Jalles, João To a ; Pessino, Ca ola; Calde on, Ana C is ina
Wo king Pape
Fiscal consolida ions in La in Ame ica and he Ca ibbean:
Do inequali y, co up ion, and in o mali y ma e ?
IDB Wo king Pape Se ies, No. IDB-WP-1668
P o ided in Coope a ion wi h:
In e -Ame ican De elopmen Bank (IDB), Washing on, DC
Sugges ed Ci a ion: Jalles, João To a ; Pessino, Ca ola; Calde on, Ana C is ina (2025) : Fiscal
consolida ions in La in Ame ica and he Ca ibbean: Do inequali y, co up ion, and in o mali y
ma e ?, IDB Wo king Pape Se ies, No. IDB-WP-1668, In e -Ame ican De elopmen Bank (IDB),
Washing on, DC,
h ps://doi.o g/10.18235/0013447
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/315949
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/
WORKING PAPER No IDB-WP-1668
Fiscal Consolida ions in La in
Ame ica and he Ca ibbean
Do Inequali y, Co up ion, and In o mali y
Ma e ?
João To a Jalles
Ca ola Pessino
Ana C is ina Calde on
In e -Ame ican De elopmen Bank
Ins i u ions o De elopmen Sec o
Fiscal Managemen Di ision
Ma ch 2025
Fiscal Consolida ions in La in Ame ica
and he Ca ibbean
Do Inequali y, Co up ion, and In o mali y
Ma e ?
João To a Jalles
Ca ola Pessino
Ana C is ina Calde on
In e -Ame ican De elopmen Bank
Ins i u ions o De elopmen Sec o
Fiscal Managemen Di ision
Ma ch 2025
Ca aloging-in-Publica ion da a p o ided by he
In e -Ame ican De elopmen Bank
Felipe He e a Lib a y
To a Jalles, João.
Fiscal consolida ions in La in Ame ica: do inequali y, in o mali y and co up ion ma e ? / João
To a Jalles, Ca ola Pessino, Ana C is ina Calde on.
p. cm. — (IDB Wo king Pape Se ies ; 1668)
Includes bibliog aphical e e ences.
1. Pho o ol aic powe gene a ion-La in Ame ica. 2. Pho o ol aic powe gene a ion-Ca ibbean
A ea. 3. Sola ene gy-La in Ame ica. 4. Sola ene gy-Ca ibbean A ea. I. Pessino, Ca ola.
II. Calde ón, Ana C is ina. III. In e -Ame ican De elopmen Bank. Fiscal Managemen Di ision.
IV. Ti le. V. Se ies.
IDB-WP-1668
JEL Codes: C23, E21, E62, H5, H62
Keywo ds: iscal adjus men s, il e ing, panel da a, bina y choice models, local p ojec ion, iscal
mul iplie s, nonlinea i ies, co up ion, shadow economy
h p://www.iadb.o g
Copy igh © 2025 In e -Ame ican De elopmen Bank ("IDB"). This wo k is subjec o a C ea i e Commons
license CC BY 3.0 IGO (h ps://c ea i ecommons.o g/licenses/by/3.0/igo/legalcode). The e ms and
condi ions indica ed in he URL link mus be me and he espec i e ecogni ion mus be g an ed o he
IDB.
Fu he o sec ion 8 o he abo e license, any media ion ela ing o dispu es a ising unde such license shall
be conduc ed in acco dance wi h he WIPO Media ion Rules. Any dispu e ela ed o he use o he wo ks o
he IDB ha canno be se led amicably shall be submi ed o a bi a ion pu suan o he Uni ed Na ions
Commission on In e na ional T ade Law (UNCITRAL) ules. The use o he IDB's name o any pu pose o he
han o a ibu ion and he use o he IDB's logo shall be subjec o a sepa a e w i en license ag eemen
be ween he IDB and he use and is no au ho ized as pa o his license.
No e ha he URL link includes e ms and condi ions ha a e an in eg al pa o his license.
The opinions exp essed in his wo k a e hose o he au ho s and do no necessa ily e lec he iews o he
In e -Ame ican De elopmen Bank, i s Boa d o Di ec o s, o he coun ies hey ep esen .
1
Abs ac *†‡*
Widening income dispa i ies, highe co up ion, and inc eased
in o mali y in many eme ging ma ke and de eloping economies
(EMDEs)—all wi h p essing and moun ing iscal p oblems—ha e
ekindled in e es in he empi ical analysis o he key ac o s
de e mining he occu ence o iscal consolida ions. Using disc e e
choice models, his pape examines he d i e s o iscal consolida ion
episodes in a sample o 148 EMDEs be ween 1980 and 2019, wi h a
ocus on La in Ame ican and Ca ibbean coun ies. Inequali y does no
seem o d i e consolida ions—which a e mo e likely du ing good
economic imes—while mo e in o mali y inc eases he p obabili y o
hei occu ence and co up ion dec eases i . In u n, when examining
he d i e s o success ul consolida ions, la ge income inequali y ac s
as a boos , while in o mali y is a hinde ance. In ac , while he size o
he public in es men mul iplie in La in Ame ica and he Ca ibbean is
la ge han in o he egions, when in o mali y is high, he mul iplie
e ec is educed o a much lowe and insigni ican magni ude. Resul s
a e obus o se e al sensi i i y and obus ness es s.
* João To a Jalles: Uni e si y o Lisbon–Lisbon School o Economics and Managemen (ISEG); Resea ch
in Economics and Ma hema ics and Resea ch Uni on Complexi y and Economics, Uni e sidade de
Lisboa–ISEG; Economics o Policy, Uni e sidade No a de Lisboa–No a School o Business and
Economics and IPAG Business School. Email: [email protected].
† Ca ola Pessino: In e -Ame ican De elopmen Bank. Email: cpessi[email p o ec ed].
‡ Ana C is ina Calde on: In e : In e -Ame ican De elopmen Bank. Email: [email p o ec ed].
* The au ho s hank And ea Luzu iaga o help ul commen s and sugges ions. The opinions exp essed
he ein a e hose o he au ho s and no necessa ily hose o hei employe s. Any emaining e o s a e
he au ho s’ sole esponsibili y.
2
1. In oduc ion
The COVID-19 pandemic has le a la ge den in he go e nmen budge s ac oss he
wo ld. Du ing 2020, go e nmen s had no choice bu o inc ease public spending o
igh he pandemic a a ime when sh inking economic ac i i y dep essed hei
alling e enues. Consequen ly, public deb ose by 5 pe cen age poin s (pp) on
a e age in 2020, o 46 pe cen o GDP. In 2021, highe - han-p ojec ed in la ion
educed public deb - o-GDP a ios in ad anced economies (AEs) and eme ging
ma ke s (EMs) (IMF, 2022). This ep ie e is likely o be sho li ed. As in la ion ises,
go e nmen bonds become less a ac i e o in es o s, and he cos s o bo owing
ises. The e a e a conside able numbe o low-income de eloping coun ies (LIDCs)
ha a e in deb dis ess and would need o emba k on iscal consolida ion in he
o eseeable u u e, pa icula ly i deb elie is no o hcoming (Clemen s e al., 2021).
In addi ion, in he con ex o EMs and LIDCs, he In e na ional Mone a y Fund (IMF)
ecen ly es ima ed ha sca ing om he COVID-19 pandemic will inc ease he
al eady sizable inancing needed o achie e he Sus ainable De elopmen Goals
(Benedek e al., 2021).
Gi en hese de elopmen s, he issue o iscal consolida ion will emain
pe inen in he o eseeable u u e o bo h ad anced economies and eme ging
ma ke s, while being o g ea e u gency in EMs and LIDCs because o deple ing (o a
lack o ) iscal space. Fiscal e enchmen o ge public inances back on he
sus ainabili y ack and add ess iscal sol ency conce ns is he e o e he pa h o
ollow. In e na ional ins i u ions such as he IMF p ojec ha a la ge sha e o
coun ies will pu sue iscal consolida ion in he coming yea s. Such p ojec ions a e
only ealis ic, howe e , o he ex en ha cu en economic and poli ical condi ions
a e su icien ly conduci e o iscal adjus men . The e is conside able li e a u e on
iscal consolida ions in AEs— e lec ing g ea e a ailabili y o da a— ocused on he
ype o iscal adjus men (spending a he han ax-based) ha is du able and mo e
g ow h iendly. By con as , s udies on EMs and LIDCs a e limi ed, in pa icula hose
on he La in Ame ican and Ca ibbean (LAC) egion. A he same ime, many
de eloping coun ies su e om highe - han-a e age le els o inequali y and
in o mali y and poo ins i u ional quali y (high co up ion), all o which can
comp omise iscal consolida ions' abili y o ake o and/o succeed.4 This is
4 Some o hese issues a e being in es iga ed in Eu opean coun ies ha unde wen consolida ion a e
he global inancial c isis o 2008 and ailed o achie e i because o hese issues (Pappa, Sajedi, and
Vella, 2015).
3
pa icula ly e iden in he LAC egion, whe e, despi e nea ly all coun ies
expe iencing a signi ican decline in income inequali y since he ea ly 2000s,
s uc u al challenges such as weak labo ma ke s, limi ed social mobili y, and
pe sis en dispa i ies in access o educa ion and heal h ca e con inue o hinde
equi able economic g ow h.5
In ac , mo e gene ally, he egion has aced signi ican iscal challenges o e
he yea s, wi h a ious coun ies in he egion g appling wi h high le els o public
deb , pe sis en iscal de ici s, and mac oeconomic ins abili y (Eyzagui e and San os,
2018). These challenges ha e posed obs acles o sus ainable economic g ow h, social
de elopmen , and inancial s abili y. Some key ac o s con ibu ing o hese iscal
challenges include mac oeconomic ola ili y,6 e enue dependence,7 in o mal
economy,8 public expendi u e p essu es,9 and poli ical and go e nance challenges.10
The pu pose o his empi ical pape is ou old. Fi s , i iden i ies and
cha ac e izes in a no el way iscal consolida ions in a la ge sample o eme ging
ma ke s and de eloping economies (EMDEs). Second, i empi ically s udies he main
de e minan s o iscal consolida ion. Thi d, i adds p e iously neglec ed dimensions
o inequali y, co up ion, and in o mali y. Finally, i looks a whe he d i e s a e
di e en depending on he composi ional na u e o he consolida ion p og am; hei
deg ee o success is also inspec ed.
The pape seeks answe s o he ollowing mo e de ailed se o ques ions. Wha
a e he key s ylized ac s cha ac e izing iscal consolida ions in LAC coun ies? Wha
a e he key mac oeconomic conside a ions ha induce hem o implemen iscal
consolida ions? Is i high deb , slowing g ow h, o wo sening e ms o ade? To wha
5 See, o example, López Cal a and Lus ig (2010), Gaspa ini, C uces, and To na olli (2011), Gaspa ini and
Lus ig (2011), Aze edo, Inchaus e, and San elice (2013), Székely and Mendoza (2015), Gaspa ini, C uces,
and To na olli (2016), and Székely and Mendoza (2016).
6 LAC economies ha e expe ienced high le els o mac oeconomic ola ili y, including luc ua ions in
commodi y p ices, exchange a es, and in e es a es. These ola ile condi ions can dis up iscal
planning and make i challenging o main ain iscal s abili y (Céspedes, Chang, and Velasco, 2017).
7 Many coun ies in he egion ely hea ily on a ew key sou ces o e enue, such as commodi y expo s
o speci ic indus ies. This dependence exposes go e nmen s o signi ican e enue luc ua ions,
making i di icul o sus ain s able iscal policies and mi iga e iscal isks (Tal i and Végh, 2005).
8 A sizable in o mal economy exis s in many LAC coun ies, leading o ax e asion, low e enue
collec ion, and limi ed iscal space. In o mali y poses challenges o e ec i e iscal managemen
because i unde mines ax compliance, hampe s e enue mobiliza ion e o s, and limi s he each o
social p o ec ion p og ams (Melguizo and Sca ascini, 2015).
9 The egion aces a ious p essu es on public expendi u e, including demands o social p og ams,
educa ion, heal h ca e, and in as uc u e de elopmen . Mee ing hese demands while main aining
iscal sus ainabili y equi es ca e ul expendi u e p io i iza ion, e iciency imp o emen s, and e ec i e
public adminis a ion (Ce ángolo and Lema, 2017).
10 Poli ical ins abili y, co up ion, and weak go e nance ha e a ec ed iscal managemen in se e al LAC
coun ies. Lack o poli ical consensus, sho - e m policy ocus d i en by elec o al cycles, and inadequa e
ins i u ions can hinde e ec i e iscal consolida ion e o s (Le y-Yeya i and Panizza, 2011).
4
ex en do inequali y, co up ion, and in o mali y ma e ? A e he e meaning ul
di e ences in e ms o he composi ion o iscal adjus men (i.e., in e ms o whe he
i is expendi u e based o ax based)? Wha makes a consolida ion succeed?
To answe hese ques ions, we ely on a new da ase cons uc ed using he
i s di e ence o new es ima es o he cyclically adjus ed p ima y balance (CAPB)
ob ained by means o he Hamil on (2018) il e . This is done o a la ge sample o 148
EMDEs be ween 1980 and 2019. We ind ha he du a ion o a iscal episode is highe
in AEs han in LAC and ha he ini ial iscal condi ions p e ailing jus be o e a gi en
episode seem o ha e had an impac on he size o subsequen iscal e o s. In
addi ion, he ypical iscal consolida ion episode is o sho du a ion (i.e., 2–3 yea s)
and in ol ed ela i ely modes gains. Then, elying on bina y choice models, we ind
ha a consolida ion is mo e likely o ake place in “good imes”: when g ow h is high,
coun ies expe ience posi i e e ms o ade shocks, and in la ion is low. High deb
emains a signi ican de e minan o consolida ion because LAC coun ies ha e
limi ed access o inancial ma ke s compa ed wi h AEs. Inequali y in he egion does
no seem o d i e consolida ions, while mo e in o mali y and less co up ion inc ease
he p obabili y o hei occu ence. Mo e-co up coun ies seem less inclined o ca y
ou expendi u e-based iscal adjus men s. The opposi e is ue o hose wi h la ge
in o mal sec o s. In ac , while he size o LAC’s public in es men mul iplie is la ge
han in o he coun y g oups, when in o mali y is high he mul iplie e ec ge s
educed o a much lowe and insigni ican magni ude. Resul s a e obus o se e al
sensi i i y and obus ness es s, including, in e alia, al e na i e consolida ion
measu es ha use o ecas e o s, na a i e app oaches, and he use o o he
es ima o s.
The emainde o he pape is o ganized as ollows. Sec ion 2 e iews he
ele an li e a u e. Sec ion 3 discusses he empi ical me hodology and Sec ion 4
p esen s he da a oge he wi h key s ylized ac s. Sec ion 5 discusses he empi ical
esul s and Sec ion 6 concludes and elabo a es on policy implica ions.
2. Li e a u e Re iew
This pape ocuses on he condi ions ha may lead coun ies o emba k on iscal
consolida ion, he success o iscal consolida ion, and o wha ex en inequali y,
co up ion, and in o mali y ha e an impac on such an adjus men p ocess. Hence,
he pape ela es o ou s ands o li e a u e: (i) he de e minan s o iscal
consolida ions, (ii) he ela ionship be ween inequali y and iscal consolida ions, (iii)
5
he ela ionship be ween co up ion and iscal consolida ions, and (i ) he
ela ionship be ween in o mali y and iscal consolida ions. Mos o he exis ing
li e a u e is su icien ly gene al and, consequen ly, does no speci ically ac o in o
add ess idiosync asies (o p oblems) in he LAC egion ha migh hinde he success
o a iscal consolida ion p og am.
Fiscal consolida ions—which in ol e educing budge de ici s and s abilizing
public deb —a e c ucial o sus ainable economic g ow h and inancial s abili y.
Unde s anding he de e minan s o iscal consolida ions can p o ide insigh s o
policymake s o design e ec i e consolida ion s a egies. The li e a u e examining
he de e minan s o iscal consolida ions ocuses on he ac o s ha in luence he
success o ailu e o such consolida ion e o s, including mac oeconomic ac o s,
poli ical ac o s, and s uc u al ac o s.
Fi s is he se o mac oeconomic ac o s. Fi s and o emos , economic g ow h
plays a signi ican ole in de e mining he success o iscal consolida ions (Alesina
and A dagna, 2010; Fa ás and Miho , 2013). Highe economic g ow h can lead o
inc eased ax e enues and educed spending on unemploymen bene i s, he eby
easing he iscal adjus men p ocess. S udies ha e shown ha iscal consolida ions
implemen ed du ing pe iods o economic expansion end o be mo e success ul
han hose implemen ed du ing ecessions. Howe e , he ela ionship be ween
g ow h and consolida ion is complex because consolida ion measu es hemsel es
can impac economic g ow h. The e o e, policymake s need o s ike a balance
be ween iscal discipline and suppo ing economic ac i i y. Second, in e es a es
ha e implica ions o he cos o public deb , which is a c ucial conside a ion in iscal
consolida ions (Gia azzi and Pagano, 1990; Pe o i, 1996). Highe in e es a es can
inc ease deb se icing cos s, making iscal adjus men s mo e challenging.
Con e sely, lowe in e es a es can c ea e mo e a o able condi ions o
consolida ions by educing he bu den o in e es paymen s. Howe e , he
ela ionship be ween in e es a es and iscal consolida ions is in luenced by ac o s
such as mone a y policy, in la ion expec a ions, and ma ke pe cep ions o a
coun y's c edi wo hiness. Thi d, in la ion can impac iscal consolida ions h ough
i s e ec s on public deb dynamics and economic g ow h (Alesina and Pe o i, 1996;
Alesina, A dagna, and T ebbi, 2006). Highe in la ion e odes he eal alue o public
deb (denomina ed in local cu ency), making i easie o achie e deb educ ion
12
1.5 imes he sample s anda d de ia ion (o equal o one sample s anda d de ia ion,
on a e age, o e wo yea s). As he e is no single, ag eed-upon de ini ion in he
li e a u e, and being awa e o bes p ac ices e iewed abo e, we adop a middle-
g ound app oach in de ining CAPB- o-GDP change h esholds o he de e mina ion
o iscal consolida ion episodes. We op o he Alesina and Pe o i (1997) app oach,
unde which a iscal consolida ion episode is de ined as a minimum annual
imp o emen in he CAPB- o-GDP a io o 0.5 pp o e wo consecu i e yea s.13
Ano he ele an issue is he CAPB measu e o choice. CAPB da a can be
ob ained ei he ia a publicly a ailable sou ce (e.g., he IMF's Wo ld Economic
Ou look [WEO] da abase) o compu ed using a il e ing app oach (by decomposing
GDP and go e nmen e enues in o hei cyclical and end componen s). In ela ion
o his, he e is no clea consensus in he li e a u e ega ding he “op imal” way o
es ima e po en ial ou pu . Acco ding o Bo io, Disya a , and Juselius (2017), pas
s udies ha e applied (i) uni a ia e s a is ical app oaches, usually consis ing o il e ing
ou he end componen om he cyclical one, o (ii) s uc u al app oaches, de i ing
he es ima es di ec ly om a heo e ical model. Awa e o he sho comings o using
ei he o he wo app oaches14 and he disad an age o no maximizing he o al
numbe o obse a ions in ou panel da abase when using he WEO CAPB,15 we apply
a il e ing echnique.
Once he po en ial ou pu (and, consequen ly, he ou pu gap) is ob ained, we
use i o compu e a new measu e o he CAPB. Re lec ing he ac ha he elas ici y
o go e nmen e enues (REV) o ou pu g ow h is close o one while p ima y
expendi u e (PEXP) is la gely inelas ic o g ow h (i.e., we assume he same as
Gi oua d and And é, 2005), we mul iply go e nmen e enues by he ac o
[1/(1+OG/100)] o calcula e REV!"# (adjus ed go e nmen e enues), wi h OG being he
ou pu gap ob ained ia he Hod ick-P esco (HP) o Hamil on il e s.16
Ma hema ically, we ha e:
13 The s a yea o a iscal consolida ion episode is, he e o e, he yea in which he e is a minimum
annual imp o emen in he CAPB- o-GDP a io o 0.5 pp, i he e is also a minimum annual
imp o emen in he CAPB- o-GDP a io o 0.5 pp in he ollowing yea . Acco dingly, he end yea o a
iscal consolida ion episode is he las yea (in a sequence o yea s) wi h a minimum annual
imp o emen in he CAPB- o-GDP a io o 0.5 pp, a e which he annual CAPB- o-GDP ei he imp o es
by less han 0.5 pp o wo sens (i.e., dec eases).
14 S a is ical me hods su e om he end-poin p oblem— ha is, hey a e ex emely sensi i e o he
addi ion o new da a and o eal- ime da a e isions. S uc u al models, on he o he hand, may be
di icul o implemen consis en ly in c oss-sec ional en i onmen s and ely on he imposi ion o p e-
de e mined assump ions.
15 The IMF does no ha e an o icial me hod o compu ing po en ial ou pu . While he mos common
IMF app oach elies on a p oduc ion unc ion, assump ions a y g ea ly ac oss coun ies.
16 Fo a discussion o hese app oaches, see, o example, Hamil on (2018).
13
CAPB =REV!"# −PEXP. (1)
Composi ion-wise, a iscal consolida ion episode is de ined as an expendi u e-
based episode i he a io o he cumula i e all in he p ima y expendi u e- o-GDP
a io (de ined as he sum o all annual changes in he p ima y expendi u e- o-GDP
a io wi hin he episode) o he cumula i e adjus men (de ined as he sum o all
annual changes in he CAPB- o-GDP a io) is la ge han (o equal o) 2/3 in absolu e
alue. I he sum o all annual changes in he p ima y expendi u e- o-GDP a io
wi hin a iscal consolida ion episode is posi i e, he episode is classi ied as a ax-
based consolida ion. All emaining cases a e classi ied as “mixed” consolida ion
episodes. Succinc ly, a iscal consolida ion episode is de ined as expendi u e based
when |∆&'(&)_+,&|
|∆)-&.)_+,&|≥ 2/3/and ∆PEXPC_GDP < 0, wi h CAPBC_GDP and PEXPC_GDP deno ing
cumula i e CAPB and p ima y expendi u e (in pe cen o GDP) wi hin a gi en
episode. Con e sely, a iscal consolida ion episode is de ined as ax based when
∆PEXPC_GDP/ ≥ 0. I ollows ha any episodes ha do no sa is y he c i e ia se o h
abo e a e classi ied as mixed iscal consolida ion episodes.
Weighing he a o emen ioned ac o s, ou p e e ed speci ica ion o he
de ini ion o iscal consolida ion episodes in his pape will employ a CAPB change
h eshold o 0.5 pp o e wo consecu i e yea s (Alesina and Pe o i, 1997), wi h he
CAPB da a ob ained based on he Hamil on il e (yielding ou “Hamil on-based”
c i e ion o iscal consolida ions). This speci ica ion comes wi h se e al ad an ages,
namely: (i) allowing us o maximize ou sample size by iden i ying mo e han 1,000
(450) iscal consolida ion yea s (episodes) ac oss 185 coun ies (37 AEs and 148
EMDEs) be ween 1980 and 2019 ( o ou pu poses only EMDEs will be used in he
empi ics), (ii) ensu ing b oad consis ency and compa abili y wi h he al eady-
es ablished li e a u e on iscal consolida ions (mos o which elies on CAPB me ics),
and (iii) p io i izing ela i ely du able iscal consolida ions as opposed o one-o
shocks o CAPB le els (gi en ha he CAPB h eshold c i e ion is applied o CAPB-
o-GDP changes o e wo yea s as opposed o a single yea ).
14
Table 1. Fiscal Consolida ion Episodes in LAC by C i e ia
Coun y
WEO FC episode
HP FC episode
Hamil on FC
episode
A gen ina
2002–2004, 2018–2019
2002–2004, 2017–2019
2002–2004, 2017–
2019
Belize
none
2004–2008, 2010–
2011, 2016–2018
2001–2002, 2004–
2006, 2016–2018
Boli ia
none
1994–1995, 2003–2006
1989–1990, 1994–
1995, 1999–2000,
2003–2006
B azil
1998–1999, 2016–2017
1998–1999
none
Chile
2004–2006, 2010–2011
2003–2006, 2010–2011
2003–2006, 2010–
2011
Colombia
2011–2012
1990–1991, 2000–2001,
2012–2012
1990–1991, 2000–
2001, 2011–2012
Cos a Rica
none
2005–2007
2005–2007
Dominican Republic
2004–2005
2004–2005, 2013–2015
2004–2005, 2013–
2015
Ecuado
none
1999–2000, 2010–2011,
2017–2018
1999–2000, 2010–
2011, 2017–2018
El Sal ado
2003–2004, 2016–2017
2003–2004, 2016–2017
1993–1994, 2003–
2004, 2016–2017
Guyana
2009–2010, 2012–2013
2007–2008
2008–2009
Hondu as
none
1991–1992, 1994–1995,
2004–2005, 2010–
2011, 2014–2015
1991–1992, 1994–
1995, 2004–2005,
2010–2011, 2014–
2015
Jamaica
none
1998–2000, 2012–2013
1998–2000, 2012–
2013
Mexico
1999–2001, 2016–2017
2000–2001, 2015–2017
Panama
1999–2000, 2005–2007
1995–1996, 1999–2000,
2005–2007, 2015–2016
1995–1996, 1999–
2000, 2005–2007
Pa aguay
none
1985–1986, 1989–1990,
1993–1994, 2003–2004,
2010–2011
1985–1986, 1989–
1990, 1997–1998,
2003–2004, 2010–
2011
Pe u
2006–2007, 2010–2011,
2018–2019
2006–2007, 2010–
2011, 2018–2019
1999–2001, 2016–
2017
Su iname
2006–2007, 2016–2018
1992–1996, 2003–
2007, 2016–2017
1994–1995, 2006–
2007, 2017–2018
T inidad and Tobago
none
1989–1991, 2018–2019
1989–1991, 2007–
2008, 2018–2019
U uguay
2001–2003
2001–2004
2002–2003
Venezuela
none
1989–1990, 1995–1996,
1999–2000, 2002–
2005
1989–1990, 1995–
1996, 1999–2000,
2002–2005, 2019
Sou ce:
Au ho s' elabo a ion.
No e:
"FC" = iscal consolida ion.
3.2. Empi ical App oach
Ou aim is o explo e whe he income inequali y, co up ion, and in o mali y (ICI o
sho ) a ec he likelihood o consolida ing public inances while con olling o o he
a iables iden i ied in he li e a u e as a ec ing he implemen a ion o iscal
consolida ions. Hence, ou main dependen a iable o in e es is he occu ence o
a iscal consolida ion episode. To cap u e his, we ely on a iscal consolida ion (FC)
15
dummy o coun y
i
in yea
ha akes he alue o 1 i coun y
i
is in a iscal
consolida ion episode (as de ined abo e) in yea
(0 o he wise). Based on his bina y
cha ac e iza ion, ou baseline empi ical exe cise consis s o es ima ing logis ic
eg essions o assess he likelihood o a gi en coun y expe iencing a iscal
consolida ion yea , wi h he coun e ac ual being he opposi e o his (i.e., no
expe iencing a iscal consolida ion yea ). We es ima e he ollowing model:
P ob(FC = 1|𝑋) = 𝛷(𝐼𝐶𝐼′𝜶 + 𝑋′𝜷), (2)
whe e α,/𝜷 a e ec o s o he pa ame e s o be es ima ed; 𝐼𝐶𝐼 is a p oxy o
inequali y, co up ion, o in o mali y; 𝑋 is a ec o o con ol a iables; and 𝛷(⋅) is he
logis ic unc ion. Ou lis o con ol a iables includes he eal GDP g ow h a e, he
a e o in la ion, and he deb - o-GDP a io. These a iables a e sou ced om he Ap il
2022 IMF WEO in age. We also add ade openness (p oxied by he alue o impo s
and expo s in pe cen o GDP), pe cen changes in he e ms o ade, and pe cen
changes in he eal e ec i e exchange a e om he Wo ld Bank’s Wo ld
De elopmen Indica o s (WDI) da abase as con ols. The model associa ed wi h
equa ion (2) can be w i en as:
𝐹𝐶/0 = 𝜶𝐼𝐶𝐼/012 + 𝜷𝑋/012 + 𝜀/0, (3)
whe e, again, he FC a iable akes he alue 1 i a iscal consolida ion episode akes
place (i.e., we allow o mul i-yea iscal consolida ion episodes): 𝐹𝐶i = 1 i a iscal
consolida ion akes place in coun y
i
du ing yea
(0 o he wise);
i
= 1, …, N;
is he
yea ; and 𝜀/0/is an independen and iden ically dis ibu ed (i.i.d.) e o e m. In his
case, each se o es ima es 𝜶/Cand 𝜷/
Dis in e p e ed as showing ( he ise o all in) he
likelihood o a iscal consolida ion yea being expe ienced by coun y
i
.
4. Da a and S ylized Fac s
Mac oeconomic da a come om he IMF’s Ap il 2022 WEO da abase. These include
eal GDP, he budge -balance- o-GDP a io, CAPB (pe cen o GDP), o al
go e nmen e enues (pe cen o GDP), p ima y go e nmen expendi u es (pe cen
o GDP), he CPI in la ion a e (pe cen ), and go e nmen g oss deb (pe cen o GDP).
Addi ional in o ma ion on ade openness ( alue o expo s and impo s, pe cen o
16
GDP), changes in he e ms o ade, and changes in he eal exchange a e come
om he Wo ld Bank's WDI da abase as men ioned abo e.
Inequali y p oxies a e gi en by he Gini index, which goes om 0 o 100, wi h
he la e deno ing mo e unequal income dis ibu ion. Se e al sou ces a e used. The
i s is Sol ’s (2009, 2020) S anda dized Wo ld Income Inequali y Da abase (SWIID),
which co e s 177 coun ies om 1960 o he p esen and includes bo h g oss (ma ke )
and ne (dispensable) Gini. The SWIID da ase combines income in o ma ion om
he Uni ed Na ions Wo ld Income Da abase (UNWIDER) and he Luxembou g
Income S udy (LIS). SWIID p o ides compa able s anda dized Gini coe icien s o
measu e income inequali y based on es ima es o ma ke (p e- axes and ans e s)
and ne (pos - axes and ans e s) income inequali y. This hus allows he compa ison
o income dispa i ies be o e and a e edis ibu ion by axa ion and ans e s o e
ime. No e ha axes de e mine households’ disposable income a ailable o
consump ion and hus in luence he income dis ibu ion. Howe e , disposable
income does no conside indi ec axes. This c ea es a limi a ion when only
disposable income is conside ed. As a esul , we look a bo h p e- ax-and- ans e s
and pos - ax-and- ans e s Gini indices.17 Acco ding o Po e ba (2007), using he
la e mi iga es he e e se causali y p oblem because pos - ax-and- ans e s a y
“mechanically” and “economically” wi h he iscal sys em whe eas he p e- ax-and-
ans e s measu e a ies solely h ough he endogenous esponses o labo supply o
he gene al equilib ium e ec on ac o p ices. We use bo h he ma ke and ne
income Gini indices, wi h high co e age ac oss coun ies and o e ime, in he
es ima ions.18
The second is Milano ic’s All Ginis da ase , which ep esen s a compila ion and
adap a ion o income o consump ion Gini coe icien s (calcula ed ac oss
households o household pe capi a, on g oss o ne basis) e ie ed om nine
sou ces and ends in 2017. Ou o hese nine sou ces he ollowing a e used due o
17 The Gini indica o s based on disposable income co e he o al ma ke income ecei ed by all
household membe s (g oss ea nings, sel -employmen income, and capi al income), plus he cu en
cash ans e s hey ecei e, less income and weal h axes, social secu i y con ibu ions, and cu en
ans e s ha hey pay o o he households.
18 The impu a ion me hodology o s anda dize obse a ions collec ed om a ious sou ces makes hese
se ies subjec o measu emen unce ain y (Jenkins, 2015). Indeed, he e a e some conce ns abou he
eliabili y o SWIID's impu ed es ima es pa icula ly in da a-poo egions (Jenkins, 2015). Tha said,
Fe ei a, Lus ig, and Teles (2015) compa ed eigh inequali y da ase s and conclude ha “al hough he e
is much ag eemen ac oss hese da abases, he e is also a non- i ial sha e o coun y/yea cells o
which subs an ial disc epancies exis ” and ha “ he me hodological di e ences […] o en appea o be
d i en by a undamen al ade-o be ween a wish o b oade co e age on he one hand, and o g ea e
compa abili y on he o he .”
17
sample limi a ions o he egion unde sc u iny (LAC): POVCAL, SEDLAC, Wo ld
Income Dis ibu ion (WID), and WIDER.
Fo co up ion wo sou ces a e used. The i s is he Wo ld Bank's Coun y
Policy and Ins i u ional Assessmen (CPIA) anspa ency, accoun abili y, and
co up ion a iable in he public sec o ( a ing 1–6, wi h 6 deno ing a be e sco e).
The second is he Co up ion Pe cep ions Index (CPI) om T anspa ency
In e na ional, which anks 180 coun ies a ound he wo ld by hei pe cei ed le els
o public sec o co up ion ( esul s a e gi en on a scale o 0 [highly co up ] o 100
[ e y clean]). Fo bo h indices hei espec i e mi o s a e compu ed so ha he
eading is as ollows: he la ge he alue, he mo e co up he coun y in a gi en
yea .
Fo in o mali y wo sou ces a e used. The i s is he Elgin e al. (2021) da ase
on in o mali y, which includes se e al p oxies: Dynamic S ochas ic Gene al
Equilib ium (DSGE) model–based es ima es o in o mal ou pu (pe cen GDP) and
mul iple indica o s mul iple causes (MIMIC) model–based es ima es o in o mal
ou pu (pe cen GDP), sel -employmen (pe cen o al employmen ), and in o mal
employmen (pe cen o al employmen ).19 The second sou ce o in o mali y is om
he In e na ional Labo O ganiza ion (ILO) and includes he o al in o mali y a e
(pe cen ).
To maximize co e age (pa icula ly wi hin LAC), he p e e ed sou ces o ou
h ee dimensions a e he SWIID o inequali y, he CPI o co up ion, and Elgin e al.
(2021) DSGE-based es ima es o in o mal ou pu (pe cen GDP) o in o mali y. Tha
said, sensi i i y is done using al e na i e p oxies o he di e en concep s.
Table 2 p o ides compa a i e summa y s a is ics be ween AEs and La in
Ame ica. We obse e ha he numbe o iscal episodes is signi ican ly lowe when
we conside he WEO-based consolida ion c i e ion compa ed wi h he HP-based o
Hamil on-based c i e ia. This is pa icula ly salien in he case o La in Ame ica, whe e
he e we e only 61 WEO-based consolida ions e sus 192 HP-based and 183 Hamil on-
based. Also, he a e age adjus men in pe cen o GDP is lowe when he o me
c i e ia is used (less han 2 pe cen o GDP when using WEO-based and mo e han 2
pe cen o GDP wi h HP- o Hamil on-based). So, using he WEO unde es ima es he
a e age size o consolida ions due o da a limi a ions. In addi ion, he du a ion o a
19 MIMIC is essen ially a s uc u al model whe e he shadow economy is es ima ed om a sys em o
equa ions composed o economic and ins i u ional a iables. The DSGE a iable comes om a
de e minis ic DGE model ini ially p oposed by Elgin and Oz unali (2012).
18
iscal episode is highe o AEs han he du a ion obse ed o La in Ame ica. In ac ,
while he epo ed du a ion is, on a e age, h ee yea s o AEs, he du a ion o iscal
episodes o La in Ame ica is sligh ly lowe a 2.5 yea s. The ull se o episodes by
La in Ame ican coun y (wi h co esponding yea s) is p o ided in Table A1 in he
Appendix. Fo ins ance, wo me hods—WEO and HP— ha de e mine iscal
consolida ion episodes on he basis o he change in he CAPB essen ially coincide
in iden i ying, o ins ance, he iscal con ac ions o A gen ina in 2002–2004 o B azil
in 1998–1999.
Table 2. Summa y S a is ics o Fiscal Consolida ions by CAPB Measu e and pe
C i e ion o Economic De elopmen
Ad anced Economies
To al # yea s o
FC episodes
A g. # FC
episodes
A g. size o adjus men in FC
episode (% GDP)
A g. du a ion o FC
episode (yea s)
WEO-
based
191
1.65
1.72
3.21
HP-based
267
2.25
1.94
3.14
Hamil on
-based
276
2.35
1.87
3.05
La in Ame ica
To al # yea s o
FC episodes
A g. # FC
episodes
A g. size o adjus men in FC
episode (% GDP)
A g. du a ion o FC
episode (yea s)
WEO-
based
61
1.59
1.99
2.34
HP-based
192
2.06
2.53
2.66
Hamil on
-based
183
2.33
2.17
2.64
Sou ce:
Au ho s' calcula ions.
No es:
"FC" = iscal consolida ion. A e age size o adjus men in FC episode is he cumula i e adjus men
(de ined as he sum o all annual changes in he CAPB- o-GDP a io wi hin he episode) wi hin a gi en
episode, di ided by he du a ion ( o al numbe o yea s) o he episode. A e age du a ion o FC episode
(yea s) is he sum o all yea s du ing which a coun y has consolida ed wi hin a gi en episode.
Figu e 1 epo s he dis ibu ion o changes in CAPB (pe cen o GDP) in LAC. In he
es o he ex we will e ain om p esen ing WEO-based e idence and ocus
ins ead on HP- and Hamil on-based consolida ions.
19
Figu e 1. Changes in CAPB (pe cen o GDP) du ing Fiscal Consolida ions in La in
Ame ica and he Ca ibbean
HP-based consolida ions
Hamil on-based consolida ions
Sou ce:
Au ho s' calcula ions.
No e:
The igu e plo s bo h he his og am and Ke nel densi ies o each c i e ion (HP and Hamil on) o
LAC.
As a as he cha ac e is ics o iscal consolida ion episodes, ini ial iscal condi ions
p e ailing jus be o e a gi en consolida ion episode seem o ha e had an impac on
he size o subsequen iscal e o s (Figu e 2). The lowe he CAPB, he la ge he
ensuing iscal consolida ion. This may e lec ha la ge budge de ici s made i mo e
necessa y o consolida e and, a he same ime, aised public awa eness o he ex en
o he iscal imbalance p oblem, making i easie o ac .
Figu e 2. Ini ial Fiscal Imbalance and Subsequen Adjus men s o Fiscal
Consolida ions in La in Ame ica and he Ca ibbean
HP-based consolida ions
Hamil on-based consolida ions
Sou ce:
Au ho s' calcula ions.
0.1 .2 .3
Densi y
-10 -5 0 5 10
dcapbhp
0.1 .2 .3
Densi y
-10 -5 0 5 10
dcapbham
-15 -10 -5 0 5
Budge posi ion in he yea be o e consolida ion
0 5 10 15
Imp o emen in budge posi ion
95% CI Fi ed alues
CAPB_hp_po py
wi h linea i -LAC
Ini ial Fiscal posi ions and subsequen adjus men
-15 -10 -5 0 5
Budge posi ion in he yea be o e consolida ion
0 5 10 15
Imp o emen in budge posi ion
95% CI Fi ed alues
CAPB_ham_po py
wi h linea i -LAC
Ini ial Fiscal posi ions and subsequen adjus men
20
Mos o he iscal consolida ion episodes in LAC we e o sho du a ion (see he op
wo panels in Figu e 3, bu wi h some excep ions—see Table A1 in he Appendix) and
in ol ed ela i ely modes gains (see bo om panels o Figu e 3). Howe e , he e we e
a numbe o la ge e o s, amoun ing o imp o emen s o mo e han 10 pe cen o
GDP. I is also possible o obse e ha , in gene al, sizable iscal consolida ions las ed
o longe pe iods and smalle consolida ions had a sho e du a ion (Figu e 4).
Figu e 3. Du a ion and Size o Consolida ion Episodes in La in Ame ica and he
Ca ibbean
HP-based consolida ions
Hamil on-based consolida ions
Sou ce:
Au ho s' calcula ions.
No es:
Top panels: budge posi ion measu ed by he CAPB (pe cen o po en ial GDP). Bo om panels:
imp o emen measu ed du ing he consolida ion yea s o he iden i ied episode.
050 100
Numbe o episodes
2 3 4 5
LAC
The dis ibu ion o episodes by du a ion, in yea s
050 100 150
Numbe o episodes
2 3 4 6
LAC
The dis ibu ion o episodes by du a ion, in yea s
010 20 30
Numbe o episodes
010 20 30 40
Imp o emen in budge posi ion du ing he episode
LAC
The Dis ibu ion o episodes by he size o consolida ion
010 20 30
Numbe o episodes
010 20 30
Imp o emen in budge posi ion du ing he episode
LAC
The Dis ibu ion o episodes by he size o consolida ion
21
Figu e 4. Rela ionship be ween Du a ion and Size o Consolida ion Episodes in
La in Ame ica and he Ca ibbean
HP-based consolida ions
Hamil on-based consolida ions
Sou ce: Au ho s' calcula ions.
No es:
Budge posi ion measu ed by he CAPB (pe cen o po en ial GDP). Imp o emen measu ed
du ing he consolida ion yea s o he iden i ied episode.
5. Empi ical Resul s
This sec ion p esen s he main indings on he ole o inequali y, co up ion, and
in o mali y as de e minan s o iscal consolida ions, based on he speci ica ions
discussed in Sec ion 3. Based on he discussion in Sec ions 3 and 4 and o he
eg ession speci ica ions men ioned below, we ely on an unbalanced panel
da abase a an annual equency. Summa y s a is ics o he a iables included in he
eg essions a e shown in Table A1 in he Appendix.
5.1. De e minan s o Fiscal Consolida ions
We begin wi h he es ima ion o logis ic eg ession (3) o explo e he main
de e minan s o iscal consolida ions in EMDEs and La in Ame ica (bo h excluding
he Ca ibbean coun ies) in Tables 3 and 4 espec i ely.20 Bo h show he esul s when
we: (i) ini ially exclude he ICI a iables (column 1) and (ii) indi idually add each ICI
a iable one a a ime.
Fo he EMDE sample, we obse e ha he wo se he le el o in o mali y, he
mo e likely i is o a coun y o consolida e, bu he esul is su ounded by g ea
unce ain y (compa e columns 8–10 in Table 3 wi h column 11). Co up ion comes ou
20 Resul s a e based on he Hamil on-based CAPB c i e ion o iden i ying iscal consolida ion episodes.
The WEO-based esul s a e a ailable upon eques . The HP-based esul s a e pa ially shown in he
Appendix and discussed in Sec ion 5.3 on obus ness and sensi i i y.
010 20 30 40
Imp o emen in budge posi ion du ing he episode
2 3 4 5
Du a ion o he episode (yea s)
LAC
Du a ion and size o consolida ion
010 20 30
Imp o emen in budge posi ion du ing he episode
2 3 4 5 6
Du a ion o he episode (yea s)
LAC
Du a ion and size o consolida ion
28
Table 6a. De e minan s o Success ul Fiscal Consolida ions, EMDEs
Speci ica ion
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
Success de ini ion
V1
V1
V1
V1
V2
V2
V2
V2
V3
V3
V3
V3
Real GDP g ow h ( -1)
0.180***
0.248***
0.161***
0.726***
0.120***
0.201**
0.148***
0.136
0.110***
0.193**
0.105***
0.099
(0.042)
(0.089)
(0.040)
(0.252)
(0.043)
(0.082)
(0.047)
(0.255)
(0.035)
(0.083)
(0.034)
(0.086)
Deb a io ( -1)
0.000
0.014
0.006+
0.005
0.007**
0.019*
0.010**
0.122*
0.006
0.014
0.009+
-0.006
(0.005)
(0.010)
(0.004)
(0.018)
(0.004)
(0.011)
(0.004)
(0.066)
(0.005)
(0.012)
(0.006)
(0.016)
In la ion ( -1)
0.066**
-0.014
0.042*
0.031
0.024
-0.187*
0.007
-0.252
0.059+
0.033
0.048+
0.001
(0.030)
(0.040)
(0.022)
(0.069)
(0.026)
(0.096)
(0.025)
(0.203)
(0.042)
(0.046)
(0.034)
(0.076)
T ade openness ( -1)
-0.001
-0.001
0.004
0.003
-0.003
-0.007
-0.001
0.030
-0.003
0.008
0.001
0.006
(0.004)
(0.004)
(0.004)
(0.010)
(0.003)
(0.006)
(0.004)
(0.022)
(0.004)
(0.006)
(0.005)
(0.012)
Te ms o ade g ow h ( -1)
-0.006
-0.082+
-0.024
-0.159
-0.005
0.009
-0.023
-0.185+
-0.005
0.042
0.011
-0.086+
(0.025)
(0.053)
(0.022)
(0.148)
(0.024)
(0.054)
(0.024)
(0.124)
(0.022)
(0.053)
(0.017)
(0.059)
SWIID Ma ke Gini ( -1)
0.030*
0.012
0.011
(0.015)
(0.024)
(0.019)
Co up ion Pe cep ions Index
(CPI) (in e ed) ( -1)
-0.010
0.033+
-0.006
(0.018)
(0.022)
(0.020)
MIMIC-based in o mali y ( -1)
-0.006
-0.000
0.013
(0.011)
(0.011)
(0.015)
In o mal employmen ( -1)
-0.068**
0.021
-0.028*
(0.031)
(0.033)
(0.016)
Obse a ions
224
92
254
34
224
92
254
34
224
92
254
34
McFadden Pseudo-R2
0.096
0.069
0.074
0.295
0.052
0.095
0.060
0.340
0.067
0.088
0.061
0.079
No es:
The dependen a iable is a dummy aking he alue o 1 in a success ul consolida ion yea (0 o he wise), de ined using he Hamil on-based c i e ion and
ollowing one o he de ini ions o “success” desc ibed in he main ex . S anda d e o s in pa en heses. Cons an e m omi ed. +, *, **, *** deno e s a is ical
signi icance a he 15, 10, 5, and 1 pe cen le els, espec i ely.
29
Table 6b. De e minan s o Success ul Fiscal Consolida ions, La in Ame ica
Speci ica ion
(1)
(2)
(3)
(4)
Success de ini ion
V1
V2
V3
V3
Real GDP g ow h ( -1)
0.476+
0.253
0.170**
0.170
(0.329)
(0.264)
(0.069)
(0.256)
Deb a io ( -1)
0.012
0.022**
0.051***
-0.023
(0.013)
(0.011)
(0.017)
(0.052)
In la ion ( -1)
0.269+
0.161
0.363**
0.864+
(0.183)
(0.242)
(0.158)
(0.566)
T ade openness ( -1)
0.038
0.020
0.025
-0.084
(0.028)
(0.027)
(0.038)
(0.075)
Te ms o ade g ow h ( -1)
-0.047+
-0.069
-0.059
-0.188
(0.033)
(0.085)
(0.082)
(0.253)
SEDLAC Gini ( -1)
0.144**
0.056
-0.317
(0.062)
(0.142)
(0.561)
MIMIC-based in o mali y ( -1)
0.073
(0.110)
Obse a ions
32
32
32
18
McFadden Pseudo-R2
0.306
0.218
0.404
0.574
No es:
The dependen a iable is a dummy aking he alue o 1 in a success ul consolida ion yea (0
o he wise), de ined using he Hamil on-based c i e ion and ollowing one o he de ini ions o “success”
desc ibed in he main ex . S anda d e o s in pa en heses. Cons an e m omi ed. +, *, **, *** deno e
s a is ical signi icance a he 15, 10, 5, and 1 pe cen le els, espec i ely.
Box 1. Non-Linea Fiscal Mul iplie E ec s Depending on Inequali y, Co up ion,
and In o mali y
He e we es ima e he condi ional esponse o eal GDP ollowing a go e nmen
spending shock, condi ioned on he le el o inequali y, co up ion, and in o mali y.
We use Jo dá’s (2005) non-linea local p ojec ions o ob ain impulse esponses ha
a e allowed o a y acco ding o a con inuous unc ion 𝐹(𝑧/0),/as ollows:
𝑦/,045 − 𝑦/,012 = 𝛼/+ 𝜏/+ [𝛽5
6× τ × 𝑆/,0] + [𝛽5
7× (1 − τ) × 𝑆/,0] + θ𝑋/,0 + 𝜀/,0, (B1)
whe e
y
is he eal GDP (in logs) and 𝑆/,0 is he go e nmen spending shock;
τ/is/an/indica o / unc ion/ ha / akes/ he/ alue = 1/𝑖𝑓/𝑧80
c
c
c
<𝑚𝑒𝑑9:;<=> such ha 𝑧8
h is a
coun y-speci ic a e age o an indica o o inequali y, co up ion, o in o mali y and
𝑚𝑒𝑑9:;<=>/is he co esponding sample median alue.a We a e in e es ed pa icula ly
in he 𝛽5
7 coe icien — ha is, when he s a e is o high inequali y, high co up ion, o
high in o mali y. Equa ion B1 is es ima ed using O dina y Leas Squa es (OLS) wi h
D iscoll-K aay (1998) obus s anda d e o s (clus e ed a he coun y le el).
To un equa ion B1, he big issue ela es o iden i ying 𝑆/,0. The obse ed
he e ogenei y in he es ima es o iscal spending mul iplie s e lec s o a g ea ex en
he gene al challenges associa ed wi h he iden i ica ion o exogenous shocks in
public spending. To da e, se e al app oaches ha e been employed o add ess his
issue.b We will ely on he app oach ha uses go e nmen spending o ecas e o s
employed by Aue bach and Go odnichenko (2012, 2013), Abiad, Fu ce i, and Topalo a
(2016), Fu ce i and Li (2017), and Colombo e al. (2022). These au ho s a gue ha his
me hodology o e comes he obs acles ha o en con ound he causal es ima ion o
he e ec o iscal policy on economic pe o mance.c The measu e o go e nmen
spending shocks is he di e ence be ween he ac ual eal alue o public in es men
o consump ion and he co esponding alue expec ed by analys s as o Oc obe o
he same yea . Mo e o mally, and simila ly o Colombo e al. (2022), he iden i ica ion
30
o unexpec ed iscal policy shocks using o ecas e o s in go e nmen spending is
done in wo s eps. Fi s , we compu e he -pe iod o ecas e o o public spending
o coun y
i
, 𝐹𝐸/,0|012 = ∆𝑙𝑛𝐺/,0 − ∆𝑙𝑛𝐺/,0|012, whe e ∆𝑙𝑛𝐺/,0 de ines he ac ual
go e nmen spending g ow h a e and ∆𝑙𝑛𝐺/,0|012 is he -1 IMF o ecas o ∆𝑙𝑛𝐺/,0 made
a ime -1 o yea
.d This app oach also sol es, by cons uc ion, he p oblem o “ iscal
o esigh ,” which a ises when agen s eac o an icipa ed a he han ealized shocks
(see, e.g., Fo ni and Gambe i, 2010; Ben Zee and Pappa, 2015). Second, he
co esponding public spending componen is eg essed on lags o he
co esponding o ecas e o (plus coun y and ime e ec s) o pu ge p edic able
componen s and by aking he esidual o his p ojec ion (no malized by he a e age
sha e in pe cen o GDP o he public spending a iable) as he iscal shock. Rela i e
o Colombo e al. (2022), we ha e expanded he ange o o ecas da a (and, hence,
he spending shocks) om 1995–2015 as in hei pape o he 1980–2019 pe iod.e
Resul s gi en by he impulse esponse a e shown in Figu e B1, which plo s he
uncondi ional esponses (dashed blue line) oge he wi h he condi ional ones (solid
black line) wi h associa ed con idence bands. We obse e a posi i e and insigni ican
uncondi ional ou pu esponse o go e nmen consump ion shocks, bu a posi i e
and signi ican uncondi ional esponse o go e nmen in es men shocks. The size
o he public in es men mul iplie in La in Ame ica is la ge han in o he g oups o
coun ies, such as OECD economies. When in o mali y is high he public in es men
mul iplie e ec ge s educed o a much lowe and insigni ican magni ude. The
e ec on ou pu om a public consump ion shock when in o mali y is high emains
s a is ically insigni ican . These esul s a e b oadly in line wi h hose ob ained by
Colombo e al. (2022) using a sho e ime span bu o a la ge sample o coun ies.
The unde es ima ion o iscal mul iplie s seems mo e p onounced in coun ies wi h
a highe le el o in o mali y, which o Pappa, Sajedi, and Vella (2015) can be
measu ed h ough ax e asion. The same applies o high inequali y. Fo co up ion,
he ou pu impac o a public consump ion shock is no s a is ically di e en om
ze o and is no dis inguishable om he baseline. In he case o a public in es men
shock he e ec emains posi i e and signi ican bu , again, s a is ically no di e en
om he uncondi ional esul .
Figu e B1. Condi ional Response o Real GDP G ow h o Go e nmen Spending
Shock: The Role o Inequali y, Co up ion, and In o mali y (in pe cen )
-2 0 2 4 6
0 1 2 3 4
yea
public cons. shock in high in o mali y
-1 0 1 2
0 1 2 3 4
yea
public in . shock in high in o mali y
31
No es
: The
x
axis is in yea s; =0 is he yea o he iscal spending shock (i.e., an unan icipa ed 10 pe cen
inc ease); =1 is he i s yea o impac . Solid black lines deno e he esponse o a iscal spending shock
when inequali y, co up ion, and in o mali y a e high; he da k g ay a ea deno es 90 pe cen con idence
bands and he ligh g ay a ea deno es 68 pe cen con idence bands, based on s anda d e o s clus e ed
a coun y le el. The dashed blue line deno es he uncondi ional esul oge he wi h 90 pe cen
con idence bands depic ed as dashed ed lines.
a Due o a lack o su icien con inuous obse a ions o inequali y, co up ion, and in o mali y p oxies,
he app oach discussed by Aue bach and Go odnichenko (2012, 2013) o es ima e he local p ojec ion
in he con ex o a smoo h ansi ion au o eg essi e (STAR) model de eloped by G ange and Te äs i a
(1993) is no possible.
b These include iden i ica ion based on he assump ion ha go e nmen spending does no espond
o mac oeconomic shocks in he same pe iod in a S uc u al Vec o Au o eg essi e (SVAR) amewo k
(Blancha d and Pe o i, 2002; Ilze zki, Mendoza, and Végh, 2013); he na u al expe imen app oach
exploi ing a ia ion in he mili a y spending buildups (Ramey and Shapi o, 1998; Ramey, 2011a, 2011b;
Ramey and Zubai y, 2018) and o icial lending (K aay, 2012, 2014) as sou ces o exogenous luc ua ions
in go e nmen spending.
c Acco ding o Ramey (2016), shocks should be exogenous wi h espec o o he cu en and lagged
endogenous a iables, hey should be unco ela ed wi h o he exogenous shocks, and hey should
ep esen unan icipa ed mo emen s in exogenous a iables. The o ecas -e o app oach is a guably
able o add ess all h ee.
d Acco ding o An e al. (2018), IMF iscal o ecas s a e accu a e and p e e ed compa ed o p i a e-sec o
o ecas s.
e Lack o a la ge comp ehensi e and c oss-coun y compa able da ase on ax e enue o ecas s p e en
us om explo ing he o he side o he budge , namely ax hikes o shocks— ha is, e enue-based
consolida ions.
Basile, Gi a di, and Miele (2016) exploi I alian da a on ax e asion and un epo ed income o in es iga e
he esponse o he o mal and in o mal sec o s o public expendi u e shocks. They ind ha in I aly
iscal expansions cause a educ ion in he sha e o un epo ed income.
5.3. Robus ness and Sensi i i y
Se e al sensi i i y and obus ness exe cises we e conduc ed, beginning wi h he
sensi i i y exe cises o add ess omi ed a iable bias. Resul s a e shown in Table A3
in he Appendix. Tes ing o he “o iginal sin” (i.e., coun ies’ obliga ion o bo ow in
o eign cu ency while being able o pay in domes ic cu ency), we ob ain (based on
-2 0 2 4 6
0 1 2 3 4
yea
public cons. shock in high inequali y
-1 0 1 2
0 1 2 3 4
yea
public in . shock in high inequali y
-5 0 5 10
0 1 2 3 4
yea
public cons. shock in high co up ion
0.5 11.5 2
0 1 2 3 4
yea
public in . shock in high co up ion
32
he speci ica ion unde column 2 o Table 3) a s a is ically insigni ican coe icien o
(lagged) in e es paymen s. We also es ed o he “u gency o consolida e” (i.e., he
dis ance be ween a coun y’s deb -s abilizing p ima y balance and CAPB in pe cen
o GDP), ob aining a posi i e and s a is ically signi ican coe icien o he EMDE
sample bu no o La in Ame ica. Finally, we explo ed he ole o (lagged)
in e na ional ese es minus gold e ie ed om he Wo ld Bank WDI da abase and
ob ained a s a is ically insigni ican coe icien . We ound ha in La in Ame ica, he
la ge he s ock o ese es, he less likely i is o a coun y o consolida e.
Tables A4 and A5 in he Appendix epea Tables 3 and 4 o he EMDE and LAC
samples, espec i ely, by eplacing he bina y dependen a iable wi h a con inuous
a iable gi en by he yea ly change in he Hamil on-based CAPB (pe cen GDP). This
will cap u e all s uc u ally adjus ed imp o emen s in he o e all balance, om he
smalles o he la ges —in con as wi h he dummy a iable logis ic app oach. OLS
eg essions sugges ha esul s a e b oadly simila wi h a key di e ence: o bo h he
EMDE and LAC samples, he la ge he le el o in o mali y, he la ge he ensuing
imp o emen in he CAPB.
In e ms o obus ness, we s a by using an al e na i e CAPB-based de ini ion
o iden i y he consolida ion yea s—namely, he HP. Resul s p esen ed in Table A6 in
he Appendix, while sligh ly weake in s a is ical e ms, a e gene ally consis en wi h
hose in Tables 3 and 4 o he EMDE and LAC samples, espec i ely.
Nex , we cons uc al e na i e measu es o consolida ion. The i s builds on
Gali and Pe o i (2003), Golinelli and Momigliano (2009), Alesina and A dagna (2010),
Co se i, Meie , and Mülle (2012), and Aue bach and Go odnichenko (2012). We
es ima e a iscal policy ule o he o m:
∆𝐶𝐴𝑃𝐵/0 = α/+ 𝜇0+ β∆𝐶𝐴𝑃𝐵/012 + γGAP/0 + δDEBT/012 + 𝜀/0, (4)
whe e α/ s ands o unobse ed coun y e ec s, 𝜇0 cap u es ime e ec s,
∆𝐶𝐴𝑃𝐵/0s ands o he yea ly change in he CAPB as a pe cen o GDP using he
Hamil on-based app oach, GAP/0 is he ou pu gap ob ained using he Hamil on-
based app oach, and DEBT/0 s ands o he deb - o-GDP a io. To accoun o he
con empo aneous co ela ion be ween he ou pu gap and he dependen a iable,
we es ima ed equa ion (4) by means o an ins umen al a iable echnique whe e
he ou pu gap is ins umen ed by each own lag and he i s lag o eal GDP g ow h
a e. Gi en ha o La in Ame ica N=13<T=40, he es ima es a e less suscep ible o he
33
so-called Nickell bias (Nickell, 1981). Acco ding o Gali and Pe o i (2003), he esponse
o he dependen a iable o ou pu gap e lec s he sys ema ic disc e iona y iscal
policy componen , while 𝜀/0 is he andom componen . This e lec s he non-
sys ema ic iscal policy esponse o he unan icipa ed iscal policy shocks, which a e
independen ac oss coun ies. The iscal consolida ion shock ha is used as an
al e na i e o he bina y dummy a iable used as baseline is de ined as ollows: 𝐷/0 =
1 i 𝜀/0 > 0 and 𝐷/0 = 0 i 𝜀/0 ≤ 0 (i.e., i has posi i e alue du ing imes o iscal
consolida ion, which implies ha he CAPB inc eases). Table 7.a shows he esul s.
The block o ixed con ols emains quali a i ely simila , and as be o e we s ill ge
ha , o La in Ame ica, inequali y does no seem o d i e consolida ions, while mo e
in o mali y inc eases he p obabili y o hei occu ence.
34
Table 7a. Panel Analysis: Fiscal Consolida ion Based on Two-S ep Reg ession, EMDEs e sus La in Ame ica
Speci ica ion
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Sample
EMDEs
La in Ame ica
Real GDP g ow h ( -1)
0.141***
0.111**
0.147***
0.129*
0.222***
0.122
0.342***
0.384**
(0.025)
(0.054)
(0.026)
(0.073)
(0.074)
(0.238)
(0.110)
(0.164)
Deb a io ( -1)
0.010***
0.011+
0.012***
0.033***
0.022+
0.075**
0.059***
0.085***
(0.002)
(0.007)
(0.003)
(0.013)
(0.016)
(0.034)
(0.022)
(0.026)
In la ion ( -1)
0.005
0.035
0.010
-0.104**
0.030
-0.196+
0.066
-0.070+
(0.009)
(0.025)
(0.009)
(0.044)
(0.025)
(0.124)
(0.061)
(0.048)
T ade openness ( -1)
0.001
0.011***
0.002
-0.003
-0.007**
0.032
-0.002
-0.016+
(0.002)
(0.004)
(0.002)
(0.007)
(0.003)
(0.025)
(0.013)
(0.010)
Te ms o ade g ow h ( -1)
0.002
0.012
0.004
0.025
0.007
0.069*
-0.005
0.016
(0.006)
(0.014)
(0.006)
(0.031)
(0.014)
(0.041)
(0.026)
(0.025)
SWIID Ma ke Gini ( -1)
0.022**
(0.009)
Co up ion Pe cep ions Index
(CPI) (in e ed) ( -1)
0.010
-0.005
(0.015)
(0.029)
DSGE-based in o mali y ( -1)
0.005
(0.007)
SEDLAC Gini ( -1)
-0.043
(0.035)
MIMIC-based in o mali y ( -1)
0.009
(0.008)
In o mal employmen ( -1)
-0.003
0.061***
(0.010)
(0.019)
Obse a ions
1,275
514
1,451
214
130
73
115
89
McFadden Pseudo-R2
0.055
0.059
0.057
0.088
0.126
0.278
0.217
0.272
No es:
The dependen a iable is a dummy aking he alue o 1 in a iscal consolida ion yea (0 o he wise). S anda d e o s in pa en heses. Cons an e m
omi ed. +, *, **, *** deno e s a is ical signi icance a he 15, 10, 5, and 1 pe cen le els, espec i ely.
35
The second consolida ion measu e elies on he al e na i e iden i ica ion app oach
ha uses o ecas e o s. Aue bach and Go odnichenko (2013) use o ecas e o s o
examine how he iscal mul iplie a ies wi h he business cycle in OECD economies.
This measu e o go e nmen shocks is compu ed as he di e ence be ween he
ac ual public spending and he public spending expec ed p e iously by p o essional
o ecas e s. Using o ecas e o –based shocks, Abiad, Fu ce i, and Topalo a (2016)
iden i y he causal impac o highe public in es men on ou pu , p i a e in es men ,
unemploymen , and public deb a ios. Abiad, Fu ce i, and Topalo a (2016) a gue
ha his me hodology o e comes he obs acles ha o en con ound he causal
es ima ion o he e ec o iscal policy on economic pe o mance. The Aue bach and
Go odnichenko (AG) app oach was also u ilized by Fu ce i and Li (2017), Honda,
Miyamo o, and Taniguchi (2020), and Miyamo o e al. (2020). This me hodology has
he ad an age o o e coming he p oblem o “ iscal o esigh ” (see Fo ni and
Gambe i, 2010; Leepe , Rich e , and Walke , 2012; Leepe , Walke , and Yang, 2013;
and Ben Zee and Pappa, 2017).
Mo e o mally, we cons uc a measu e o con ac iona y iscal policy shocks
as an unexpec ed imp o emen in he budge balance as a sha e o GDP. These
“ iscal policy” o ecas e o s a e he di e ences be ween he ac ual budge balances
epo ed in he ollowing yea and he one-yea -ahead o ecas s in he au umn
edi ion o he IMF’s WEO epo . Ma hema ically, we ha e 𝐹𝐸/,0|012 =𝐵𝐵/,0 −𝐵𝐵?/,0|012,
whe e 𝐵𝐵/,0 de ines he ac ual budge balance in pe cen o GDP and 𝐵𝐵?/,0|012 is he
IMF o ecas o 𝐵𝐵/,0 made a ime -1. We use budge balance o ecas s om 2003
un il 2019. Posi i e o ecas e o s mean ha he ac ual budge balance was la ge
han he o ecas , sugges ing an unan icipa ed iscal e enchmen . These ake he
alue 1 and 0 o he wise. Table 7.b shows he esul s. Fo EMDEs we ge sligh ly
con lic ing esul s on he in luence o in o mali y on he p obabili y o occu ence o
a consolida ion, while o La in Ame ica he posi i e and signi ican esul keeps
su acing as be o e.
36
Table 7.b. Panel Analysis: Fiscal Consolida ion Based on Fo ecas -E o App oach, EMDEs e sus La in Ame ica
Speci ica ion
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Sample
EMDEs
La in Ame ica
Real GDP g ow h ( -1)
0.071***
-0.008
0.053**
0.078
0.128+
-0.188*
0.123
0.094**
(0.026)
(0.026)
(0.021)
(0.056)
(0.087)
(0.111)
(0.127)
(0.037)
Deb a io ( -1)
0.007
-0.007*
0.005
-0.008
0.024
-0.007
0.010
-0.012*
(0.005)
(0.004)
(0.004)
(0.007)
(0.025)
(0.012)
(0.023)
(0.007)
In la ion ( -1)
0.010
-0.013
0.016
0.051
0.154**
-0.070
0.242***
-0.004
(0.011)
(0.025)
(0.013)
(0.065)
(0.072)
(0.058)
(0.043)
(0.076)
T ade openness ( -1)
-0.003
0.002
0.000
0.011+
0.005
0.001
-0.023
0.006
(0.003)
(0.002)
(0.003)
(0.007)
(0.018)
(0.011)
(0.017)
(0.010)
Te ms o ade g ow h ( -1)
0.004
0.014
0.006
-0.044**
-0.075*
-0.005
-0.096***
-0.048+
(0.010)
(0.020)
(0.009)
(0.021)
(0.044)
(0.051)
(0.034)
(0.033)
SWIID Ma ke Gini ( -1)
0.016
(0.016)
Co up ion Pe cep ions Index
(CPI) (in e ed) ( -1)
-0.008
-0.004
(0.008)
(0.021)
DSGE-based in o mali y ( -1)
0.022**
(0.009)
SEDLAC Gini ( -1)
0.098
(0.092)
MIMIC-based in o mali y ( -1)
0.033***
(0.008)
In o mal employmen ( -1)
-0.013*
0.002
(0.007)
(0.010)
Obse a ions
1,275
514
1,451
214
130
73
115
89
McFadden Pseudo-R2
0.020
0.009
0.017
0.048
0.157
0.037
0.183
0.050
No es:
The dependen a iable is a dummy aking he alue o 1 in a iscal consolida ion yea (0 o he wise). S anda d e o s in pa en heses. Cons an e m
omi ed. +, *, **, *** deno e s a is ical signi icance a he 15, 10, 5, and 1 pe cen le els, espec i ely.
37
The hi d and inal consolida ion measu e elies on he na a i e app oach da ase
pu oge he by Da id and Leigh (2018). This da ase o iscal consolida ions o 14 LAC
economies du ing 1989–2016 includes he size o he disc e iona y changes in axes
and expendi u es join ly and sepa a ely. This allows us o pe o m a composi ion
examina ion o he d i e s simila ly o he CAPB baseline app oach we did ea lie .
Table 7.c shows he esul s; hey a e gene ally weake , bu his could be he esul o
bo h a di e en sample composi ion and a di e en ime span unde sc u iny.
44
Aze edo, J. P., G. Inchaus e, and V. San elice. 2013. Decomposing he Recen
Inequali y Decline in La in Ame ica. Wo ld Bank Policy Resea ch Wo king
Pape No. 6715. Washing on, DC: Wo ld Bank.
Basile, R., A. Gi a di, and V. Miele. 2016. Tax e asion, public expendi u e and he
business cycle in I aly.
Economic Modelling
52: 40–49.
Benedek, J., I. Kinga, I. Tö ök, A. Teme dek, and J. Holobâcă. 2021.
Indica o -based
assessmen o local and egional p og ess owa d he Sus ainable
De elopmen Goals (SDGs): An in eg a ed app oach om Romania
.
Sus ainable De elopmen
29(5): 860–875.
Ben Zee , N., and E. Pappa. 2015. Mul iplie s o Unexpec ed Inc eases in De ense
Spending: An Empi ical In es iga ion.
Jou nal o Economic Dynamics and
Con ol
57(C): 205–226.
———. 2017. Ch onicle o a Wa Fo e old: The Mac oeconomic E ec s o An icipa ed
De ense Spending Shocks.
Economic Jou nal
127(603): 1568–1597.
Be g, A., and J. D. Os y. 2011. Inequali y and Unsus ainable G ow h: Two Sides o he
Same Coin? IMF S a Discussion No e No. 11/08. Washing on, DC: In e na ional
Mone a y Fund.
Besley, T., and T. Pe sson. 2011.
Pilla s o P ospe i y: The Poli ical Economics o
De elopmen Clus e s.
P ince on, NJ: P ince on Uni e si y P ess.
Blancha d, O., and R. Pe o i. 2002. An Empi ical Cha ac e iza ion o he Dynamic
E ec s o Changes in Go e nmen Spending and Taxes on Ou pu .
Qua e ly
Jou nal o Economics
117(4): 1329–1368.
Bo io, C., P. Disya a , and M. Juselius. 2017. Re hinking po en ial ou pu : Embedding
in o ma ion abou he inancial cycle.
Ox o d Economic Pape s
69(3): 655–677.
Céspedes, L. F., R. Chang, and A. Velasco. 2017. Financial In e media ion, Exchange
Ra es, and Uncon en ional Policies in an Open Economy.
Jou nal o
In e na ional Economics
108: S76-S86.
Ce ángolo, O., and D. Lema. 2017. Inclusi e G ow h and Fiscal Policies: Challenges o
La in Ame ica. Uni ed Na ions Economic Commission o La in Ame ica and
he Ca ibbean (ECLAC) Wo king Pape No. 388. San iago de Chile: Economic
Commission o La in Ame ica and he Ca ibbean.
Clemen s, B., S. Gup a, J. Jalles, and V. Mylonas. 2021. Low-Income De eloping
Coun ies Will Su ely Need Mo e Deb Relie Down he Line. Cen e o Global
De elopmen blog pos . June 23. Washing on, DC: Cen e o Global
De elopmen .
45
Cingano, F. 2014. T ends in Income Inequali y and I s Impac on Economic G ow h.
OECD Social, Employmen and Mig a ion Wo king Pape No. 163. Pa is: OECD
Publishing.
Colombo, E., D. Fu ce i, P. Pizzu o, and P. Ti elli. 2022. Fiscal Mul iplie s and In o mali y
(No. 2022-2082). Washing on, DC: In e na ional Mone a y Fund.
Co se i, G., A. Meie , and G. J. Mülle . 2012. Wha De e mines Go e nmen Spending
Mul iplie s?
Economic Jou nal
122(561): 555–579.
Da id, A., and D. Leigh. 2018. A New Ac ion-Based Da ase o Fiscal Consolida ion in
La in Ame ica and he Ca ibbean. IMF Wo king Pape 18/94. Washing on, DC:
In e na ional Mone a y Fund.
Da id, A., T. Koma suzaki, and S. Penknagu a. 2020. The Mac oeconomic E ec s o
S uc u al Re o ms in La in Ame ica and he Ca ibbean. IMF Wo king Pape
20/195. Washing on, DC: In e na ional Mone a y Fund.
De ies, P., J. Guaja do, D. Leigh, and A. Pesca o i. 2011. A New Ac ion-Based Da ase
o Fiscal Consolida ion. IMF Wo king Pape 11/128. Washing on, DC:
In e na ional Mone a y Fund.
D azen, A., and M. Esla a. 2010. Elec o al Manipula ion Via Vo e -F iendly Spending:
Theo y and E idence.
Jou nal o De elopmen Economics
92(1): 39–52.
D ehe , A., and M. Gassebne . 2013. G easing he Wheels? The Impac o Regula ions
and Co up ion on Fi m En y.
Public Choice
155(3–4): 413–432.
D iscoll, J. C., and A. C. K aay. 1998. Consis en Co a iance Ma ix Es ima ion wi h
Spa ially Dependen Panel Da a.
Re iew o Economics and S a is ics
80(4):
549–560.
Elgin, C., M. A. Kose, F. Ohnso ge, and S. Yu. 2021. Unde s anding In o mali y. CEPR
Discussion Pape No. 16497. Washing on, DC: Cen e o Economic and Policy
Resea ch.
Elgin, C., and O. Oz unali. 2012. Shadow Economies a ound he Wo ld: Model Based
Es ima es. Bogazici Uni e si y Depa men o Economics Wo king Pape s
2012/05.
Eyzagui e, N., and P. San os. 2018. Tackling Fiscal Challenges in La in Ame ica:
P og ess and Oppo uni ies. IMF Regional Economic Ou look: Wes e n
Hemisphe e. Washing on, DC: In e na ional Mone a y Fund.
Fa ás, A., and I. Miho . 2013. Fiscal Policy as a S abiliza ion Tool.
Handbook o
Mac oeconomics
2: 2301–2358.
46
Fe ei a, F., N. Lus ig, and D. Teles. 2015. App aising C oss-Na ional Income Inequali y
Da abases: An In oduc ion.
Jou nal o Economic Inequali y
13: 497–526.
Fo ni, M., and L. Gambe i. 2010. Fiscal Fo esigh and he E ec s o Go e nmen
Spending. CEPR Discussion Pape 049. Washing on, DC: Cen e o Economic
and Policy Resea ch.
Fu ce i, D., and B. Li. 2017. The Mac oeconomic (and Dis ibu ional) E ec s o Public
In es men in De eloping Economies. IMF Wo king Pape 17/217. Washing on,
DC: In e na ional Mone a y Fund.
Gali, J., and R. Pe o i. 2003. Fiscal Policy and Mone a y In eg a ion in Eu ope.
Economic Policy
18(37): 533–572.
Gao, S., and J. Shen. 2007. Asymp o ic P ope ies o a Double Penalized Maximum
Likelihood Es ima o in Logis ic Reg ession.
S a is ics and P obabili y Le e s
77: 925–930.
Gaspa ini, L., G. C uces, and L. To na olli. 2011. Recen T ends in Income Inequali y in
La in Ame ica.
Economia
10(2): 147–201.
———. 2016. Ch onicle o a Decele a ion Fo e old: Income Inequali y in La in Ame ica
in he 2010s.
Re is a de Economía Mundial
43: 2546.
Gaspa ini, L., and N. Lus ig. 2011. The Rise and Fall o Income Inequali y in La in
Ame ica. Wo king Pape No. 1110. New O leans: Tulane Uni e si y.
Ga es, S. 2001. Empi ically Assessing he Causes o Ci il Wa . Wo king Pape
(unpublished).
Gia azzi, F., and M. Pagano. 1990. Can Se e e Fiscal Con ac ions Be Expansiona y?
Tales o Two Small Eu opean Coun ies.
NBER Mac oeconomics Annual
5: 75–
122.
———. 1996. Non-Keynesian E ec s o Fiscal Policy Changes: In e na ional E idence
and he Swedish Expe ience.
Swedish Economic Policy Re iew
3: 67–103.
Gi oua d, N., and C. And é. 2005. Measu ing Cyclically-Adjus ed Budge Balances o
he OECD Coun ies. OECD Economics Depa men Wo king Pape No. 434.
Pa is: OECD Publishing.
Golinelli, R., and S. Momigliano. 2009. The Cyclical Reac ion o Fiscal Policies in he
Eu o A ea: The Role o Modelling Choices and Da a Vin ages.
Fiscal S udies
30(1): 39–72.
G ange , C. W. J., and T. Te as i a. 1993.
Modeling Nonlinea Economic Rela ionships
.
New Yo k: Ox o d Uni e si y P ess.
47
Guaja do, J., D. Leigh, and A. Pesca o i. 2014. Expansiona y Aus e i y: New
In e na ional E idence.
Jou nal o he Eu opean Economic Associa ion
12(4):
949–968.
Guicha d, S., M. Kennedy, and E. Wu zel. 2007. Wha P omo es Fiscal Consolida ion:
OECD Coun y Expe iences. OECD Economics Depa men Wo king Pape No.
553. Pa is: OECD Publishing.
Hall, R. E., and C. I. Jones. 1999. Why Do Some Coun ies P oduce So Much Mo e
Ou pu pe Wo ke Than O he s?
Qua e ly Jou nal o Economics
114(1): 83–116.
Halle be g, M., and P. Ma ie . 2004. Execu i e Au ho i y, he Pe sonal Vo e, and
Budge Discipline in La in Ame ican and Ca ibbean Coun ies.
Ame ican
Jou nal o Poli ical Science
48(3): 571–587.
Hamil on, J. D. 2018. Why You Should Ne e Use he Hod ick-P esco Fil e .
Re iew
o Economics and S a is ics
100(5): 831–843.
Hilsche , J., A. Ra i , and R. Reis. 2022. In la ing Away he Public Deb ? An Empi ical
Assessmen .
Re iew o Financial S udies
35(3): 1553–1595.
Honda J., H. Miyamo o, and M. Taniguchi. 2020. Explo ing he Ou pu E ec o Fiscal
Policy Shocks in Low Income Coun ies. IMF Wo king Pape 2020/012.
Washing on, DC: In e na ional Mone a y Fund.
Ilze zki, E., E. G. Mendoza, and C. A. Végh. 2013. How Big (Small?) A e Fiscal Mul iplie s?
Jou nal o Mone a y Economics
60(2): 239–254.
IMF (In e na ional Mone a y Fund). 2017.
Fiscal Moni o : Tackling Inequali y.
Washing on, DC: In e na ional Mone a y Fund.
———. 2022. IMF Fiscal Moni o : “Fiscal Policy om Pandemic o Wa .” Washing on, DC:
In e na ional Mone a y Fund.
Jenkins, S. P. 2015. Wo ld Income Inequali y Da abases: An Assessmen o WIID and
SWIID.
Jou nal o Economic Inequali y
13(4): 629–671.
Jo dà, O. 2005. Es ima ion and In e ence o Impulse Responses by Local P ojec ions.
Ame ican Economic Re iew
95: 161–182.
Jo dà, Ò., and A. Taylo . 2016. The Time o Aus e i y: Es ima ing he A e age
T ea men E ec o Fiscal Policy.
Economic Jou nal
126(590): 219–255.
Kalemli-Ozcan, S., B. E. Sø ensen, and O. Yosha. 2012. Fiscal Policy in a Small Open
Economy wi h Oil Sec o and Non-Oil Sec o .
Jou nal o he Eu opean
Economic Associa ion
10(7): 1577–1618.
Kau mann, D., and P. C. Vicen e. 2011. Legal Co up ion.
Economics and Poli ics
23(2):
195–219.
48
Keen, M., and B. Lockwood. 2010. The Value-Added Tax: I s Causes and
Consequences.
Jou nal o De elopmen Economics
92(2): 138–151.
King, G., and L. Zeng. 1999a. Logis ic Reg ession in Ra e E en s Da a. Depa men o
Go e nmen . Camb idge, MA: Ha a d Uni e si y.
———. 1999b. Es ima ing Absolu e, Rela i e, and A ibu able Risks in Case-Con ol
S udies. Depa men o Go e nmen . Camb idge, MA: Ha a d Uni e si y.
———. 2001. Explaining Ra e E en s in In e na ional Rela ions.
In e na ional
O ganiza ion
55: 693–715.
Kopi s, G., and S. Symansky. 1998. Fiscal Policy Rules (No. 162). Washing on, DC:
In e na ional Mone a y Fund.
K aay, A. 2012. How La ge Is he Go e nmen Spending Mul iplie ? E idence om
Wo ld Bank Lending.
Qua e ly Jou nal o Economics
127(2) 829–887.
———. 2014. Go e nmen Spending Mul iplie s in De eloping Coun ies: E idence om
Lending by O icial C edi o s.
Ame ican Economic Jou nal: Mac oeconomics
6(4) 170–208.
Lambe ini, L., and J. Ta a es. 2005. Exchange Ra es and Fiscal Adjus men s: E idence
om he OECD and Implica ions o he EMU.
Con ibu ions in
Mac oeconomics
5(1): 1–28.
Leepe , E. M., A. W. Rich e , and T. B. Walke . 2012. Quan i a i e E ec s o Fiscal
Fo esigh .
Ame ican Economic Jou nal: Economic Policy
4(2): 115–144.
Leepe , E. M., T. B. Walke , and S. Yang. 2013. Fiscal Fo esigh and In o ma ion Flows.
Econome ica
81(3): 1115–1145.
Le y-Yeya i, E., and U. Panizza. 2011. The Elusi e Cos s o So e eign De aul s.
Jou nal
o De elopmen Economics
94(1): 95–105.
López Cal a, L., and N. Lus ig (eds). 2010.
Declining Inequali y in La in Ame ica: A
Decade o P og ess?
New Yo k/Washing on, DC: Uni ed Na ions De elopmen
P og amme/B ookings Ins i u ion P ess.
Maloney, W. F., and F. Valencia. 2017. Enginee s o G ow h? Economic De elopmen
and he Role o he Enginee ing Sec o . Wo ld Bank Policy Resea ch Wo king
Pape No. 8211. Washing on, DC: Wo ld Bank.
Ma i, A., and J. Tho n on. 2005. De elopmen s in Domes ic Fuel P ices, Taxes, and
Subsidies in Selec ed Coun ies (unpublished). Fiscal A ai s Depa men .
Washing on, DC: In e na ional Mone a y Fund.
Mau o, P. 1998. Co up ion and he Composi ion o Go e nmen Expendi u e.
Jou nal
o Public Economics
69(2): 263–279.
49
Melguizo, A., and C. Sca ascini. 2015. In o mali y, P oduc i i y, and S a e Suppo o
he Ma ke s. IDB Wo king Pape No. 573. Washing on, DC: In e -Ame ican
De elopmen Bank.
Miyamo o, H., e al. 2020. G ow h Impac o Public In es men and he Role o
In as uc u e Go e nance. In G. Schwa z e al. (eds),
Well Spen : How S ong
In as uc u e Go e nance Can End Was e in Public In es men
. Washing on,
DC: In e na ional Mone a y Fund.
Nickell, S. 1981. Biases in Dynamic Models wi h Fixed E ec s.
Econome ica
49(6): 1417–
1426.
Pappa, E., R. Sajedi, and E. Vella. 2015. Fiscal Consolida ion wi h Tax E asion and
Co up ion.
Jou nal o In e na ional Economics
96: S56-S75.
Pe o i, R. 1996. G ow h, Income Dis ibu ion, and Democ acy: Wha he Da a Say.
Jou nal o Economic G ow h
1(2): 149–187.
———. 1999. Fiscal Policy in Good Times and Bad.
Qua e ly Jou nal o Economics
114(4):
1399–1436.
Pe y, G., W. F. Maloney, and O. A ias. S. 2007.
In o mali y: Exi and Exclusion
.
Washing on, DC: Wo ld Bank.
Pe sson, T., and G. Tabellini. 1994. Is Inequali y Ha m ul o G ow h?
Ame ican
Economic Re iew
84(3): 600–621.
Po a ke, N. 2013. Poli ical Cycles and Economic Pe o mance in OECD Coun ies:
Empi ical E idence om 1951–2006.
Public Choice
155(3–4): 393–409.
Po e ba, J. M. 2007. Income Inequali y and Income Taxa ion.
Jou nal o Policy
Modeling
29(4): 623–633.
Ramey, V. A. 2011a. Can Go e nmen Pu chases S imula e he Economy?
Jou nal o
Economic Li e a u e
3(49): 673–685.
———. 2011b. Iden i ying Go e nmen Spending Shocks: I ’s All in he Timing.
Qua e ly
Jou nal o Economics
1(126): 51–102.
———. 2016. Mac oeconomic Shocks and Thei P opaga ion. NBER Wo king Pape
Se ies. Wo king Pape 21978. Camb idge, MA: Na ional Bu eau o Economic
Resea ch.
Ramey, V. A., and M. D. Shapi o. 1998. Cos ly Capi al Realloca ion and he E ec s o
Go e nmen Spending.
Ca negie-Roches e Con e ence Se ies on Public
Policy
48: 145–194.
50
Ramey, V. A., and S. Zubai y. 2018. Go e nmen Spending Mul iplie s in Good Times
and in Bad: E idence om US His o ical Da a.
Jou nal o Poli ical Economy
126(2): 850–901.
Rome , C. D., and D. H Rome . 2010. The Mac oeconomic E ec s o Tax Changes:
Es ima es Based on a New Measu e o Fiscal Shocks.
Ame ican Economic
Re iew
100(3): 763–801.
Schneide , F., and D. H. Ens e. 2000. Shadow Economies: Size, Causes, and
Consequences.
Jou nal o Economic Li e a u e
38(1): 77–114.
Slem od, J. 2019. Tax Compliance and En o cemen .
Jou nal o Economic Li e a u e
57(4): 904–54.
Sol , F. 2009. S anda dizing he Wo ld Income Inequali y Da abase.
Social Science
Qua e ly
90(2): 231–242.
———. 2020. Measu ing Income Inequali y ac oss Coun ies and o e Time: The
S anda dized Wo ld Income Inequali y Da abase.
Social Science Qua e ly
101(3): 1183–1199.
Székely, M., and P. Mendoza. 2015. Is he Decline in Inequali y in La in Ame ica He e
o S ay?
Jou nal o Human De elopmen and Capabili ies
16: 397–419.
———. 2016. Declining Inequali y in La in Ame ica: S uc u al Shi o Tempo a y
Phenomena?
Ox o d De elopmen S udies
44.
Tal i, E., and C. Végh. 2005. Tax Base Va iabili y and P ocyclical Fiscal Policy in
De eloping Coun ies.
Jou nal o De elopmen Economics
78(1): 156–190.
Tommasi, M., and A. Velasco. 1996. Choosing and Reneging: The S abili y o Exchange
Regimes.
NBER Mac oeconomics Annual
11: 235–283.
Wei, S. J. 2001. Co up ion and Economic De elopmen .
Paci ic Economic Re iew
6(3): 317–337.
51
Appendix
Table A1. Summa y S a is ics o Reg ession Va iables o he Whole EMDE Sample
Va iable
Obse a ions
Mean
S anda d
de ia ion
Minimum
Maximum
Real GDP
g ow h
2,222
4.11
4.30
-41.88
38.2
Deb a io
1,953
55.36
37.62
0.07
451.33
In la ion
2,213
6.94
12.86
-72.72
256.10
Change in Real
E ec i e
Exchange Ra e
(REER)
889
-0.40
7.90
-77.93
36.20
T ade
openness
2,000
79.46
41.19
6.07
347.99
Te ms o ade
g ow h
2,024
0.24
8.93
-97.51
150.28
Gini Disposable
1,564
41.20
7.21
23.4
66.4
Gini Ma ke
1,564
45.55
6.79
22.4
70
gini_SEDLAC
153
51.40
4.55
40.97
59.41
Gini POVCAL
703
40.44
9.11
24.03
65.77
~cpi_in
631
62.38
11.96
26
92
sims_in o mal
133
59.65
21.10
22.47
88.64
DSGE_in o mal
1,738
35.97
9.77
8.55
66.43
in o mal_emp
220
62.91
21.33
18.91
99.65
52
Table A2. Panel Analysis Including he Exchange Ra e: Hamil on-Based Fiscal
Consolida ions, EMDEs
Speci ica ion
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
eg esso s
Real GDP g ow h
( -1)
-0.045*
-0.029
-0.022
-0.052
-0.062+
-0.135**
-0.141**
-
0.072***
-
0.063***
(0.028)
(0.031)
(0.031)
(0.042)
(0.043)
(0.055)
(0.058)
(0.024)
(0.023)
Deb a io ( -1)
-0.006
-0.004
-0.006
-0.016*
-0.015+
0.014
-0.016
-0.002
-0.003
(0.006)
(0.006)
(0.007)
(0.008)
(0.009)
(0.018)
(0.016)
(0.006)
(0.006)
In la ion ( -1)
-0.021
-
0.045**
-0.036+
-0.033+
-0.022
-0.070
0.067*
-0.041*
-0.024
(0.021)
(0.023)
(0.023)
(0.021)
(0.031)
(0.077)
(0.037)
(0.024)
(0.024)
REER g ow h ( -1)
-
0.058***
-
0.061**
-
0.067***
-
0.104***
-
0.098***
-0.078*
-0.075*
-0.053**
-
0.054***
(0.022)
(0.025)
(0.024)
(0.025)
(0.032)
(0.040)
(0.047)
(0.022)
(0.021)
T ade openness
( -1)
0.008***
0.013***
0.012***
0.011***
0.012***
0.002
-0.010
0.011***
0.010***
(0.003)
(0.003)
(0.003)
(0.004)
(0.004)
(0.008)
(0.011)
(0.003)
(0.003)
Te ms o ade
g ow h ( -1)
0.009
0.001
0.001
0.012
0.008
0.141***
0.005
0.009
0.008
(0.010)
(0.010)
(0.010)
(0.012)
(0.011)
(0.052)
(0.010)
(0.010)
(0.009)
SWIID Ma ke
Gini ( -1)
-0.037*
(0.020)
SWIID Disposable
Gini ( -1)
-0.027
(0.023)
POVCAL Gini ( -1)
-
0.038**
(0.019)
Wo ld Bank Gini
( -1)
-0.014
(0.020)
Co up ion
Pe cep ion Index
(CPI) (in e ed) ( -
1)
-0.001
(0.027)
Coun y Policy
and Ins i u ional
Assessmen
(CPIA) (in e ed)
( -1)
-
0.726+
(0.486)
DSGE-based
in o mali y ( -1)
0.012
(0.015)
MIMIC-based
in o mali y ( -1)
0.011
(0.015)
Obse a ions
652
534
534
288
211
205
170
606
621
McFadden
Pseudo-R2
0.044
0.076
0.072
0.114
0.109
0.130
0.119
0.064
0.054
No es:
The dependen a iable is a dummy aking he alue o 1 in a iscal consolida ion yea (0
o he wise), de ined using he Hamil on-based c i e ion. S anda d e o s in pa en heses clus e ed a he
coun y le el. Cons an e m omi ed. +, *, **, *** deno e s a is ical signi icance a he 15, 10, 5, and 1 pe cen
le els, espec i ely.
53
Table A3. Panel Analysis Add essing Omi ed Va iable Bias: Hamil on-Based Fiscal
Consolida ions, EMDEs e sus La in Ame ica
Speci ica ion eg esso s
(1)
(2)
(3)
(4)
(5)
(6)
Sample
EMDEs
La in
Ame ica
EMDEs
La in
Ame ica
EMDEs
La in
Ame ica
Real GDP g ow h ( -1)
-0.047***
0.007
-0.020
0.049
-0.044**
0.024
(0.017)
(0.066)
(0.022)
(0.079)
(0.019)
(0.077)
Deb a io ( -1)
0.006**
0.014**
0.004+
0.014+
0.002
0.009
(0.003)
(0.005)
(0.003)
(0.010)
(0.003)
(0.007)
In la ion ( -1)
-0.007
0.028
0.004
0.077*
-0.011
0.022
(0.009)
(0.031)
(0.010)
(0.045)
(0.011)
(0.036)
T ade openness ( -1)
0.005*
0.006
0.004*
0.005
0.004+
-0.004
(0.002)
(0.006)
(0.002)
(0.007)
(0.002)
(0.008)
Te ms o ade g ow h ( -1)
0.015*
0.005
0.014*
0.000
0.024***
0.009
(0.008)
(0.009)
(0.008)
(0.007)
(0.008)
(0.010)
In e es paymen s ( -1)
-0.056
-0.023
(0.050)
(0.153)
U gency o consolida e ( -1)
0.044**
0.047
(0.018)
(0.046)
Log ese es minus gold ( -1)
-0.059
-0.329*
(0.045)
(0.174)
Obse a ions
1,595
271
1,554
265
1,461
274
McFadden Pseudo-R2
0.019
0.048
0.022
0.044
0.020
0.071
No es:
The dependen a iable is a dummy aking he alue o 1 in a iscal consolida ion yea (0
o he wise), de ined using he Hamil on-based c i e ion. S anda d e o s in pa en heses clus e ed a he
coun y le el. Cons an e m omi ed. +, *, **, *** deno e s a is ical signi icance a he 15, 10, 5, and 1 pe cen
le els, espec i ely.
