Moće ić, Amina; Lazo ic-Pi a, Lejla
A icle
De e minan s o local go e nmen deb in he Fede a ion
o Bosnia and He zego ina
Financial In e ne Qua e ly
P o ided in Coope a ion wi h:
Uni e si y o In o ma ion Technology and Managemen , Rzeszów
Sugges ed Ci a ion: Moće ić, Amina; Lazo ic-Pi a, Lejla (2024) : De e minan s o local go e nmen
deb in he Fede a ion o Bosnia and He zego ina, Financial In e ne Qua e ly, ISSN 2719-3454,
Sciendo, Wa saw, Vol. 20, Iss. 3, pp. 58-69,
h ps://doi.o g/10.2478/ iq -2024-0019
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/329880
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-nc-nd/3.0/
10.2478/ iq -2024-0019
Abs ac The heo y o iscal ede alism sugges s ha deb o subcen al le els o go e nmen is closely
ela ed o he in e nal o ganiza ion o he coun y, de ined h ough in e go e nmen al iscal
ela ions. Due o a g ea impo ance o subcen al le els o go e nmen in he heo y and p ac-
ice o iscal ede alism, he aim o he esea ch is o empi ically in es iga e he de e minan s o
local go e nmen uni s’ (LGUs) deb in he Fede a ion o Bosnia and He zego ina (FBiH). Based
on panel da a om 2011 o 2019 and by using a gene alized me hod o momen s (GMM), we
es ima e he model using se s o iscal, ins i u ional, economic, and demog aphic a iables. The
esul s indica e ha he deb o LGUs in FBiH is signi ican ly de e mined by in e go e nmen al
iscal ela ions de ined h ough a g oup o iscal and ins i u ional a iables and demog aphic ac-
o s. Signi ican de e minan s o LGU’s deb in FBiH in he bo h sho and a long e m include
inancing o special expendi u e needs o some LGUs oge he wi h LGUs wi hin a can on wi h
special expendi u e needs. Fu he mo e, ans e s om o he go e nmen le els a e also signi i-
can in he sho and a long e m. Based upon ou esul s, we conclude wi h se e al policy impli-
ca ions impo an o he upcoming in e go e nmen al decision-making p ocesses in FBiH.
JEL classi ica ion: H68, H74
Keywo ds: De e minan s o Deb , Fede a ion o Bosnia and He zego ina, Local Go e nmen Uni s,
Recei ed: 16.05.2024 Accep ed: 26.07.2024
Ci e his:
Moće ić, A. & Lazo ić-Pi a, L. (2024). De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina. Financial In e ne Qua -
e ly 20(3), pp. 58-69.
© 2024 Amina Moće ić & Lejla Lazo ić-Pi a, published by Sciendo. This wo k is licensed unde he C ea i e Commons A ibu ion-NonComme cial-
NoDe i a i es 3.0 License.
1 Minis y o Finance o Sa aje o Can on, Bosnia and He zego ina, email: amina.moce [email p o ec ed].go .ba, ORCID: h ps://o cid.o g/0009-0000-5839-
2199.
2 Uni e si y o Sa aje o, School o Economics and Business, email: lejla.lazo ic@e sa.unsa.ba, ORCID: h ps://o cid.o g/0000-0001-9421-1842.
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
1. Wha a e he de e minan s o LGU deb in FBiH in
he sho and he long un?
2. Do coun y-speci ic iscal and ins i u ional a iables
a ec LGUs deb in FBiH in he sho and he long
un?
By applying dynamic GMM modelling, we measu e
he de e minan s o LGU deb o he pe iod 2011-2019
by using o al local deb (in BAM) as a dependen a ia-
ble. Based upon p e ious academic li e a u e, inde-
penden a iables a e di ided in o se e al g oups: is-
cal, ins i u ional, economic, and socio-demog aphic
a iables. A e he in oduc ion, he a icle is di ided
in o ou addi ional chap e s – li e a u e e iew ol-
lowed by he me hodology sec ion and esul s and dis-
cussion sec ion. The conclusion sec ion ies o shed
some ligh on he impo ance o in e go e nmen al
iscal ela ions in all aspec s o public inance as well as
in public bo owing, especially o small, pos - ansi ion
coun ies wi h complex cons i u ional o ganiza ion.
The exis ing academic li e a u e ela ed o de e mi-
nan s o LGU o subcen al deb has been gaining
p ominence in he las hi y yea s. Howe e , due o
each coun y’s speci ics, in e ms o empi ical model-
ling, he e is no single econome ic model ha ep e-
sen s a scien i ic s anda d in de e mining he ac o s o
local deb a he subcen al le el wi hin coun ies. In
his pape , we p o ide an o e iew o he li e a u e
ha analysed he deb o LGUs. In he li e a u e e iew,
we will p esen he me hods, dependen and g oup o
independen a iables used in p e ious academic e-
sea ch.
Ea ly esea ch o he 42 la ges Ame ican ci ies
om 1980-1990 applied a pooled ime-se ies eg es-
sion model in he wo ks o C op and Wandel (1997).
The au ho s concluded ha ci y deb was in luenced by
in e go e nmen al g an s and ede al e o m o e
which mos ci ies can ha e li le o no con ol. The au-
ho s also ind ha local poli ical ac o s do no ha e
a de e mining in luence on he deb policy o he ob-
se ed Ame ican ci ies. The a iables used we e classi-
ied as poli ical a iables (deb limi s, he exis ence o
e e endums on bo owing, ax and expendi u e e-
s ic ions, e c.) and socioeconomic a iables (popu-
la ion densi y, ede al g an s, he ax e o m law o
1986, e c.).
Feld and Ki chgassne (1999) by using he SUR
(Seemingly Un ela ed Reg ession) model o he yea
1990 and by analysing 131 Swiss municipali ies con-
cluded ha he p ocedu e " op-down" does no a ec
he deb and ha o e s hemsel es ca e mo e abou
iscal discipline han hei elec ed ep esen a i es e en
i he e a e no limi s on bo owing. The a iables hey
Public deb is an impo an indica o o sus ainable
public inances and an ins umen o iscal policy in
each coun y. A e he global inancial c isis (GFC) in
2008-2009, local go e nmen uni s (LGUs) in Bosnia
and He zego ina (BiH) we e ine i ably aced wi h
a ‘scisso s e ec ’ o inc eased local public expendi u es
and, unde high dependence on sha ed e enues, all-
ing public e enues which lead o highe local public
bo owing. Basa ić e al., (2018) s a e ha om 2008
un il 2015 he o al agg ega e local public deb (loans
and issued secu i ies) o all LGUs in BiH mo e han
doubled.
Bosnia and He zego ina (BiH) consis s o wo sepa-
a e en i ies (FBiH and Republika S pska (RS)) whe e
FBiH is mo e iscally decen alised en i y ( ede al le el,
can ons, and municipali ies and ci ies (LGUs)) and
whe e LGUs in bo h en i ies ha e a low le el o iscal
au onomy. This unique and complex cons i u ional
s uc u e ep esen ed as an asymme ic con ede a ion
ine i ably a ec s in e go e nmen al iscal ela ions and
he e o e equi es u he analysis o he posi ion o
LGUs wi hin such a cons i u ional s uc u e.
Public deb is a complex phenomenon and empi i-
cal esea ch a emp s o cap u e i s complexi y. Aca-
demic in e es using di e en me hods and a ailable
li e a u e on de e minan s o subcen al public deb
ac oss (see: C op & Wandel, 1997, Feld e al., 2011,
Balague -Coll e al., 2015, Galiński, 2023 and o he s)
and be ween coun ies (Bello e al., 2017) has been
inc easing o e he las h ee decades.
Howe e , he e is no a ailable empi ical li e a u e
ha in es iga es de e minan s o subcen al o local
deb o LGUs in BiH (FBiH) by applying dynamic panel
modelling. Balague -Coll e al. (2015) in hei esea ch
on he de e minan s o public deb indica e ha he
same egula o y ules canno be applied o all coun-
ies, no can he same ules be applied o all LGUs
wi hin one coun y. The e o e, ou model will include
he speci ics o ules ha egula e in e go e nmen al
iscal ela ions in FBiH as a iables o in e es o his
esea ch and measu e hei impac on LGU local public
deb . Bea ing in mind he unique and complex cons i u-
ional o ganiza ion o BiH ha ine i ably a ec s sub-
cen al public e enues, expendi u es and bo owing
capaci y, we wish o explo e he de e minan s o LGU
deb in FBiH in he sho and in he long e m. The spe-
ci ic eason behind ou in es iga ion lies in he u u e
policy ecommenda ions o in e go e nmen al iscal
ela ions ha would conside he impac ha iscal,
legal and ins i u ional egula ion has on LGU deb in he
sho - e m and long- e m. Fo such pu poses, we de-
ine he ollowing esea ch ques ions:
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
he sha e o own income in he o al income o he
municipali y, he numbe o axpaye s in he communi-
y, he ax a e o he a e age axpaye , he ole o he
Minis e o Finance, he numbe o poli ical pa ies in
he execu i e b anch, a dummy a iable o he iscal
e e endum, e c.
Veiga and Veiga (2014) by using FE and Sys em
GMM o he pe iod 1979-2012 and he pe iod 2003-
2012 examined a sample o 308 Po uguese municipali-
ies and concluded ha highe a e age sala ies o he
p i a e sec o a he municipal le el led o lowe public
bo owing and ha he public deb had g own in he
elec ion yea and by he highe unemploymen a e.
The explana o y a iables hey used we e sha e o capi-
al expendi u es in o al expendi u es, sha e o ax e -
enues in o al e enues, pe cen age o popula ion o e
65-yea s, unemploymen a e, popula ion densi y,
mon hly sala y pe inhabi an , sha e o he popula ion
employed in he p i a e sec o , pe cen age o he mu-
nicipal a ea assigned o u ban use, elec ion yea , num-
be o ou is s o e s pe inhabi an , e c.
Bogumil e al. (2014) by using OLS examined he
budge deb o 1,111 ci ies and municipali ies wi h o e
10,000 inhabi an s in Ge many in 2010 and concluded
ha exogenous ac o s such as he sha e o he popula-
ion in basic social secu i y, he impo ance o he place
o wo k (Eas o Wes Ge many), and he s a us as
a dis ic - ee ci y we e signi ican explana o y ac o s
o he deb o ci ies and municipali ies and ha he
deb inc eased wi h socio-economic p oblems. The
explana o y a iables hey used a e unemploymen
a e, sales ax e enue, numbe o social wel a e ecipi-
en s, agmen a ion o local Pa liamen , membe ship o
he CDU mayo , numbe o CDU membe s in he ci y
council, place o wo k (Eas o Wes Ge many) and o h-
e s.
Galiński (2015) by using OLS o he pe iod 1995-
2013 in es iga ed he de e minan s o LGU bo owing
in Poland and concluded ha he deb g ew ega dless
o in es men ac i i ies, which lead o he conclusion
ha some dys unc ions wi hin he sys em o LGU e e-
nues and expendi u es lead o g ea e bo owing.
Balague -Coll e al. (2015) by using OLS quan ile
eg ession o he yea 2008 analysed 1381 municipali-
ies in Spain wi h o e 1000 inhabi an s concluded ha
he a iables such as non- inancial su plus/de ici , he
a io o ax e enues o o al e enues, capi al expendi-
u es, and he le el o economic ac i i y ha e had
a signi ican impac on public bo owing. Balague -Coll
e al. (2015) says: “(…) he impac o he con ollable
deb d i e s depends on he speci ic poin o depa -
u e, so he imposi ion o s ic and igid limi s on he
le el o deb may ha e an undesi able impac on he
in es men and de elopmen plans o weal hy munici-
used we e he unemploymen a e, he numbe o ax-
paye s in he municipali y, he a e age axable income,
he sha e o le -wing pa ies in he execu i e powe ,
he numbe o pa ies in he execu i e powe , a dum-
my a iable i he e was democ a ic decision-making on
he budge de ici , and indices ela ed o budge plan-
ning, e c.
Beni o and Bas ida (2004) by using OLS be ween-
g oups (BG) and wi hin-g oups (WG) o he pe iod
1994-1998 analysed a sample o 130 ci ies in Spain and
concluded ha non- inancial su plus/de ici , inancial
independence, capi al expendi u e, and capi al income
a e he a iables ha bes explain he indeb edness a
he municipal le el in Spain. The explana o y a iables
hey used we e non- inancial su plus/de ici , a io o
ax e enue o o al e enue, capi al expendi u e, capi-
al e enue, popula ion, economic le el (measu ed by
household disposable income), and poli ical ideology.
Ashwo h e al. (2005) by using a FMOLS (Fully
Modi ied O dina y Leas Squa es) eg ession model o
he pe iod 1977-2000 analysed a sample o 298 Flemish
municipali ies (Belgium) and concluded ha he e is
a s ong ela ionship be ween public bo owing and
income, popula ion, and deb se icing cos s. The ex-
plana o y a iables hey used we e income, popula ion,
income pe capi a, deb se icing cos , numbe o po-
li ical pa ies, and numbe o poli ical pa ies in coali-
ion.
B idges (2005) by using simple eg ession analysed
he pe capi a deb o he popula ion o 6 Ken ucky
ci ies in 2005 and concluded ha go e nmen s wi h
highe pe capi a ax e enues we e less deb - inanced
han go e nmen s wi h lowe ax e enues, and ha
densely popula ed a eas ha e had highe public deb
han spa sely popula ed a eas.
Cabases e al. (2007) by using panel da a s a i ied
by popula ion analysed a sample o 8102 municipali ies
in Spain in he pe iod 1988-2000 and concluded ha
in es men s de e mine he public bo owing o Spanish
municipali ies and ha municipali ies wi h g ea e is-
cal capaci y bo ow less. The explana o y a iables hey
used we e municipal GDP pe capi a, a io o eal in-
es men o GDP, a io o own ax e enue o GDP,
a io o deb se ice o cu en e enue in he p e ious
yea pe size o popula ion, ne sa ing in deb se icing
cos s, and dummy a iables.
Feld e al. (2011) by using a c oss-sec ional analysis
o he yea s 1990 and 2004 examined he local deb
pe axpaye o 137 la ge Swiss owns and illages and
concluded ha di ec democ a ic igh s and a high de-
g ee o iscal au onomy led o a lowe le el o local
deb and ha he sha e o own e enues in o al e e-
nues has a signi ican nega i e e ec on local deb . The
explana o y a iables hey used we e unemploymen ,
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
u es and bo owing. The explana o y a iables hey
used we e capi al expendi u e and in es men in in a-
s uc u e asse s, ne su plus/ne de ici , dummy a ia-
ble o he yea s 2007 and 2008 (GFC), o he asse s
(excep in as uc u e), o he income (excep income
ha is assigned o hem a ce ain a es).
Mos ecen ly, Galiński (2023) by using panel da a
models wi h ixed e ec s (FE) o he pe iod 2010-2021
o 222 municipali ies om Wielkopolski Region in Po-
land concluded ha he e is a di ec ela ionship be-
ween he sha e o PIT and CIT in own e enues and
indeb edness o he municipali ies. The dec ease o
unemploymen a e and he g ow h o business densi y
con ibu ed o he dec ease in he deb o e enue
a io. The inc ease o he indeb edness esul ed om
an inc ease in he in es men ac i i y, he sha e o
spendings on wages as well as expendi u es on educa-
ion.
As pe a ailable li e a u e indica ed in he p e ious
sec ion, Table 1 ep esen s a iables o in e es we
used in ou esea ch. O e he pas hi y yea s, da a
a ailabili y has always been he majo obs acle o any
analysis wi hin BiH/FBiH, and he e o e i emains so
o speci ic in es iga ion in o he opic o LGU deb and
i s de e minan s.
pali ies, while a i ing la e o iscally i esponsible mu-
nicipali ies (p. 31).”
The explana o y a iables hey used a e non-
inancial su plus /de ici , a io o ax e enues o o al
e enues, capi al expendi u es, popula ion densi y,
le el o ou ism de elopmen , le el o economic ac i i-
y (measu ed by he a io o co po a e income ax o
o al municipal e enues), ne sa ings ( unds o in es -
men ) and o he s.
Med e-Balin and Bohle (2016) by using RE
(Random E ec s Model) o he pe iod om 2007-2013
analysed he ela ionship be ween public bo owing
and EU g an s and p ojec s inanced by EU unds in
Poland and Hunga y and concluded ha access o EU
unds caused an inc ease in deb a he subcen al le el
due o he need o co- inance such p ojec s. The in-
c ease in he deb o he local le el in Hunga y has e-
sul ed in he aking away o ce ain esponsibili ies
om he local le el, hus educing hei ole o a mini-
mum.
Ehalaiye e al. (2017) by using pooled OLS eg es-
sion o he pe iod 2004-2015 analysed he deb o 78
local and egional go e nmen s in New Zealand and
concluded ha he own e enues o egions and p o -
inces a e he key de e minan o hei bo owing (no
in as uc u al spending) and ha he pos -GFC pe iod
is associa ed wi h an inc ease in in as uc u e expendi-
Table 1: Lis o a iables in he esea ch
Va iable Label Desc ip ion o a iable
Dependen a iable
The deb o LGUs DEB Deb o LGUs, which includes o eign bo owing, domes ic bo ow-
ing, secu i ies, and issued gua an ees (in BAM)
Fiscal/budge a iables
The budge de ici in
a p e ious yea DEF-lag
De ici shown in he inancial epo s o LGUs. I a su plus is shown,
he uni is p esen ed as ha ing no de ici o he gi en yea , i.e., wi h
a alue o ze o (in BAM)
Capi al expendi u es/
To al expendi u es CAPITAL Capi al ans e s and expendi u es o he acquisi ion o ixed asse s
(in BAM) / To al expendi u es (in BAM)
Sha e o ax e enues in
o al e enues TAX Tax e enues (in BAM)/ To al e enues (in BAM)
Sha e o cu en expend-
i u es in o al expendi-
u es
EXP Cu en expendi u es (in BAM) /To al expendi u es (in BAM)
Cu en and capi al
ans e s o all le els o
go e nmen
TRANS The sum o cu en and capi al ans e s ecei ed by all le els o go -
e nmen (Fede a ion, can ons and o he LGUs) (in BAM)
T ans e s acco ding o
he decisions o he Fed-
e al Go e nmen
TRANS
FBiH
Amoun s o alloca ed ans e s aken om FBiH go e nmen deci-
sions (in BAM)
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
model o e s he possibili y o gene a ing in e nal in-
s umen s, so he ea men o po en ial endogenei y is
comp ehensi e and he es ima ion mo e consis en
(Roodman, 2006 and 2007; Baum, 2006). Based on he
abo e, we use a dynamic gene alized me hod o mo-
men s (GMM) es ima ion es ablished by A elano and
Bo e (1995) as well as Blundell and Bond (1998). In he
si ua ion o imbalanced panel models, he gaps and he
lack o da a a e bes ackled in he SGMM (Roodman,
2007). The ini ial gene al speci ica ion o ou model is
as ollows:
(1)
The dependen a iable in model (1) is local public
deb p esen ed by DEB, while DEBi -1 p esen s he i s
lag o he dependen a iable, and ui he e o includ-
ing all unobse ed impac s on local public deb . The
index “i ” p esen s he LGUs in FBiH “i” in ime “ ”. Xi is
1 x k ec o o con ol a iables iden i ied as impo an
de e minan s o local public deb , “d” p esen s k x 1
ec o o pa ame e s o be es ima ed while “T” p e-
sen s he ec o o ime dummy a iables included in
he analysis. To examine he de e minan s o LGUs
deb in FBiH, he a icle includes all 79 local uni s (N
= 79) in FBiH o e 2011-2019 pe iod (T = 9). The de-
penden a iable is deb o LGUs, simila o p e ious
esea ch as indica ed in he wo k Ehalaiye e al. (2017).
We combined he da a om he ollowing ins i u-
ions: Fede al Minis y o Finance, Cen e o he Rep-
esen a ion o Ci il In e es s (www.cpi.ba), Financial
and In o ma ion Agency (FIA), Fede al O ice o S a is-
ics, Published decisions o he FBiH Go e nmen , Dun
& B ads ee and he au ho s’ calcula ions (dummy
a iables).
Acco ding o he li e a u e e iew, he e is no
unique me hodology used in he esea ch o deb a
he subcen al le el. In he li e a u e e iew sec ion,
we ha e de e mined se e al ypes o econome ic es i-
ma ions and me hods deployed o he pu poses o
es ima ing local deb a he subcen al le el (OLS, OLS
quan ile eg ession, FE, RE, SUR model and SGMM).
Among he s udies p e e ing he dynamic panel model
in he analysis o deb a he local le el wi hin he
coun y, we single ou Veiga and Veiga (2014). As ou
esea ch aim is o dis inguish be ween sho - e m and
long- e m de e mina ion o deb a he local le el in
FBiH, he dynamic panel model is a sui able me hod o
es ima ion. Speci ic cha ac e is ics o he sample
(co e ing 79 local uni s o 9 yea s), in he si ua ion
when T < N, is an impo an a gumen o choosing
a dynamic panel model (G eene, 2008). The dynamic
panel model is also a good me hod o es ima ion when
po en ial endogenei y is conside ed (G eene, 2008),
which is he case in ou model. The dynamic panel
Va iable Label Desc ip ion o a iable
Ins i u ional a iables
LGUs ha ha e a special
expendi u e needs due
o he inancing o ma e-
ial cos s o p ima y
schools and ha ha e
mo e han 60,000 inhab-
i an s
LAW
SPEC
dummy
The dummy a iable has a alue o 1 o LGUs ha ha e a special
expendi u e needs, 0 o LGUs ha do no ha e special expendi u e
needs.
LGUs in can ons ha
ha e a special
expendi u e need
CANTONSPEC
dummy
The dummy a iable has a alue o 1 o LGUs in can ons ha ha e
special expendi u e needs, 0 o LGUs in can ons ha do no ha e
special expendi u e needs.
Economic a iables
A e age annual ne sala-
y in LGUs ACTIVITY A e age annual ne sala y in LGUs in all sec o s acco ding o NACE
e . 2 classi ica ion (in BAM)
T ade balance o LGUs TRDB Disagg ega ed ade balance as a di e ence be ween expo s and
impo s a he le el o LGUs (in BAM)
G oss p o i o compa-
nies on an annual le el in
LGUs
B-PROFIT The sum o he g oss p o i achie ed by companies in LGUs o each
yea o obse a ion (in BAM)
Demog aphic a iables
Popula ion POP To al numbe o esiden s o LGUs
Popula ion densi y pe
km2 DENSITY LGUs popula ion/ LGUs a ea
Sou ce: Au ho ’s own wo k.
1
* * *
i i i i i
DEB DEB X T u
−
= + + + +
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
legally and ins i u ionally es ablished ules in FBiH,
which in ou model a e measu ed by he a iables o
in e es (LAW - SPECdummy, CANTON - SPECdummy
and TRANS) ha e sho and long- e m e ec s on deb
o LGUs and we wan o in es iga e his impac .
The esul s o ou speci ica ion model a e p esen -
ed in Table 2. We ea he lagged e ec o deb as p e-
de e mined, i.e., endogenous a iables ins umen ed
wi h lagged le els and di e ences. All o he independ-
en a iables a e ea ed as ins umen s. The sample
has 79 g oups, i.e., LGUs, and he model was es ima ed
by using 25 ins umen s. We es ima ed speci ica ion by
using he SGMM me hod and epo ed he ob ained
esul s in Table 2.
Speci ica ion o he inal dynamic panel model is in
he ollowing de eloped o m:
(2)
Fo an ini ial empi ical check, we es ima e he
Pea son coe icien o co ela ion and examine he co-
e icien s o co ela ion o all a iables in he model.
We ind ha he e is no p oblem o high co ela ion
be ween a iables (co ela ion coe icien s as a ule o
humb a e below he le el o 0.7, Appendix 1).
Speci ically wi h i s policy implica ions, we ex-
pec ha in e go e nmen al iscal ela ions and iscally,
0 1 1 2 1 3
4 5 6 7
8 9 10 1
11 12
13 14
i i i i
i i i i
i i i
i i
i
DEB DEB DEF CAPITAL
TAX EXP TRANS TRANS FBiH
TRDB ACTIVITY B PROFIT
LAW SPECdummy CANTON SPECdummy
DENSITY POP
−−
−
= + + +
+ + + + −
+ + + −
+ − + −
++
i i
+
Table 2: Base model - SGMM dynamic panel - one-s ep obus es ima e
Model diagnos ics
Numbe o obse a ions 598.000
Numbe o g oups (i.e. LGUs) 79.000
Numbe o ins umen s 25.000
F es o join signi icance
H0: Independen a iables a e join ly equal o ze o
Wald chi2(21) = 1,357.870
P ob > chi2 = 0.000
A ellano-Bond es o AR (1) in i s di e ences
H0: The e is no i s -o de se ial co ela ion in esiduals
z = -2.140
P > z = 0.032
A ellano-Bond es o AR (2) in second di e ences
H0: The e is no second-o de se ial co ela ion in esiduals
z = 0.350
P > z = 0.732
The dependen a iable is deb on subcen al le el – LGUs in FBiH in cu en BAM (DEB)
Va iables (Sho explana ion o a iable) Coe icien s z-s a is ic p- alue
Cons an (In e cep e m) -7,018,575.000 -2.08 0.037
DEB_lag (lagged dependen a iable, 1s lag) 0.595 5.55 0.000
DEF_lag 0.248 0.73 0.466
CAPITAL 179.780 1.07 0.287
TAX 12,043.550 0.94 0.346
EXP 44,405.060 1.99 0.047
TRANS 0.753 1.68 0.092
TRANSFBiH -0.340 -0.37 0.715
LAWSPECdummy 1,196,506.000 2.40 0.017
CANTONSPECdummy -672,626.000 -1.97 0.049
TRDB -0.000 -0.05 0.963
ACTIVITY 2,376.690 1.69 0.092
BPROFIT_lag -0.000 -0.45 0.649
DENSITY -623.030 -1.96 0.050
POP 36.840 2.50 0.012
Se o ime dummy a iables included
_Iyea _2012
_Iyea _2014
_Iyea _2015
_Iyea _2016
_Iyea _2017
_Iyea _2018
_Iyea _2019
-1,089,931.000
-86,299.830
187,430.300
-13,587.310
-13,350.530
-834,142.000
950,225.800
-2.30
-0.20
0.54
-0.06
-0.04
-1.78
2.18
0.022
0.844
0.588
0.954
0.971
0.075
0.029
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
pendi u e in o al expendi u e in LGUs; TRANS - ans-
e s o all le els o go e nmen o LGU; LAWSPECdum-
my - LGUs ha ha e a special expendi u e needs due
o: inancing ma e ial cos s in p ima y schools, +60,000
inhabi an s, o ha e been assigned a special expendi-
u e needs; CANTONSPECdummy - LGUs belonging o
can ons wi h special expendi u e needs; ACTIVITY -
a e age ne sala y in he LGUs; DENSITY - popula ion
densi y in LGUs, POP – popula ion in LGUs.
Ou esea ch esul s indica e he signi icance o he
coun y speci ic a iables, namely iscal and ins i u ion-
al a iables. Some o hem ep esen he speci ics o
iscal ela ions wi hin FBiH meaning hey could no be
he subjec o ea lie esea ch. Speci ically, he sho -
e m impac o he LAWSPECdummy is in e p e ed ha
LGUs ha ha e a special expendi u e need (due o he
inancing o ma e ial cos s in p ima y schools o be-
cause o he numbe o inhabi an s o e 60,000 o
hose LGUs ha ha e been assigned special expendi-
u e needs) end o ha e a highe le el o deb on a e -
age by 1,196,506 BAM compa ed o hose LGUs ha
ha e no been assigned special expendi u e need
(ce e is pa ibus, s a is ically signi ican a 5%). Fu he -
mo e, he sho - e m impac o CANTONSPECdummy
means ha LGUs ha belong o can ons ha ha e spe-
cial expendi u e needs end o ha e a lowe le el o
deb by 672,626 BAM on a e age compa ed o hose
LGUs ha do no belong o can ons ha a e assigned
a special expendi u e need (ce e is pa ibus, s a is ically
signi ican a 5%). Simila ly, he sho - e m impac o
he TRANS a iable indica es ha an inc ease in e-
cei ed ans e s om all le els o go e nmen by
1 BAM leads o an inc ease in LGU deb by 0.75 BAM
(ce e is pa ibus, s a is ically signi ican a 10%). The
esen ed esul s con ibu e owa ds he in es iga ion
o seconda y esea ch ques ion ega ding he impac
ha speci ic ins i u ional and legal a iables ha e on
LGU deb in FBiH.
Fu he mo e, he sho - e m impac o o he a ia-
bles on he deb o LGUs indica e ha an inc ease in
cu en expendi u es o o al expendi u es by 1 pe -
S a is ical diagnos ics sugges ha all ele an es s
o he SGMM ype o modelling a e app op ia e. Ac-
co ding o A elano and Bond (1991), he GMM e-
qui es ha he e is i s -o de se ial co ela ion bu
ha he e is no second-o de se ial co ela ion in he
esiduals. As we may see om Table 2, hose es s sup-
po he alidi y o he model speci ica ion. Baum
(2006) a gues ha he Hansen J- es is he mos used
diagnos ic in GMM es ima ion o assessmen o he
sui abili y o he model. The Hansen es o o e iden i-
ying es ic ions does no ejec he null hypo hesis a
any con en ional le el o signi icance (p = 0.254) and i
is an indica ion ha he model has alid ins umen a-
ion. The es ima ed coe icien o lagged dependen
a iable in he model has a alue less han uni y ac-
co ding o Roodman (2007). The es ima ed coe icien
o lagged dependen a iable in he model is 0.59
which means ha he s eady-s a e assump ion holds.
Bond (2002) sugges s addi ional de ec ion o he dy-
namic panel es ima es alidi y by checking i he es i-
ma ed coe icien o he lagged dependen a iable lies
be ween he alues ob ained om OLS and FE es ima-
o s, which is con i med in ou model (i.e., he ollow-
ing alues a e ob ained: OLS = 0.81 > GMM = 0.59 > FE
= 0.33). In ou esea ch, 25 ins umen s o igina ed om
he es ic ions o use one lag o le els and h ee o
di e ences in he da a (i.e., he es ic ion is se o (13)
in x bond2) using special use -w i en command col-
lapse o dec easing ins umen s. The numbe o ins u-
men s is lowe han he numbe o obse a ions and
numbe o g oups indica ing ha he model is app op i-
a e.
The model indica es ha he deb o LGUs in FBiH
is au o eg essi e in he obse ed pe iod indica ing ha
local deb om he p e ious pe iod is a signi ican de-
e minan o he cu en local deb . The s a is ical sig-
ni icance o he lag-dependen a iable (a he con en-
ional 1% le el) sugges s ha he cu en le el o local
deb p esen s he e lec ion o he his o ical e ec o
he deb a LGUs o FBiH. O he independen a iables
ha we e s a is ically signi ican a e: EXP - cu en ex-
Model diagnos ics
Hansen J- es o o e iden i ying es ic ions
H0: Model speci ica ion is co ec and all o e iden i ying es ic ions
(all o e iden i ied ins umen s) a e co ec (exogenous)
chi2(3) = 4.070
P ob > chi2 = 0.254
Di e ence in Hansen es s o exogenei y o GMM ins umen subse s:
Hansen es excluding SGMM ins umen s (i.e., he di e enced ins u-
men s)
H0: GMM di e enced ins umen s a e exogenous
chi2(2) = 3.570
P ob > chi2 = 0.168
Di e ence in Hansen es s o exogenei y o GMM ins umen subse s:
H0: sys em - GMM ins umen s a e exogenous, and hey inc ease Han-
sen J- es
chi2(1) = 0.500
P ob > chi2 = 0.476
Sou ce: Au ho ’s calcula ions using STATA 15.
Amina Moće ić, Lejla Lazo ić-Pi a
De e minan s o local go e nmen deb in he Fede a ion o Bosnia and He zego ina Financial In e ne Qua e ly 2024, ol. 20 / no. 3
1%) and an inc ease in popula ion densi y by 1 pe son
pe km2 leads o a educ ion o LGU deb by 623 BAM
on a e age (ce e is pa ibus, s a is ically signi ican a
5%).
The long- un coe icien sugges s ha all a iables
ha we e s a is ically signi ican in he sho un a e
also impo an in he long un and ha e he same di ec-
ion o impac on local deb (Table 3).
cen age poin is associa ed wi h an inc ease in LGU
deb by 44,405.06 BAM (ce e is pa ibus, s a is ically
signi ican a 5%). Simila ly, an inc ease in he a e age
sala y by 1 BAM leads o an inc ease in LGU deb on
a e age by 2,376 BAM (ce e is pa ibus, s a is ically sig-
ni ican a 10%). An inc ease in he numbe o inhabi -
an s by 1 leads o an a e age inc ease in he LGU deb
by 36 BAM (ce e is pa ibus, s a is ically signi ican a
Table 3: De e minan s o LGUs deb in FBiH in he long un
The dependen a iable is deb on he subcen al le el – LGUs in cu en BAM (DEB)
Independen a iables Coe icien s S anda d e o z
EXP 109,797.000 37,657.140 2.92
TRANS 1.860 0.754 2.47
LAWSPECdummy 2,958,532.000 875,035.400 3.38
CANTONSPECdummy -1,663,164.000 725.820 -2.29
ACTIVITY 5,876.000 3,084.000 1.91
DENSITY -1,540.540 725.360 -2.12
POP 91.100 28.140 3.24
Independen a iables P>|z| [95% In e al]
EXP 0.004 35,991.170 183,604.50
TRANS 0.014 0.382 3,341.00
LAWSPECdummy 0.001 1,243,494.000 4,673,570.00
CANTONSPECdummy 0.022 -3,085,746.000 -240,582.00
ACTIVITY 0.057 -167.950 11,921.00
DENSITY 0.034 -2,962.230 -118.85
POP 0.001 35.950 146.25
Sou ce: Au ho ’s own wo k.
indings because hey ha e indica ed a posi i e impac
o ans e s o highe le els on he public bo owing o
LGUs. The speci ics o he legal egula ions wi hin he
FBiH on public deb con ibu ed o he ac ha ans-
e s ha e a posi i e impac on public deb . In he LGUs
o FBiH ans e s oge he wi h o he LGU e enues,
p esen he basis o de e mining he uppe limi o he
pe mi ed deb le el. In o he empi ical wo ks (Feld and
Ki chgassne , 2006) au ho s ind ha he subcen al
le els o de eloped coun ies ha e a s ong in e es in
main aining hei so e eign y and he e o e y o
a oid a si ua ion whe e hey mus be escued (bailed
ou ) by a highe le el o go e nmen . Au onomy and
so e eign y in decision-making is a esou ce ha he
subcen al le els o de eloped coun ies s i e o p e-
se e wi hou comp omise. Unlike de eloped coun-
ies, BiH/FBiH has s ill no eached ha le el o de el-
opmen wi hin he amewo k o public inances, which
can be seen h ough he legisla ion i sel . The cu en
legal amewo k does no ecognize his mu ual in lu-
ence be ween he wo pilla s o iscal ede alism
( ans e s and public bo owing).
Ou esul s also include ime dummy a iables o
he yea s in which local and gene al elec ions we e
Resul s o ou esea ch lead o he answe o ou
i s esea ch ques ion which implies ha he dynamics
o deb o LGUs in FBiH a e signi ican ly de e mined by
se s o iscal, ins i u ional, and demog aphical de e mi-
nan s bo h in he sho and in he long e m.
Coun y speci ic indings o he a iable LAW-
SPECdummy indica e ha special expendi u e needs
o LGUs ha ha e mo e han 60,000 inhabi an s and/
o inance ma e ial cos s in p ima y schools (LGUs o
He zego ina-Ne e a Can on, Wes He zego ina Can-
on and Can on 10) o whe e he LGU is assigned spe-
cial expendi u e needs, inc eases deb in hose LGUs.
CANTONSPECdummy indica es he ac ha LGUs be-
longing o can ons ha ha e de ined special expendi-
u e needs in he Law on Alloca ion o Public Re enues
in FBiH bo ow less han LGUs belonging o can ons
ha do no ha e special expendi u e needs. Howe e ,
esea ch ha analyses deb wi hin a coun y mus con-
side he speci ics o de ined in e go e nmen al iscal
ela ions.
The s udies ha used in e go e nmen al ans e s
in hei speci ica ion a e C op and Wendel (1998), Feld
and Ki chgassne (2006), Kim and Lim (2018). In hei
analysis, ans e s lead o a lowe le el o deb (bail-ou
e ec ). Ou esul s a e no complemen a y o o he
