Dis ibu i e Poli ics and Spa ial Equi y: The Alloca ion o Public In es men
in Chile
DOI:10.1080/00343404.2017.1309013
Felipe Li e
Economics and Business Facul y, Albe o Hu ado Uni e si y, E asmo Escala 1835,
San iago 8340539, Chile. Tel: (+0056) 28897366. E-mail: [email p o ec ed]
Xabie Gainza
Depa men o Applied Economics I, Uni e si y o he Basque Coun y UPV/EHU, A da.
Lehendaka i Agui e 83, Bilbao 48015, Spain. Tel: (+0034) 946013639. E-mail:
[email p o ec ed]
Abs ac
This pape analyzes he in luence o elec o al conce ns on in es men dis ibu ion om
he cen al go e nmen o Chilean municipali ies. D awing upon panel da a, he pape
shows ha in es men is media ed by po k-ba el and poli ical budge cycles, as well as
a ou ing he ela i ely be e -o a eas. Es ima ions also e eal ha esou ces a e
channelled o he municipali ies whe e he o e ma gin in local elec ions is la ge whe eas
na ional esul s a e no ele an , indica ing ha local go e nmen s’ lobbying capaci y is
p io i ized o e na ional elec o al in e es s. Based on hese esul s, he implica ions o
egional go e nance and o educing he ma gin o a bi a y alloca ions a e discussed.
Polí ica Dis ibu i a y Equidad Espacial: La Asignación de la In e sión
Pública en Chile
Resumen
This is an Accep ed Manusc ip e sion o he ollowing a icle, accep ed o publica ion in Li e , F., & Gainza, X. (2017). Dis ibu i e poli ics and
spa ial equi y: he alloca ion o public in es men in Chile. Regional S udies, 52(3), 403–415, h ps://doi.o g/10.1080/00343404.2017.1309013. I is
deposi ed unde he e ms o he C ea i e Commons A ibu ion-NonComme cial-NoDe i a i es License (h p://c ea i ecommons.o g/licenses/by-
nc-nd/4.0/), which pe mi s non-comme cial e-use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed, and is
no al e ed, ans o med, o buil upon in any way.
Es e a ículo analiza la in luencia de ac o es elec o ales en la dis ibución de la in e sión
desde el gobie no cen al a los municipios chilenos. A pa i de da os de panel, el a ículo
mues a que la in e sión es á condicionada po una polí ica de co e elec o alis a y po el
ciclo polí ico, además de a o ece las á eas más p óspe as. Las es imaciones ambién
e elan que los ecu sos se di igen a los municipios donde el ma gen de ic o ia en las
elecciones locales es mayo mien as que los esul ados nacionales no son ele an es, lo
que indica que la p esión de los gobie nos locales es decisi a en e a los in e eses
elec o ales nacionales. A pa i de es os esul ados, se discu en las implicaciones pa a la
gobe nanza egional y pa a educi el ma gen de una asignación a bi a ia.
Keywo ds: Dis ibu i e poli ics; po k-ba el; poli ical budge cycle; equi y-e iciency
ade-o ; egional go e nance; Chile
JEL: D72; H50; H72; R58
In oduc ion
How is public in es men alloca ed? Is i dis ibu ed in esponse o social equi y o
economic e iciency c i e ia, o media ed by poli ical conside a ions? Economic and
egional de elopmen li e a u e has adi ionally analyzed he ole o public in es men
as a ade-o be ween equi y and e iciency (Richa dson, 1979). This deba e has ecen ly
in ensi ied wi h he con ibu ions o he new economic geog aphy li e a u e, which s a e
ha spa ial agglome a ions o economic ac i i y may bene i na ional g ow h and, hus,
policies ha seek o educe egional economic inequali ies may in ac be na ionally
ine icien (Ma in, 2008). In e mingled in his deba e s and he ques ions abou he spa ial
o ganiza ion o he s a e and whe he decen aliza ion may a ou a mo e e icien and
balanced spa ial pa e n o , con e sely, be a u he sou ce o spa ial inequali ies and
ine iciencies ( o a heo e ical discussion, see Rod íguez‐Pose and Gill (2005); o
empi ical e idence o he e ec s o de olu ion on egional dispa i ies Rod íguez-Pose
and Ezcu a (2009) and Lessmann (2012)).
Beyond e iciency and equi y conside a ions, in he pas wo decades a bu geoning
li e a u e on dis ibu i e poli ics has documen ed he signi icance o pa isan and elec o al
conce ns o public esou ce dis ibu ion in many coun ies a ound he wo ld (Golden and
Min, 2013; K amon and Posne , 2013). The unde lying hypo hesis is ha poli icians a e
mo i a ed by he desi e o e ain public o ice and, consequen ly, elec ed o icials alloca e
speci ic ypes o goods o speci ic g oups o cons i uen s in elec o al dis ic s a speci ic
imes in he elec o al cycle (Golden and Min, 2013, p.78). Poli ical in e e ence includes
manipula ing iscal a iables along he elec o al cycle o con ince o e s ha he
poli icians ha e ecen ly been doing ‘an excellen job’, i.e. he poli ical business cycle
(Rogo , 1990), showing poli ical a ou i ism o cul u ally cons uc ed popula ion
subg oups (K amon and Posne , 2013) o bene i ing pa icula a eas o maximize hei
e-elec ion chances, i.e. ‘po k-ba el’ (Ta i s, 2009).
This pape analyzes he exis ence o poli ical conce ns in he dis ibu ion o in es men
unds om he cen al go e nmen o municipali ies in Chile, and i s implica ions o
e i o ial cohesion. In pa icula , i d i es a en ion o h ee aspec s o dis ibu i e
poli ics. Fi s , he exis ence o elec o al conce ns in in es men dis ibu ion is explo ed
and con as ed wi h he capaci y o municipali ies o gene a e e enues and o o mula e
in es men p ojec s. The aim is o es whe he public in es men , when media ed by
po k-ba el and poli ical business cycle conside a ions, pu sues e iciency by
concen a ing on he ela i ely be e -o a eas, a he expense o spa ial equi y. Second,
he pape explo es how g an alloca ion a ies along he elec o al cycle in bo h,
municipali ies ha a e uled by poli ical pa ies o he cen al go e nmen ’s uling
coali ion, and i al municipali ies. Thi d, i examines whe he he cen al go e nmen
ans e s esou ces o he municipali ies whe e he o e ma gin in local and na ional polls
is la ge o , con e sely, whe e compe i ion is s onge . The pu pose is o de e mine no
only whe he co e o swing municipali ies a e bene i ed, bu also i dis ibu ion ollows
na ional go e nmen ’s in e es o , on he con a y, is due o municipal go e nmen s’
capaci y o pu p essu e on he cen al go e nmen , encou aged by hei elec o al powe .
Based on hese esul s, he pape explo es he implica ions o a ea go e nance and he
mechanisms o educe he chances o a poli ical bias on in es men und alloca ions.
The analysis d aws on panel da a o public in es men unds om he cen al go e nmen
o he 345 Chilean municipali ies o e he pe iod 2004-2014. The pe iod co e s h ee
local polls and h ee na ional polls, in which he winning coali ion was om bo h poli ical
b anches, wo le -wing and one igh -wing. The poli ical a iables conside ed a e
municipal and na ional elec o al esul s, whe eas socioeconomic da a includes municipal
e enues pe capi a, municipal s a expenses pe capi a, popula ion and he pe cen age
o people below he na ional po e y line. The econome ic analysis elies on ixed-e ec s
and gene alised me hod o momen s (GMM) (A ellano and Bo e , 1995; Blundell and
Bond, 1998) o con ol o po en ial endogenei y.
This pape adds h ee no el ies o he li e a u e on he ela ionship be ween dis ibu i e
poli ics and egional de elopmen . Fi s ly, i analyzes how he bu den o using in es men
o elec o al conce ns is dis ibu ed along he elec o al cycle and be ween allies and i al
local go e nmen s. O he s udies o which he au ho s a e awa e ocus on he a ia ions
o iscal ans e s o axes in elec ion yea s, bu how he e ec s o his iscal manipula ion
a e dis ibu ed along he e m in o ice and ac oss locali ies wi h di e en poli ical
alignmen emains unde s udied. Ano he o iginal con ibu ion is ha i is es ed whe he
poli ical in luence is exe cised o bene i he elec o al p ospec s o he cen al go e nmen
o , con e sely, i is due o local go e nmen s’ lobbying capaci y in a bo om-up p ocess.
This esul is pa icula ly impo an because unde s anding he na u e o he poli ical
game is undamen al o he success o any policy mean o educe he ma gin o
disc e iona y alloca ions. One las no el y is ha he pape explo es, o he i s ime, he
exis ence o poli ical in luence on in es men g an ans e s o Chilean municipali ies.
P e ious analyses ocused on he in luence o economic a iables o p esiden ial
elec ions (Ce da and Ve ga a, 2007) and on he access o publicly unded bene i s om
he pe spec i e o o e s (Cal o and Mu illo, 2012) bu , o he bes o he au ho s’
knowledge, no s udy has e ealed he in luence o poli ical ac o s in public in es men
dis ibu ion in Chile. The exis ence o a poli ical bias is ele an because, being Chile a
highly unequal coun y, he goal o balancing socio-spa ial di e ences may be
unde mined i edis ibu ion is media ed by elec o al conce ns. Beyond his pa icula
case, he pape also discusses he ins i u ional mechanisms o cope wi h poli ical
dis o ions, con ibu ing in his way o he deba e abou hei implica ions o egional
de elopmen .
The emainde o he pape is o ganized as ollows. In he nex sec ion he heo e ical
amewo k is buil om he li e a u e on dis ibu i e poli ics. In sec ion h ee, he case
s udy is cha ac e ized in e ms o Chile’s socio-spa ial cha ac e is ics, he mul i-le el
go e nance sys em and he sub-na ional inance mechanisms. Sec ion ou p esen s he
da a and he me hodology o he empi ical analysis. In sec ion i e he esul s a e
explained. The pape concludes by con ex ualizing he esul s and discussing hei
implica ions o local go e nance.
Dis ibu i e poli ics and egional de elopmen
In b oad e ms, dis ibu i e poli ics e e s o how public au ho i ies con e geog aphically
concen a ed bene i s while di using cos s ac oss o e s (Weingas , Shepsle and Johnsen,
1981). This de ini ion, hough, may be oo b oad, since i includes po k-ba elling as well
as esou ce alloca ion in lagging a eas, which may be a poli ical goal unde e i o ial
cohesion c i e ia. To na ow he concep ual amewo k, S okes e al. (2013) di e en ia e
among dis ibu i e s a egies acco ding o hei p og amma ic s. non-p og amma ic
na u e. Fo a dis ibu ion o be p og amma ic, ules o dis ibu ion ha e o be o malized
and public, and hey ha e o shape he dis ibu ion o esou ces, whe eas in non-
p og amma ic dis ibu ions he c i e ia o alloca ions a e no ende ed public.
Wi hin non-p og amma ic dis ibu ions, wo o he mos sp ead poli ical dis o ions a e
concen a ing esou ces in pa icula geog aphical a eas wi h elec o al mo i s and
manipula ing he iming o he iscal a iables. Poli ical business (o budge ) cycle (PBC)
is he e m used o iden i y he a ia ions on budge expendi u e o axes along he
elec o al cycle. The exis ence o a PBC has been widely s udied by mac oeconomis s,
showing ha mone a y and iscal policies end o be expansiona y be o e an elec ion
(Wille and Keil, 2004). Empi ical analyses ha e demons a ed he exis ence o a PBC
in se e al coun ies a ound he wo ld, bu he magni ude di e s ac oss coun ies. This is
so because PBCs depend on he ins i u ional condi ions unde which incumben poli ical
leade s ha e g ea e abili y and incen i es o manipula e policy a iables along he
elec o al cycle (Al and Rose, 2006).
Al hough he bulk o he li e a u e on PBCs has used na ional-le el da a and
mac oeconomic a iables, o he s udies ha e ocused on he local le el. In ac , Veiga
and Veiga (2007) a gue ha he hypo hesis o he PBC is be e sui ed o he local le el
because he economic e alua ion o local se ice deli e y, such as educa ion,
in as uc u es o ameni ies, is easie han he economic assessmen o na ional se ices
such as de ence, legal ins i u ions o income edis ibu ion. E idence o manipula ing
iscal a iables a he local le el has been ound in bo h, de eloped and de eloping
coun ies. In F ance, Foucaul , Madies and Pa y (2008) ound ha local go e nmen s
inc eased all ca ego ies o public spending in p e-elec o al pe iods. Likewise,
oppo unis ic beha iou has been epo ed o Po ugal, bu Veiga and Veiga (2007)
sugges ha mayo s ended o educe axes and inc ease expendi u e on highly isible
i ems sho ly be o e elec ions, whe eas expendi u es emained he same o e en dec eased
o he less isible i ems. Along he same line, D azen and Esla a (2010) showed o
Colombian locali ies ha poli icians a ge ed in as uc u e spending p io o elec ions
while o he ypes o expendi u e, such as in e es paymen s, ans e s o e i ees, and
paymen s o empo a y wo ke s, ell in elec ion yea s. In B azil, o e he pe iod 1980-
2005 a dec ease in he iscal su plus occu ed in elec ion yea s because local expendi u e
inc eased while local ax e enues and in es men declined (Saku ai and Menezes-Filho,
2011).
Empi ical analyses on a bi a y spa ial alloca ion o esou ces a e based on wo di e en
o mal models. Fo he ‘co e suppo e model’ poli icians will channel esou ces o
suppo g oups because he op imal s a egy o isk-a e se candida es is o edis ibu e
o co e suppo e s (Cox and Mc Cubbins, 1986). Assuming ha swing g oups a e iskie
in es men s, Cox and Mc Cubbins (1986) p edic ha poli icians will in es li le in
opposi ion g oups, somewha mo e in swing g oups, and he mos in hei suppo g oups.
By con as , Dixi and Lond egan (1996), building on Lindbeck and Weibull (1987), s a e
ha i poli ical pa ies a e equal in hei abili ies o alloca e edis ibu i e bene i s, hey
will woo g oups ha a e mos willing o swi ch hei o es. Al e na i ely, i pa ies di e
in hei abili y o a ge edis ibu i e bene i s o di e en g oups, hey will a ou hei
own co e cons i uencies (‘machine poli ics’).
Empi ical e idence on ‘po k-ba els’ suppo s bo h he co e and he swing o e
hypo hesis o , o be mo e p ecise, co e s. swing ‘dis ic s’ hypo hesis, as esea ch o e all
is based on da a om elec o al uni s a he han indi idual o e s (Golden and Min, 2013).
Mo eo e , he e is no pa e n o ac ical dis ibu ion depending on he elec o al sys em,
he g an alloca ion scheme, he scale, he spending ype o he geog aphical con ex .
Analyses o g an alloca ions o co e a eas include ad anced democ acies such as he U.S.
(Ansolabehe e and Snyde , 2006), G eece (Rod íguez-Pose, Psycha is and Tselios, 2016)
o Scandina ia (Ta i s, 2009), as well as new and de eloping democ acies like Albania
(Case, 2001), Mexico (Cos a-I-Fon , Rod iguez-O eggia and Lunapla, 2003) o Sou h
A ica (K o h, 2014). On he con a y, e idence on a ge ing swing a eas has been
epo ed o coun ies o he Global No h (i.e. Aus alia (Denema k, 2000), Sweden
(Johansson, 2003), England (John and Wa d, 2001), Sou h Ko ea (Kwon, 2005), Canada
(Milligan and Sma , 2005), Po ugal (Veiga and Pinho, 2007)) and he Global Sou h (i.e.
Ghana (Ban ul, 2011) and B azil (B ollo and Nannicini, 2012) (see Table A1 in he
Appendix o a selec ion o he empi ical esea ch).
Se e al s udies ha e analyzed he implica ions o elec o al dis o ions o egional
de elopmen , placing hem wi hin he equi y-e iciency ade-o . Cas ells and Solé-Ollé
(2005) ound ha in Spain e iciency c i e ia played a limi ed ole, while in as uc u e
needs and poli ical ac o s mos ly explained he geog aphical dis ibu ion o in as uc u e
in es men . In Tu key, Luca and Rod íguez-Pose (2015) concluded ha egional
in es men is mo i a ed by poli ics and e iciency, as he s a e concen a es esou ces in
ela i ely be e -o a eas a he han dis ibu ing hem o he poo es a eas. In Ge many,
he councils wi h he same poli ical a ilia ion as he s a e ecei ed mo e g an s, and
e iciency conside a ions appea ed o be less impo an han edis ibu i e conce ns
(Kemme ling and S ephan, 2002). Las , Cado , Rölle and S ephan (2006) ound e idence
o elec o al mo i s o egional anspo in as uc u e in es men in F ance, despi e i s
economic e u ns being low. As hey summa ize g aphically: oads and ailways a e no
buil o educe a ic jams: hey a e buil essen ially o ge poli icians e-elec ed (Cado ,
Rölle and S ephan, 2006, p. 1151).
The Chilean ins i u ional con ex
Chile is a highly cen alized coun y, in spi e o some mild s eps owa ds decen aliza ion
since 2002. Spa ially, i is o ganized in h ee go e nmen ie s. The na ional go e nmen
is he majo ins i u ion as i con ols he di e en minis ies, sub-sec e a ies and
depa men s, and he go e nance o lowe ie s. A he egional scale, he egional
go e nmen has limi ed au onomy because i ep esen s na ional go e nmen ’s in e es s
and i has no e enue aising capaci y o i s own. The egional go e nmen is also he
adminis a o o he Na ional Fund o Regional De elopmen (NFRD). Gi en he
dependence o he egional go e nmen om he na ional go e nmen , he Chilean
go e nance amewo k wo ks in eal e ms as a wo- ie sys em (na ional and local). A
he local le el, comunas a e legally au onomous, bu in p ac ice decision-making capaci y
and au onomy a e limi ed as local au ho i ies a e highly dependen on cen al unds.
Local budge s a e inanced h ough hei own sou ces o e enue, a ho izon al ans e
sys em and he cen al go e nmen g an mechanisms, such as he NFRD. Comunas
gene a e hei own e enues h ough comme cial licenses (34% o he o al), p ope y
axes (32%), ci cula ion pe mi s (11%) and o he ees (OECD, 2013). The high
Models (2) and (3) we e in ended o es ima e how in es men is dis ibu ed along he e m
o o ice, bu while Model (2) cap u es public in es men alloca ions o coali ion- and
non-coali ion mayo s in elec ion yea s, Model (3) iden i ies whe he he e a e di e ences
in he yea s a e local elec ion. Acco ding o he es ima ions, in elec ion yea s in es men
inc eased 7.6 pe cen (Model 2), whe eas in he subsequen yea a e he municipal
elec ion in es men d opped by 9 pe cen and wo yea s a e he municipal elec ion he
dec ease was 8 pe cen (Model 3).
Al hough in es men luc ua es along he e m in o ice coun ywide, he e a e signi ican
di e ences be ween a eas depending on hei poli ical alignmen . Because Models (2)
and (3) wo k wi h in e ac ion e ms be ween he a iables yea and coali ion, he
ollowing equa ions we e calcula ed o cap u e he in e ac ion e ms, and hus e lec he
ime-e ec o he di e en municipali ies:
The es ima ed equa ion om Model (2) is:
Es ima ed pe capi a in es men = 17.61 + 0.0860(coali ion) + 0.0757 (y0) + 0.0641 (y0
x coali ion) + γ Zi + σi + τ + ui (6)
The es ima ed equa ion om Model (3) is:
Es ima ed pe capi a in es men = 17.69 + 0.15 (coali ion) - 0.0895(y1) - 0.0833 (y1 x
coali ion) + γ Zi + σi + τ + ui (7)
Equa ion (6) cap u es he pe o mance o in es men in elec ion yea s. In non-coali ion
municipali ies, in es men inc eased by a ac o o 0.076 in elec ion yea s ( he y0
coe icien ), whe eas in hose aligned wi h he cen al go e nmen , in es men was aised
by a ac o o 0.14. F om Equa ion (6) oo, he di e en le els o in es men ha would
esul depending on he budge cycle (municipal elec ion yea s. non-elec ion yea ) and
he mayo ’s poli ical a ilia ion we e es ima ed (Table 3). Apa om signi ican
di e ences depending on hei poli ical alignmen , hese esul s indica e ha in es men
is mo e s able in non-coali ion comunas because i inc eases less in elec ion yea s and
dec eases less in non-elec ion yea s.
[Inse Table 3]
Table 3. Expec ed alues o in es men acco ding o mayo s’ poli ical a ilia ion and yea
No e: he expec ed alues o in es men we e adjus ed o housand pesos o 2014 and
hese a e ound o pe capi a le el
Based on Model (3), Equa ion (7) displays he in e ac ion be ween mayo s’ poli ical
membe ship and he subsequen yea a e he elec ion. This equa ion shows ha he
bu den alls mos ly on locali ies uled by mayo s o poli ical pa ies in he uling coali ion
a he na ional le el since, gi en he in e ac ion e ms, he ne e ec in coali ion
municipali ies is -0.17 and -0.09 in non-coali ion municipali ies. So, while he le el o
public in es men dec eased in he yea s a e municipal elec ions wi hou poli ical
dis inc ion, comunas wi h coali ion mayo s we e pa icula ly esponsible o ‘ oo ing he
bill’.
Finally, i was es ima ed whe he in es men is di ec ed o co e o swing a eas and
whe he na ional o local elec ion esul s we e aken in o accoun . Models (4) and (5)
included o e ma gin as an indica o o co e o swing municipali ies, bu while Model
(4) conside s municipal elec ion esul s, Model (5) in oduces na ional elec ion da a. The
municipal o e ma gin o he mayo s aligned wi h he cen al go e nmen ’s coali ion
appea s o be ele an , whe eas he na ional o e ma gin is no . Thus, he cen al
go e nmen concen a es mo e in es men in co e comunas, bu only municipal elec o al
p ospec s a e conside ed when dis ibu ing esou ces. These esul s sugges a bo om-up
poli ical in luence as unds a e channelled o a eas whe e coali ion mayo s ha e a bigge
o e ma gin in local polls, while na ional elec ion esul s a e no ele an . The a ionale
may well be ha dis ibu i e poli ics is media ed by local go e nmen s’ abili y o ob ain
mo e unds, encou aged by hei elec o al esul s.
In addi ion, in es men dis ibu ion o coali ion and non-coali ion locali ies depending on
he municipal o e ma gin was es ima ed (see Figu e C1 in he Appendix). The coali ion
a iable changes bo h he in e cep and he slope o he municipal o e ma gin, indica ing
ha municipal o e ma gin inc eases in es men in coali ion locali ies, whe eas is sligh ly
educed in non-coali ion locali ies.
In o de o es o possible endogenei y, Table 4 p esen s GMM esul s o he i e
models. The consis ency o sys em-GMM elies on wo hypo heses: ins umen al
a iables mus no be co ela ed wi h he e o e ms, and a nega i e i s -o de
au oco ela ion (AR1) in esiduals may be obse ed, bu no second-o de au oco ela ion
(AR2). The Hansen es indica es ha ins umen al a iables a e alid, while he A ellano-
Bond es s o AR1 and AR2 show no second-o de se ial au o-co ela ion, hus indica ing
alid GMM es ima ions. Mo eo e , hese a e o e all consis en wi h FE esul s o he
i e models, since all poli ical and socioeconomic a iables excep municipal e enues
appea ed o be signi ican in he GMM es ima ions.
[Inse Table 4]
Table 4. Sys em-GMM es ima ion esul s
Discussion and conclusions
Like o he coun ies o he Global No h and Sou h, in Chile in es men g an alloca ions
a e media ed by elec o al conce ns, and no only by equi y o e iciency goals. The
analysis abo e shows he exis ence o po k-ba el poli ics because he municipali ies wi h
mayo s belonging o a pa y uling he na ional go e nmen ecei e 10 pe cen mo e
in es men unds han he es o he municipali ies. These esul s align wi h o he s udies
ha a gue ha when local (o egional) go e nmen s a e uled by he pa y in cha ge o
he cen al go e nmen , hey ob ain mo e bene i s han hose o he s con olled by he
opposi ion (Be elli and John, 2010; Solé-Ollé, 2013).
Addi ionally, elec ed o icials conside iming since on elec ion yea s in es men is up o
10 pe cen g ea e . The bu den o such a poli ical business cycle is dis ibu ed une enly
ac oss ime and space: he iscal excesses o elec ion yea s a e paid mainly du ing he
subsequen wo yea s and, al hough in es men inc eases on elec ion yea s coun ywide,
i is mainly coali ion municipali ies ha bene i om he poli ical cycle’s luc ua ions.
Subsequen ly, coali ion comunas ha e o pay mainly o his highe le el o in es men ,
whe eas in non-coali ion comunas i emains mo e s able. This esul highligh s he
impo ance o aking a dynamic and spa ial pe spec i e in he analysis o poli ical
business cycles, an issue no adequa ely add essed in he li e a u e. O he s udies ha e
ocused on he oppo unis ic beha iou in elec ion o p e-elec ion yea s, bu
unde s anding how und dis ibu ion a ies h oughou he elec o al cycle and be ween
allies and i als is necessa y o app op ia e ins i u ional esponses o cope wi h iming
in he manipula ion o iscal a iables. Some o hese a e discussed below.
One o he key issues o unde s and he na u e o he dis ibu i e game is de e mining
whose in e es s a e u he ed by a disc e iona y iscal alloca ion. In Chile, municipal
elec ion esul s a e ela ed wi h in es men dis ibu ion, whe eas na ional elec ion esul s
a e no ele an . In es men is channelled o coali ion municipali ies wi h a highe o e
ma gin in municipal elec ions, poin ing o co e comunas as he main bene icia ies o such
in es men . Ou es ima ions, hus, side wi h he empi ical e idence suppo ing a ac ical
dis ibu ion in a ou o co e a eas. Ta i s (2009) a gues ha a ge ing co e municipali ies
is he mos likely s a egy o be ollowed because o e s equi e less s ingen cogni i e
capaci ies since, i he uling pa y is he same on he local and na ional le el, he e is no
con usion abou whom o ewa d. In a simila ein, B ollo and Nannicini (2012) s a e ha
i o e s a e no able o dis inguish he sou ce o ans e s and poli ical c edi spillo e s
occu in a o o municipal go e nmen s, aligned municipali ies ecei e mo e ans e s.
Ne e heless, a he han sugges ing a conse a i e s a egy by he cen al go e nmen o
p o ec i sel om he cogni i e asymme ies o he elec o a e, o he au ho s his esul
highligh s local go e nmen s’ abili y o pu p essu e on cen al policy-making. The
cen al go e nmen decides on disbu semen s, bu municipal elec o al esul s gi e local
poli icians a s onge capaci y o exe cise in luence in he compe i ion o g an s. The
impo ance o lobbying is ein o ced by he ac ha in e e y model s a expendi u es
a e posi i ely ela ed wi h in es men . In his sense, he hypo hesis would be ha comunas
wi h s ong mayo s and la ge bu eauc a ic s uc u es ha e a g ea e capaci y o o mula e
sound p oposals and o pu p essu e on disbu se s.
The la e is a pa icula ly ele an con ibu ion o his pape . In he academic li e a u e
emains unde s udied whe he he dis ibu ion o na ional g an s is media ed by cen al
go e nmen s’ in e es s o local go e nmen s’ powe . Howe e , de e mining i und
alloca ion esponds o op-down o bo om-up p essu es is undamen al o design e icien
con ol mechanisms and has signi ican consequences o egional de elopmen . In a
cen alized s a e mo ing slowly owa ds de olu ion, local go e nmen s’ lobbying
capaci y is a u he dis o ion, pa icula ly when e i o ies a e highly he e ogeneous and
may he e o e ha e di e en , pe haps opposing, in e es s (Rod íguez‐Pose and Gill,
2005). Such dis ibu i e pa e n penalizes he e i o ies no aligned wi h he cen al
go e nmen , bu i could also lead local go e nmen s bene i ed om poli ical a ou i ism
o a iscal ap because he incen i es o p uden local budge s a e supp essed (Psycha is,
Zoi and Iliopoulou, 2015).
Apa om he pa icula conce ns o he elec o al game, he analysis yields impo an
implica ions o e i o ial cohesion. Chile has he hi d-highes Gini coe icien on
income dis ibu ion in Sou h Ame ica (ECLAC, 2014) and also egis e s he second-
highes le el o e i o ial dispa i y among OECD coun ies (OECD, 2013). Unde hese
socio-spa ial condi ions, spa ial equi y seems a a he ele an policy o aim o and ye ,
in es men is concen a ed in municipali ies wi h highe municipal e enues and highe
local s a expenses. De e mining he pa icula causes o such a eg essi e dis ibu i e
pa e n lies beyond he scope o his pape , bu i may well be due o an e iciency bias in
in es men planning. In es men a ou s he ela i ely be e -o because concen a ing
esou ces in hese a eas is usually mo e e icien om a cos -bene i pe spec i e, despi e
inc easing e i o ial dispa i ies. As such, he NFRD is mean o be a egional
edis ibu i e und bu his unc ion is clea ly limi ed, especially i i is media ed by
poli ical in luence.
Di e en policy p ocedu es and ins umen s could na ow he incen i es o a bi a iness
and lead o a mo e equi able spa ial dis ibu ion. Fi s , i non-p og amma ic alloca ions
a e ed by he absence o o malized and public ules o dis ibu ion, policy e o ms
should be o ien ed o inc ease anspa ency and accoun abili y. Indeed, he le el o
in luence o elec o al cycles on he iscal balance depends on he deg ee o anspa ency
(Al and Lassen, 2006). Inc easing anspa ency and accoun abili y mainly equi es
poli ical will, al hough o e coming poli icians’ lack o en husiasm o such e o ms may
be a a he di icul obs acle.
Second, mechanisms o mid- and long- e m in es men planning could be conside ed.
The ac ual design o he NFRD allows local go e nmen s o ely on in es men g an s o
inance ope a ional expendi u es (OCDE, 2013), bu his is a pi all ha dis o s i s
s a egic na u e and inc eases he ma gin o disc e iona y dis ibu ion. In addi ion, he
design o he und gene a es “Decembe e e ”, since i has o be disbu sed be o e ha
iscal yea inishes (Decembe 31s ) in o de no o be penalized in subsequen yea s
(Ta schys, 2003).
In his con ex , he in oduc ion o an independen planning o ice o egional in es men
would help wi h he eco e y o such a s a egic pe spec i e. Fo his body o ope a e
alongside elec o al pu poses i is no only necessa y o sepa a e he poli ical sphe e om
he adminis a i e sphe e, i is likewise necessa y o se up s able inance mechanisms.
Mul i-yea budge ing could be a aluable iscal policy because i can gi e local
go e nmen s g ea e ce ain y abou u u e unding, which enables hem o plan and
manage hei expendi u e mo e e ec i ely. Al hough mul i-yea budge ing is no ee o
poli ical clou , i could dec ease poli ical in luence i he ime ame o he budge ing
pe iod and in es men planning di e s om he e m o o ice. Besides, conside a ion
can be gi en o he use o ixed mul i-yea ceilings, simila o he Uni ed Kingdom o
Sweden (Hawkeswo h, Melcho and Robinson, 2012).
A hi d policy change should go in he di ec ion o designing ins i u ional mechanisms o
educe he gap be ween comunas. F om he au ho s’ pe spec i e, one o he mos s iking
esul s o he abo e econome ic analysis is he impo ance o municipal e enue and s a
expenses o in es men dis ibu ion. Local public employees’ capaci y o design sound
p oposals is essen ial because selec ion is based, in e alia, on he p oposals’ echnical
easibili y. Ho izon al ans e mechanisms should be imp o ed in o de o educe income
dispa i ies. Inc easing he con ibu ions o he equalisa ion und would le el he p o ision
o public goods be ween a eas, while educing di e ences in municipali ies’ capaci ies o
o mula e p ojec s i hey a e used o imp o e local bu eauc acies’ skills and esou ces.
An al e na i e me hod o esol ing his p oblem is explo ing he mechanisms by which
di e en comunas could empo a ily join up when applying o in es men g an s. Local
collabo a ions would no only lead o agglome a ion economies by sha ing s a esou ces
and echnical expe ise, hey would ein o ce he s a egic pe spec i e o in es men
planning oo i , beyond each comuna’s in e es s, he wide con ex we e conside ed.
Le us no sound naï e and p e end ha hese policy ecommenda ions will emo e e e y
poli ical obs acle. The poli ical bias is a cons i u i e elemen o he dis ibu i e game
because poli icians a e o ice-mo i a ed and he elec o a e enjoys ecei ing bene i s, e en
a he expense o ine iciencies imposed on he majo i y. As a esul , any a emp o e ase
comple ely he elec o al in luence will be unsuccess ul. Howe e , hese policy p oposals
would limi he scope o unjus i ied poli ical disc e ion and lay he basis o a mo e
equi able iscal dis ibu ion.
Endno es
1Coali ion and o e ma gin a iables change o e ime and ac oss municipali ies. Yea o
he e m o o ice a iable changes o e ime, bu i is cons an ac oss municipali ies
because elec ions a e held he same day in all municipali ies.
Acknowledgemen s
We a e g a e ul o Joachim Wehne and Ra ael Ho ala-Vall e o hei help on an ea lie
d a o he pape , and Tomasz Mickiewicz and wo anonymous e e ees o hei
commen s and sugges ions h ough he e iew p ocess. We a e also in deb o Susan
And ade o he gene ous assis ance copy-edi ing. The usual disclaime s apply.
Re e ences
Alesina, A. (1989). Poli ics and business cycles in indus ial democ acies. Economic
Policy, 4, 55-98. doi: 10.2307/1344464
Al , J. E. and Rose, S. (2006). Con ex -Condi ional Poli ical Budge Cycles. In C. Boix
and S.C. S okes (Eds.) The Ox o d Handbook o Compa a i e Poli ics (pp. 845-867).
Ox o d: Ox o d Uni e si y P ess.
Al , J. E. and Lassen, D. D. (2006). T anspa ency, poli ical pola iza ion, and poli ical
budge cycles in OECD coun ies. Ame ican Jou nal o Poli ical Science, 50, 530-
550. doi: 10.1111/j.1540-5907.2006.00200.x
Ansolabehe e, S. and Snyde , J. M. (2006). Pa y con ol o s a e go e nmen and he
dis ibu ion o public expendi u es. The Scandina ian Jou nal o Economics 108,
547-569. doi: 10.1111/j.1467-9442.2006.00470.x
A ellano, M. and Bond, S. (1991). Some Tes s o Speci ica ion o Panel Da a: Mon e
Ca lo E idence and an Applica ion o Employmen Equa ions. The Re iew o
Economic S udies, 58, 277-97. doi: 10.2307/2297968
A ellano, M. and Bo e , O. (1995). Ano he look a he ins umen al a iable es ima ion
o e o -componen s models. Jou nal o Econome ics, 68, 29-51. doi:
10.1016/0304-4076(94)01642-D
Ban ul, A. B. (2011). Do o mula-based in e go e nmen al ans e mechanisms
elimina e poli ically mo i a ed a ge ing? E idence om Ghana. Jou nal o
De elopmen Economics, 96, 380-390. doi: 10.1016/j.jde eco.2010.08.012
Be elli, A. M. and John, P. (2010). Go e nmen checking go e nmen : how pe o mance
measu es expand dis ibu i e poli ics. The Jou nal o Poli ics, 72, 545-558. doi:
10.1017/S002238160999082X
Blundell, R. and Bond, S. (1998). Ini ial condi ions and momen es ic ions in dynamic
panel da a models. Jou nal o Econome ics, 87, 115-143. doi: 10.1016/S0304-
4076(98)00009-8
B ollo, F. and Nannicini, T. (2012), Tying you enemy's hands in close aces: he poli ics
o ede al ans e s in B azil. Ame ican Poli ical Science Re iew, 106, 742-761. doi:
10.1017/S0003055412000433
Cado , O., Rölle , L. and S ephan, A. (2006). Con ibu ion o p oduc i i y o po k ba el?
The wo aces o in as uc u e in es men . Jou nal o Public Economics, 90, 1133-
1153. doi: 10.1016/j.jpubeco.2005.08.006
Cal o, E. and Mu illo, M. V. (2012). When Pa ies Mee Vo e s: Assessing Poli ical
Linkages Th ough Pa isan Ne wo ks and Dis ibu i e Expec a ions in A gen ina and
Chile. Compa a i e Poli ical S udies, 46, 851 - 882. doi:
10.1177/0010414012463882
