
0.052
(0.042)
(0.043)
(0.045)
(0.044)
(0.044)
(0.043)
(0.042)
(0.041)
(0.041)
(0.041)
(0.043)
(0.042)
(0.045)
(0.042)
(0.042)
(0.042)
(0.042)
(0.042)
(0.046)
Log. Go e nmen
du abili y
0.013 0.011 0.014 -0.044 0.012 0.015 0.001 0.028 -0.020 0.011 0.005 0.015 0.014 0.027 0.009 0.012 0.014 0.014 -0.031
(0.034)
(0.034)
(0.035)
(0.038)
(0.036)
(0.035)
(0.035)
(0.037)
(0.036)
(0.034)
(0.037)
(0.034)
(0.037)
(0.036)
(0.035)
(0.035)
(0.035)
(0.034)
(0.045)
Log. GDP pe capi a
0.326
0.373
0.498
*
0.765
**
0.312
0.353
0.327
0.278
0.529
*
0.262
0.424
0.304
0.332
0.278
0.333
0.327
0.326
0.327
0.550
*
(0.259)
(0.261)
(0.285)
(0.302)
(0.262)
(0.259)
(0.267)
(0.263)
(0.276)
(0.264)
(0.271)
(0.259)
(0.271)
(0.262)
(0.260)
(0.259)
(0.259)
(0.259)
(0.300)
-1 Con igui y
0.768
***
(0.111)
-2 Con igui y
0.624
***
(0.117)
Obse a ions 1412 1412 1329 1316 1350 1398 1373 1330 1395 1412 1380 1412 1361 1330 1378 1412 1412 1412 1178
R-squa ed
0.870
0.868
0.870
0.872
0.875
0.870
0.868
0.879
0.872
0.872
0.873
0.871
0.871
0.880
0.874
0.870
0.870
0.870
0.890
This able epo s he esul s o he i s s age IV es ima ion o Table 2. Robus s anda d e o s a e in pa en heses. All eg essions include he cons an , no epo ed in he able. * p < 0.1, ** p < 0.05, ***
p < 0.01
IDOS Discussion Pape 3/2025
13
5.2 Al e na i e subsamples and measu es
Nex , we explo e he sensi i i y o ou main esul s using al e na i e subsamples. We s a by
e-es ima ing ou baseline speci ica ion, excluding hype in la ion pe iods, de ined in Lin and Ye
(2009) as coun y-yea obse a ions wi h in la ion a es o 40% o mo e.9 The unde lying
in ui ion is ha , since such high in la ion a es would e lec signi ican economic imbalances,
his could a ec p i a e sec o in es men independen ly o mac op uden ial policies. Along he
same lines, we exclude om he sample he yea s 2008–2009, gi en he economic imbalances
esul ing om he global inancial c isis. Thi d, ou sample includes 14 agile s a es, classi ied
by he IMF as coun ies wi h s ong economic, ins i u ional and s uc u al ulne abili ies hinde ing
hei economic de elopmen . Gi en ha hese cases exhibi signi ican ly di e en cha ac e is ics
om he es o he sample, i is wo h conside ing whe he hei inclusion in he s udy a ec s
he main esul s. Hence, o obus ness, we exclude hem om he main sample. Fou h, we
explo e he sensi i i y o ou esul s o ou lie s, by excluding coun y-yea obse a ions wi h
alues abo e he 95 h pe cen ile o he sample, o he a iable o in e es and he dependen
a iable espec i ely. Fi h, we e-es ima e he baseline model using exclusi ely coun ies ha
ha e implemen ed a leas one mac op uden ial ins umen du ing he s udy pe iod. In o he
wo ds, we exclude om he main sample he eigh coun ies o which he mac op uden ial policy
index is ze o o e he en i e s udy pe iod hus ocusing solely on wi hin-coun y a ia ion.10 In
all cases (Columns [1]-[6] o Table A1 in he appendix), he esul s a e e y simila o ou
baseline esul s, indica ing ha ou lie s o speci ic subsamples do no d i e ou esul s.
Fu he mo e, he Kleibe gen-Paap F-s a is ics and he esul s o he i s -s age equa ion suppo
he ele ance and he ins umen in all cases.
In Column [7] o Table A1, we u he conside an al e na i e measu e o he ins umen al
a iable, weigh ing he ins umen al a iable by annual GDP o assign g ea e in luence o
egional neighbou s wi h la ge economic size. Nex , he mac op uden ial policy da a used so
a is d awn om he Ce u i e al. (2017) da abase, which co e s 162 ad anced and de eloping
coun ies om 2000 o 2017. The in eg a ed Mac op uden ial Policy (iMaPP) da abase,
published by he In e na ional Mone a y Fund (Alam e al., 2019), also p o ides a summa y
measu e o mac op uden ial ac ions o a panel o 135 coun ies o e 1990–2021. The main
ad an age o he Ce u i e al. index is ha i co e s a la ge sample o coun ies, enabling
g ea e in e na ional compa abili y, albei o e a ela i ely sho e pe iod han he IMF da abase.
The iMaPP da ase includes 17 main ca ego ies o mac op uden ial ools, classi ying hem in o
no ac ion (0), igh ening (+1), o loosening (-1). Based on his da abase, we conduc addi ional
obus ness checks (Table A1), conside ing he iMaPP da ase . In Column [8], we use he sum
o he 17 ins umen s. In Column [9], we ollow Se e and Yücel (2022) and compu e igh ening
episodes h ough a dummy a iable equal o 1 i he numbe o igh ening episodes ac oss
mon hs is g ea e han he numbe o easing episodes in a gi en yea , and 0 o he wise. In bo h
cases, he coe icien o he a iable o in e es is nega i e and signi ican , al hough he
magni ude o he coe icien inc eases sligh ly compa ed o ha o he main model.
In Appendix A, we pe o m addi ional obus ness es s, based on he wo-s ep Sys em-GMM
me hod and using a iables agg ega ed in o non-o e lapping h ee-yea a e age o educe
s a iona i y issues, espec i ely. The esul s emain s able.
9 The esul s emain obus when conside ing al e na i e h esholds, such as in la ion a es abo e 50%,
70%, 90%, o 100%. These esul s a e no epo ed bu a e a ailable on eques .
10 These coun ies include: Bu kina Faso, Guyana, Madagasca , Mali, Nige , Senegal, Togo, and Venezuela.
IDOS Discussion Pape 3/2025
14
6 He e ogenei y
This sec ion conduc s a se ies o he e ogenei y analyses, i s dis inguishing he e ec o mac o-
p uden ial policies be ween bo owe - a ge ed and inancial-ins i u ion- a ge ed ins umen s.11
We could expec a highe impac o policies a ge ing bo owe s on in es men , as hey ha e
he mos es ic i e e ec s on c edi and inancial inclusion (see ou discussion in Sec ion 2).
Howe e , as highligh ed by Ce u i e al. (2017), e en ools a ge ed a inancial ins i u ions end
o lowe c edi g ow h, no ably in eme ging and de eloping ma ke s – d i en by ools such as
dynamic p o isioning, le e age a ios, coun e -cyclical equi emen s, ax measu es, in e -
connec ion and concen a ion limi s. Indeed, al hough hese ins umen s a ge inancial ins i u-
ions o mi iga e sys emic isk, hey la gely in luence in e media y ools o egula e c edi . Along
he same lines, using da a on 900,000 i ms om 48 coun ies om 2003–2011, Ayyaga i e al.
(2018) ind ha young i ms ha e lowe in es men and sales g ow h a e implemen a ion o
bo h bo owe - a ge ed and inancial ins i u ion- a ge ed policies. The esul s epo ed in
Columns [1] and [2] o Table 4 show ha bo h ins umen s a ge ing bo owe s and inancial
ins i u ions signi ican ly educe in es men , almos o oughly he same scale.
Nex , we examine he ole o se e al mac oeconomic, ins i u ional, and s uc u al ac o s. Mo e
p ecisely, we conside ou main model (Equa ion 1) and augmen i wi h se e al in e ac i e
e ms. Fi s , we in e ac he mac op uden ial policy index wi h he business cycle, app oxima ed
by annual GDP g ow h and he ou pu gap, espec i ely.12 Since c edi igh ening can be mo e
p onounced du ing economic down u ns (Lown & Mo gan, 2006), one may expec he nega i e
e ec o mac op uden ial policies on in es men o be less p onounced du ing he expansion
phase o he business cycle. The esul s epo ed in Columns [3] and [4] seem o con i m ou
hypo hesis. Second, we conside inancial and mone a y ac o s, namely: he le el o inancial
de elopmen (p oxied by domes ic c edi o he p i a e sec o ), inancial openness and he
exchange a e egime.13 The po en ial e ec o mac op uden ial policies on p i a e sec o
in es men wi h ega d o he le el o inancial de elopmen is no so clea -cu . As g ea e
inancial de elopmen comes wi h g ea e economic de elopmen , and p obably wi h be e
ins i u ional amewo ks, i can be a gued ha inancially mo e de eloped coun ies a e mo e
likely o s eng hen hei mac op uden ial policies e ec i ely. On he o he hand, ollowing he
pe spec i e o Ce u i e al. (2017) pe spec i e, we can conside ha a mo e de eloped inancial
sys em also implies g ea e sophis ica ion, making he applica ion o mac op uden ial policies
mo e complex, which can weaken hei e ec i eness. Rega ding inancial openness, we can
expec bo owe s in mo e open economies o success ully ci cum en mac op uden ial policies,
by inding ways o access o he sou ces o inancing, such as non-bank o c oss-bo de banking
ac i i ies. In his case, he e ec o mac op uden ial policies on p i a e sec o in es men would
be mo e limi ed, gi en he po en ial subs i u ion be ween domes ic c edi and o he sou ces o
inancing. Rega ding he exchange a e egime, Ce u i e al. (2017) no e ha i is mo e
11 Bo owe - a ge ed ins umen s include loan- o- alue a ios and deb - o-income a ios. Financial-
ins i u ion- a ge ed ins umen s include dynamic loan-loss p o isioning; coun e cyclical capi al bu e
equi emen ; le e age a io; capi al su cha ges on sys emically impo an inancial ins i u ions; limi s
on in e bank exposu es; concen a ion limi s; limi s on o eign cu ency loans; ese e equi emen
a ios, limi s o domes ic cu ency loans; and le y/ ax on inancial ins i u ions.
12 We compu e he ou pu gap by ex ac ing po en ial ou pu om obse ed eal GDP, using he Hod ick–
P esco il e .
13 We ha e also conside ed po en ial he e ogenei y wi h ega d o he cen al bank in e es a e, since
mone a y policy decisions ha e signi ican demand e ec s as well. The in e ac i e e m does no
sugges any he e ogenei y be ween mac op uden ial policies and he cen al bank in e es a e.
Howe e , hese esul s should be in e p e ed wi h cau ion, as mone a y policy decisions may
hemsel es be s ongly endogenous o mac op uden ial policies (see Kim & Meh o a, 2018 o a
comp ehensi e discussion).
IDOS Discussion Pape 3/2025
15
challenging o economies o con ol o e all c edi in mo e lexible exchange a e egimes,
pa icula ly gi en he impac o exchange a e app ecia ions o dep ecia ion on capi al
mo emen s. This sugges s ha he e ec o mac op uden ial policies may be mo e limi ed in
lexible exchange a e egimes. The esul s epo ed in Columns [5]-[7] e eal ha he nega i e
e ec o mac op uden ial policies on in es men is less p onounced in coun ies wi h mo e
de eloped inancial sys ems and lexible exchange a e egimes. Howe e , no he e ogenei y
seems o eme ge wi h ega d o inancial openness. Thi d, Column [8] examines po en ial
he e ogenei y in mac op uden ial policies acco ding o he size o he in o mal sec o , wi h he
idea ha policies a ge ing bo owe s could lead hem o shi owa ds in o mal inancial se ices
as an al e na i e o m o inancing.14 The esul s seem o co obo a e his hypo hesis. In
Columns [9]-[11], we c oss he mac op uden ial policy index wi h pe capi a income – using a
dummy a iable based on de ia ions om he sample mean – and he quali y o ins i u ions –
p oxied by co up ion con ol and he le el o democ acy, espec i ely. We ind s ong e idence
ha he ad e se e ec o mac op uden ial policies on in es men is mi iga ed in economically
and ins i u ionally mo e de eloped coun ies. These esul s can be aligned wi h ou indings and
discussion on he deg ee o inancial de elopmen . In he las column, we di e en ia e he e ec
based on he pe iods be o e and a e he 2008–2009 global inancial c isis, in he idea ha he
impac o mac op uden ial policies may ha e been mo e p onounced a e he c isis, as hese
policies we e signi ican ly in ensi ied du ing ha ime. Howe e , one may equally expec he
e ec o be less p onounced a e he c isis, as he pos -c isis in ensi ica ion o mac op uden ial
policies may also be associa ed wi h g ea e complexi y in he applica ion o mac op uden ial
ools and he inancial en i onmen , hus hinde ing he e ec i e and igo ous implemen a ion o
hese ools. The esul s appea o align wi h he second hypo hesis, indica ing a diminished
impac o mac op uden ial policies in he pos -c isis pe iod. Finally, Table A4 (in he appendix),
which epo s he esul s o he i s -s age equa ion, does no in alida e he ele ance o he
ins umen , ega dless o he he e ogenei y analysed.
14 The in o mal sec o al is om he Wo ld Bank’s P ospec s G oup (Elgin e al., 2021) which measu es
he in o mal economic ac i i y using Mul iple Indica o s Mul iple Causes (MIMIC) model-based
es ima es o in o mal ou pu .
IDOS Discussion Pape 3/2025
16
Table 4: The e ec o mac op uden ial policies on p i a e domes ic in es men :
he e ogenei y
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Bo owe -
Ta ge ed
Ins umen s
-1.464**
(0.695)
Financial
Ins i u ion-
Ta ge ed
Ins umen s
-1.977***
(0.654)
MPI
-1.406***
-1.750***
-3.246***
-1.495***
-4.554***
-2.610***
-2.813***
-3.235***
-6.271***
-2.599***
(0.434)
(0.544)
(1.048)
(0.471)
(1.518)
(0.778)
(0.833)
(1.204)
(2.232)
(0.837)
MPI x Annual
GDP g ow h
0.065***
(0.018)
MPI x Ou pu
gap
0.661***
(0.184)
MPI x
Financial
de elopmen
0.032***
(0.009)
MPI x Capi al
openness
-0.002
(0.076)
MPI x
Exchange
a e egime
0.397***
(0.137)
MPI x
In o mal
sec o
0.060***
(0.019)
MPI x High
income
1.883***
(0.521)
MPI x
Democ acy
0.557**
(0.221)
MPI x
Co up ion
con ol
2.167***
(0.777)
MPI x Pos
2008-09 c isis
1.018***
(0.344)
Obse a ions
1412
1412
1412
1412
1222
1350
1330
1373
1412
1412
1412
1412
R-squa ed
0.811
0.774
0.796
0.785
0.782
0.783
0.726
0.778
0.752
0.747
0.697
0.745
Kleibe gen-
Paap LM s a
(p- alue)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Kleibe gen-
Paap F-s a
123.55 44.67 67.46 49.39 18.40 53.55 26.07 69.23 28.51 26.90 15.29 27.99
No es: In Columns [3]-[11] ec o X a iables in isola ion (wi hou in e ac ion wi h mac op uden ial policies) and con ols a e
included bu no epo ed o he sake o space. Robus s anda d e o s a e in pa en heses. All eg essions include he
cons an , no epo ed in he able. * p < 0.1, ** p < 0.05, *** p < 0.01
IDOS Discussion Pape 3/2025
17
7 Mechanisms
So a , we ha e mainly suppo ed ou po en ial ansmission channels h ough heo e ical
discussions. In his sec ion, we a emp o examine hem empi ically, d awing on exis ing wo k
and he li e a u e dealing wi h causal media ion analysis ( o ins ance, see Acemoglu e al.,
2019; Ape i & Edoh, 2023; Bambe, 2023; Bambe e al., 2024; Imai e al., 2010). Fi s , we
es ima e he e ec o mac op uden ial policies on he po en ial channel, accoun ing o key
po en ial de e minan s o he channel. The indings om he i s h ee columns o Table 5
indica e ha mac op uden ial policies signi ican ly educe c edi g ow h, inancial inclusion and
he p obabili y o banking c ises, highligh ing he impo ance o hese ac o s as po en ial ans-
mission channels.15 Second, in columns [4]-[6] we e-es ima e he e ec o mac op uden ial
policies on p i a e domes ic in es men (Equa ion 1), including he po en ial channel among he
ec o o con ols. The esul s in Column [4] a e pa icula ly no ewo hy: when c edi g ow h is
included, he e ec o mac op uden ial policies diminishes and becomes s a is ically
insigni ican . Mo eo e , he coe icien o c edi g ow h is posi i e and signi ican , indica ing ha
he la e is a key ansmission channel. In Column [5], accoun ing o inancial inclusion sligh ly
educes he e ec o mac op uden ial policies compa ed o he coe icien o he baseline model
(Table 2, Column [1]). S ill, he coe icien o he inancial inclusion index is posi i e and
signi ican , sugges ing ha he la e is also an impo an channel. In he las column, al hough
he esul s indica e ha banking c ises educe in es men , he inclusion o his a iable does no
signi ican ly al e he coe icien o mac op uden ial policies, whose e ec emains close o ha
o he baseline model. In o he wo ds, while inancial s abili y – pa icula ly banking c ises – may
be a channel h ough which mac op uden ial policies can p omo e domes ic in es men , he
nega i e e ec o hese policies on in es men , ia he educ ion in c edi supply and inancial
inclusion, appea s o ou weigh hei po en ially bene icial impac ia inancial s abili y. In sho ,
hese esul s seem o suppo ou heo e ical hypo heses, p o iding e idence ha he educ ion
in c edi supply and inancial inclusion esul ing om mac op uden ial policies a e ele an
channels h ough which hese policies a ec p i a e sec o in es men in de eloping coun ies
– hough c edi g ow h seems o be he mos dominan channel.
15 We ollow p e ious s udies (e.g. see Bozku e al., 2018; Ozili, 2022) and compu e a composi e index
o cap u e inancial inclusion. We conside wo dimensions o inancial inclusion, using da a om he
Financial Access Su ey (IMF). The access dimension includes he numbe o comme cial bank
b anches pe 100,000 adul s and he numbe o deposi accoun s wi h comme cial banks pe 1,000
adul s. The usage dimension includes ou s anding loans om comme cial banks (pe cen o GDP).
We u he conside he a ailabili y dimension, including he numbe o ATMs pe 1,000 km2 and
100,000 adul s. The index is compu ed ollowing Ande son (2008), i.e., using gene alised leas
squa es es ima o s ha accoun o a iables wi h missing da a, gi ing hem less weigh . Since he
de e minan s o he channels conside ed may di e om hose o p i a e in es men , in Columns [1]
and [2] (Table 5) we d aw on he li e a u e on he de e minan s o inancial inclusion ( o ins ance, see
Bozku e al., 2018) and conside he ollowing con ol a iables: pe capi a GDP, he le el o
educa ion, he quali y o ins i u ions (p oxied by go e nmen du abili y), he employmen a e, he size
o he in o mal sec o , and inancial sec o e o ms. In Column [3], we conside he ollowing de e mi-
nan s o banking c ises: in la ion, lagged GDP g ow h, inancial de elopmen , ade and inancial
globalisa ion.
IDOS Discussion Pape 3/2025
24
Nickell, S. (1974). On he ole o expec a ions in he pu e heo y o in es men . The Re iew o Economic
S udies, 41(1), 1–19.
Nickell, S. (1981). Biases in dynamic models wi h ixed e ec s. Econome ica: Jou nal o he Econome ic
Socie y, 1417–1426.
Osinski, J., Seal, K., & Hoogduin, M. L. (2013). Mac op uden ial and mic op uden ial policies: owa d
cohabi a ion. In e na ional Mone a y Fund.
Ozili, P. K. (2022). Financial inclusion and sus ainable de elopmen : an empi ical associa ion. Jou nal o
Money and Business, 2(2), 186–198.
Pe sson, T., & Tabellini, G. (2009). Democ a ic capi al: The nexus o poli ical and economic change.
Ame ican Economic Jou nal: Mac oeconomics, 1(2), 88–126.
Raksmey, U., Lin, C.-Y., & Kakinaka, M. (2022). Mac op uden ial egula ion and inancial inclusion: Any
di e ence be ween de eloped and de eloping coun ies? Resea ch in In e na ional Business and
Finance, 63, 101759.
Reinha , C. (2012). The e u n o inancial ep ession (CEPR Discussion Pape No. DP8947). CEPR.
Reinha , C. M., & Reinha , V. R. (2015). Financial c ises, de elopmen , and g ow h: a long- e m
pe spec i e. The Wo ld Bank Economic Re iew, 29(suppl_1), S53–S76.
Reinha , C. M., & Sb ancia, M. B. (2015). The liquida ion o go e nmen deb . Economic Policy, 30(82),
291–333.
Rich e , B., Schula ick, M., & Shim, I. (2019). The cos s o mac op uden ial policy. Jou nal o In e na ional
Economics, 118, 263–282.
Ri e a-Ba iz, L. A., & Rome , P. M. (1991). Economic in eg a ion and endogenous g ow h. The Qua e ly
Jou nal o Economics, 106(2), 531–555.
Rubio, M., & Ca asco-Gallego, J. A. (2016). The new inancial egula ion in Basel III and mone a y policy:
A mac op uden ial app oach. Jou nal o Financial S abili y, 26, 294–305.
Sachs, J. D., & Wa ne , A. M. (2001). The cu se o na u al esou ces. Eu opean economic e iew, 45(4-
6), 827–838.
Schula ick, M., & Taylo , A. M. (2012). C edi booms gone bus : mone a y policy, le e age cycles, and
inancial c ises, 1870–2008. Ame ican Economic Re iew, 102(2), 1029–1061.
Se e , C., & Yücel, E. (2022). The e ec s o elec ions on mac op uden ial policy. Jou nal o Compa a i e
Economics, 50(2), 507–533.
Shin, H. S. (2013). Adap ing mac o p uden ial app oaches o eme ging and de eloping economies. Dealing
wi h he challenges o mac o inancial linkages in eme ging ma ke s, 17–55.
Shipan, C. R., & Volden, C. (2008). The mechanisms o policy di usion. Ame ican Jou nal o Poli ical
Science, 52(4), 840–857.
S aige , D. O., & S ock, J. H. (1997). Ins umen al a iables eg ession wi h weak ins umen s.
Econome ica, 65(3), 1997, 557–86. JSTOR, 65(3), 557–86.
S aub, L., & Ulb ich , R. (2024). Endogenous unce ain y and c edi c unches. Re iew o Economic
S udies, 91(5), 3085–3115.
Teixei a, A., & Ven e , Z. (2023). Mac op uden ial policy and agg ega e demand. In e na ional Jou nal o
Cen al Banking, 19(4), 1–40.
Teo ell, J., Dahlbe g, S., Holmbe g, S., Ro hs ein, B., Khomenko, A., & S ensson, R. (2016). The quali y
o go e nmen s anda d da ase , e sion Jan16. Uni e si y o Go henbu g: The Quali y o
Go e nmen Ins i u e. h p://www.qog.pol.gu.se doi:10.18157/QoGS dJan16
Tillmann, P. (2015). Es ima ing he e ec s o mac op uden ial policy shocks: A qual a app oach.
Economics Le e s, 135, 1–4.
Van de Gho e, A. (2021). In e ac ions and coo dina ion be ween mone a y and mac op uden ial policies.
Ame ican Economic Jou nal: Mac oeconomics, 13(1), 1–34.
IDOS Discussion Pape 3/2025
25
Appendix A: Fu he obus ness
GMM es ima es
Wi hou alid ex e nal ins umen s, he li e a u e some imes elies on al e na i e econome ic
s a egies, such as he Gene alised Me hod o Momen s (GMM), o mi iga e endogenei y issues.
In addi ion o co ec ing o endogenei y bias using in e nal ins umen s, he GMM me hod also
allows co ec ing o Nickell bias (Nickell, 1981), which is common in dynamic panel models.
The e o e, o obus ness, we ely on he wo-s ep Sys em-GMM me hod o Blundell and Bond
(1998) which combines lagged di e ences and le els o explana o y a iables as ins umen s,
hus imp o ing es ima ion e iciency. The GMM es ima es a e epo ed in Column [1] o Table
A3. The new coe icien o he a iable o in e es emains compa able o ha ob ained om he
main model, suppo ing ou main conclusions.
Th ee-yea window
The s a iona i y o ou a iables is an impo an conside a ion, as i plays a key ole in ensu ing
he eliabili y o ou esul s. Hence, o add ess po en ial s a iona i y issues, we d aw on p e ious
s udies (e.g. see De Haan & S u m, 2017 and Ape i e al., 2025) and e-es ima e ou baseline
model using a iables agg ega ed in o non-o e lapping h ee-yea a e ages. The esul s a e
epo ed in he las column o Table A3 and align wi h ou ini ial conclusions. Fu he mo e, he
magni ude o he new coe icien s is highly consis en wi h hose o he main model, indica ing
ha non-s a iona i y is unlikely o in oduce bias in o ou main es ima es. The esul s o he i s -
s age equa ion (no epo ed, bu a ailable on eques ) also suppo he alidi y o he
ins umen al a iable.
IDOS Discussion Pape 3/2025
26
Table A1: Mac op uden ial policies (MPI) and p i a e domes ic in es men : al e na i e
subsamples and measu es
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
in es in es in es in es in es in es in es in es in es
MPI -1.535*** -1.467*** -
1.576***
-2.477*** -2.019*** -1.378*** -0.918*
(0.464)
(0.437)
(0.473)
(0.622)
(0.563)
(0.509)
(0.540)
Log. Employmen
11.057
***
12.354
***
9.816
***
11.915
***
10.978
***
11.315
***
9.693
***
8.753
***
10.129
***
(2.455)
(2.524)
(2.348)
(2.689)
(2.254)
(2.317)
(2.154)
(2.027)
(2.216)
T ade globalisa ion
-0.016
-0.005
-0.016
0.003
0.003
-0.011
-0.003
-0.003
0.001
(0.017)
(0.019)
(0.017)
(0.018)
(0.017)
(0.018)
(0.016)
(0.019)
(0.019)
Co up ion con ol
0.398
**
0.368
*
0.718
***
0.631
***
0.689
***
0.505
**
0.454
***
0.375
**
0.325
(0.187)
(0.208)
(0.198)
(0.193)
(0.188)
(0.196)
(0.174)
(0.189)
(0.202)
Log. Go e nmen
du abili y
-0.305*** -0.253* -0.306*** -0.127 -0.265** -0.240** -0.290*** -0.076 -0.082
(0.110)
(0.135)
(0.113)
(0.127)
(0.115)
(0.117)
(0.103)
(0.110)
(0.112)
Log. GDP pe capi a
6.077
***
4.942
***
5.744
***
5.986
***
3.405
***
6.022
***
5.446
***
2.343
**
2.197
**
(1.267)
(1.429)
(1.211)
(1.088)
(1.031)
(1.202)
(1.131)
(1.101)
(1.106)
MPI (Alam e al.,
2019)
-5.760*
(3.409)
MPI Tigh ening
-3.045
***
(0.845)
Obse a ions
1356
1181
1244
1367
1357
1293
1412
1174
1174
R-squa ed
0.778
0.764
0.775
0.753
0.757
0.781
0.803
0.805
0.781
Kleibe gen-Paap
LM s a (p- alue)
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Kleibe gen-Paap F-
s a
54.99 61.32 53.10 35.95 36.76 45.24 35.38 31.57 89.80
This able epo s es ima es o he e ec o mac op uden ial policies (MPI) on p i a e domes ic in es men , using
al e na i e subsamples and measu es. In all cases, he ins umen al a iables (IV) is he a e age mac op uden ial
policy index in egional coun ies. In Columns [1]-[3] we e-es ima e ou baseline speci ica ion, excluding hype in la-
ion pe iods; agile s a es; and he 2008–2009 global inal c isis, espec i ely. Columns [4] and [5] exclude ou lie s,
i.e., coun y-yea obse a ions wi h alues abo e he 95 h pe cen ile o he sample, o he a iable o in e es and he
dependen a iable espec i ely. Column [6] uses exclusi ely coun ies ha ha e implemen ed a leas one mac o-
p uden ial ins umen du ing he s udy pe iod. In Column [7], we weigh he ins umen al a iable by annual GDP o
assign g ea e in luence o egional neighbou s wi h la ge economic size. Column [8] conside s he sum o he 17
mac op uden ial ools om he in eg a ed Mac op uden ial Policy (iMaPP) da abase (Alam e al., 2019). Column [9]
uses igh ening episodes h ough a dummy a iable equal o 1 i he numbe o igh ening episodes ac oss mon hs is
g ea e han he numbe o easing episodes in a gi en yea , and 0 o he wise. All eg essions include he cons an , no
epo ed in he able. * p < 0.1, ** p < 0.05, *** p < 0.01
IDOS Discussion Pape 3/2025
27
Table A2: Al e na i e subsamples and measu es: i s s age esul s
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
pi
Con igui y 0.842*** 0.910*** 0.833*** 0.629*** 0.646*** 0.754***
(0.109) (0.112) (0.109) (0.101) (0.102) (0.108)
Log. Employmen 1.168* 0.794 0.837 1.113* 0.978 0.217 0.934 -0.021 0.542*
(0.685) (0.734) (0.682) (0.648) (0.622) (0.696) (0.672) (0.065) (0.289)
T ade globalisa ion -0.004 -0.007 -0.005 -0.003 -0.002 -0.004 -0.003 0.001 0.003
(0.004) (0.005) (0.004) (0.004) (0.004) (0.005) (0.004) (0.001) (0.002)
Co up ion con ol 0.029 0.026 0.095** 0.062 0.078* 0.017 0.061 -0.004 -0.030
(0.042) (0.051) (0.043) (0.040) (0.042) (0.046) (0.042) (0.007) (0.027)
Log. Go e nmen du abili y 0.006 -0.066 0.025 0.073** 0.063** -0.005 -0.001 -0.006** -0.014
(0.034) (0.044) (0.036) (0.029) (0.029) (0.038) (0.033) (0.003) (0.013)
Log. GDP pe capi a 0.395 0.411 0.316 0.233 -0.135 0.179 0.727*** 0.013 0.006
(0.270) (0.301) (0.268) (0.214) (0.226) (0.267) (0.259) (0.024) (0.098)
Weigh ed con igui y 0.482***
(0.078)
Con igui y (Alam e al.,
2019)
1.059***
(0.181)
Con igui y MPI Tigh ening 1.023***
(0.104)
Obse a ions 1356 1181 1244 1367 1357 1293 1412 1174 1174
R-squa ed 0.870 0.865 0.865 0.848 0.881 0.859 0.866 0.598 0.586
This able epo s he esul s o he i s s age IV es ima ion o Table A1. Robus s anda d e o s a e in pa en heses.
All eg essions include he cons an , no epo ed in he able. * p < 0.1, ** p < 0.05, *** p < 0.01
IDOS Discussion Pape 3/2025
28
Table A3: Mac op uden ial policies and p i a e domes ic in es men : Sys em-GMM and
IV- h ee-yea window es ima es
(1)
(2)
Sys em-GMM IV- h ee-yea window
Lag. In es men 0.676**
(0.135)
MPI -1.097* -1.744**
(0.606) (0.823)
Log. Employmen 23.781* 8.877**
(12.998) (4.015)
T ade globalisa ion -0.143 -0.016
(0.093) (0.027)
Co up ion con ol 0.030 1.112**
(0.751) (0.362)
Log. Go e nmen du abili y -0.836 -0.395**
(1.221) (0.201)
Log. GDP pe capi a 7.381** 5.294**
(3.113) (1.876)
Obse a ions 1412 499
R-squa ed 0.8018
The AR (1), AR(2), and Hansen es p- alues epo ed in Column [1] a e espec i ely 0.002, 0.794, and 0.730. We
epo 84 g oups o 42 ins umen s in he i s column. In Column [2], we use a iables agg ega ed in o non-o e lapping
h ee-yea a e ages, o educe s a iona i y issues. Robus s anda d e o s a e in pa en heses. All eg essions include
he cons an , no epo ed in he able. * p < 0.1, ** p < 0.05, *** p < 0.01
IDOS Discussion Pape 3/2025
29
Table A4: He e ogenei y: i s s age esul s
This able epo s he esul s o he i s s age IV es ima ion o Table 4. In Columns [3]-[12] ec o X a iables in
isola ion (wi hou in e ac ion wi h mac op uden ial policies) and he in e ac i e e ms wi h mac op uden ial polices a e
no epo ed o he sake o space. Robus s anda d e o s a e in pa en heses. All eg essions include he cons an ,
no epo ed in he able. * p < 0.1, ** p < 0.05, *** p < 0.01
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Con igui y 0.920***
(0.080)
Log.
Employmen
0.245 0.599 0.837 0.608 1.024* 0.382 0.887** 0.980*** -1.104***
0.215 -1.197***
-0.663
(0.275) (0.500) (0.597) (0.611) (0.572) (0.566) (0.430) (0.306) (0.408) (0.409) (0.240) (0.544)
T ade
globalisa ion
-0.003 -0.002 -0.004 -0.005 0.008* -0.007* 0.001 -0.004* -0.012***
-0.013***
-0.002 -0.006*
(0.002) (0.003) (0.004) (0.004) (0.004) (0.004) (0.002) (0.002) (0.003) (0.002) (0.002) (0.003)
Co up ion
con ol
0.027 0.021 0.038 0.052 0.060 0.033 0.059** 0.063*** 0.030 -0.079***
-0.643***
-0.016
(0.019) (0.035) (0.039) (0.040) (0.039) (0.043) (0.025) (0.023) (0.036) (0.028) (0.045) (0.037)
Log.
Go e nmen
du abili y
0.005 0.010 0.026 0.043 0.031 0.008 0.040*** 0.063*** -0.009 -0.036** 0.017 -0.028
(0.015) (0.025) (0.030) (0.030) (0.027) (0.038) (0.014) (0.014) (0.027) (0.018) (0.014) (0.025)
Log. GDP pe
capi a
-0.125 0.432** -0.184 0.163 -0.276 0.468* -0.101 0.482* 0.430** 0.306 0.401*** 0.239
(0.098) (0.200) (0.235) (0.234) (0.190) (0.246) (0.168) (0.247) (0.180) (0.186) (0.103) (0.180)
Con igui y 0.864***
(0.124)
Con igui y 0.859*** 0.701*** 0.410*** 0.829***
0.301*** 0.473*** 0.475*** 0.352*** 0.212*** 0.468***
(0.101) (0.096) (0.091) (0.109) (0.057) (0.055) (0.085) (0.065) (0.052) (0.085)
Obse a ions 1412 1412 1412 1412 1222 1350 1330 1373 1412 1412 1412 1412
R-squa ed 0.785 0.869 0.887 0.894 0.941 0.877 0.964 0.970 0.918 0.953 0.970 0.911
IDOS Discussion Pape 3/2025
30
Appendix B: Sample and desc ip i e s a is ics
Table B1: Summa y s a is ics o he baseline model a iables
Va iables Obs. Mean Sd. Min. Max.
MPI 1,566 2.044 1.869 0 10
P i a e domes ic
in es men
1,555 13.003 6.083 0.036 33.490
Log. Employmen 1,566 4.035 0.199 3.427 4.453
T ade globalisa ion 1,566 49.849 15.471 15.934 85.464
Co up ion con ol 1,560 2.171 0.733 0.5 5
Log. Go e nmen
du abili y
1,524 2.429 1.125 0 4.595
Log. GDP pe capi a 1,548 9.058 1.035 6.588 11.453
IDOS Discussion Pape 3/2025
31
Table B2: Sample
Coun y A e age
MPI
Coun y A e age
MPI
Coun y A e age
MPI
Albania 1.28 Jo dan 2.83 Tu key 3.94
Alge ia 1.94 Kazakhs an 1.44 Uganda 2.33
Angola 1 Kenya 0.61 Uk aine 2.89
A gen ina 4.78 Kuwai 5.11 U uguay 1.89
A menia 3 Lebanon 3.83 Venezuela, RB 0
Aze baijan 2.39 Libe ia 1.67 Vie nam 1.83
Bahamas, The 2.56 Madagasca 0 Zambia 1
Bah ain 3 Malawi 1.78
Bangladesh 3.78 Malaysia 2
Bela us 1.5 Mali 0
Boli ia 2.61 Mexico 2.33
Bo swana 1.11 Moldo a 2.56
B azil 4.28 Mongolia 2.89
B unei Da ussalam 1.94 Mo occo 3
Bulga ia 2.39 Mozambique 2.94
Bu kina Faso 0 Myanma 0.11
Chile 6.5 Namibia 1.44
China 4.83 Nica agua 0.11
Colombia 6.61 Nige 0
Congo, Dem Rep 1.67 Nige ia 1.17
Cos a Rica 3 Oman 1.56
Co e d’I oi e 0.28 Pakis an 7.5
C oa ia 2 Panama 1.33
Dominican Republic 2.17 Pa aguay 3.89
Ecuado 5.11 Pe u 4.28
Egyp , A ab Rep 0.22 Philippines 2.61
El Sal ado 1 Poland 1.89
E hiopia 0.39 Romania 3.28
Gambia, The 1.78 Russian Fede a ion 1.28
Ghana 2 Saudi A abia 2.22
Gua emala 0.28 Senegal 0
Guinea-Bissau 0.06 Se bia 3.11
Guyana 0 Sie a Leone 0.67
Hai i 2.5 Sou h A ica 0.94
Hondu as 1.11 S i Lanka 1.22
Hunga y 1.78 Sudan 1.78
India 2.17 Tanzania 1
Indonesia 1.39 Thailand 1.78
I an, Islamic Rep 0.44 Togo 0
I aq 0.83 Tunisia 2.11
IDOS Discussion Pape 3/2025
32
Table B3: Sou ces o a iables
Va iables Na u e Sou ces
1. Main model a iables
Mac op uden ial Policy Index Sco es anging om 0 o 10 Ce u i e al. (2017)
P i a e domes ic in es men Con inuous In e na ional Mone a y Fund (IMF)’s In es men
and Capi al S ock da abase
Employmen a e Con inuous Wo ld Bank’s Wo ld De elopmen Indica o s
(WDI) da abase
T ade globalisa ion Index anging om 0 o 100 KOF index (D ehe , 2006a; Gygli e al., 2019)
Co up ion con ol Index anging om 0 o 6 In e na ional Coun y Risk Guide (ICRG)
Go e nmen du abili y Con inuous Poli y IV
2. Addi ional a iables
T ade openness Con inuous WDI
Te ms o ade Con inuous WDI
Capi al openness Index anging om -2 o 2 Chinn and I o (2008)
In la ion Con inuous WDI
Remi ances Con inuous WDI
Exchange a e egime Dummy Au ho s, om Ilze zki e al. (2019)
Real e ec i e exchange a e Con inuous Da as (2012)
Public in es men Con inuous IMF’s In es men and Capi al S ock da abase
Na u al esou ces Con inuous WDI
P ope y igh s Index anging om 0 o 1 V-DEM
Poli ical P essu es and
Con ols on Media Con en
Index anging om 0 o 40 Quali y o Go e nmen (Teo ell e al., 2016)
IMF p og ammes Dummy D ehe (2006b)
Economic eedom Index anging om 0 o 100 The He i age Founda ion
Regional GDP g ow h Con inuous Au ho s, om WDI
Bo owe -Ta ge ed
Ins umen s
Sco es anging om 0 o 2 Ce u i e al. (2017)
Financial Ins i u ion-Ta ge ed
Ins umen s
Sco es anging om 0 o 8 Ce u i e al. (2017)
Ou pu gap Dummy Au ho s, using eal GDP om WDI
Financial openness Index anging app oxima ely
om -2 o 2
Chinn and I o (2008)
Financial de elopmen Con inuous WDI
In o mal sec o Index anging om 0 o 100 Elgin e al. (2021)
Democ acy Index anging om 0 o 6 In e na ional Coun y Risk Guide (ICRG)
C edi g ow h Con inuous Au ho s, using da a om WDI
Financial inclusion Index anging om 0 o 100 Au ho s, using da a om he Financial Access
Su ey (IMF)
Banking c ises Dummy Lae en and Valencia (2020)
