What Makes Decentralisation in Developing Countries Pro-poor
-
Upload
independent -
Category
Documents
-
view
1 -
download
0
Transcript of What Makes Decentralisation in Developing Countries Pro-poor
What Makes Decentralisation in DevelopingCountries Pro-poor?
JOHANNES JUTTING, ELENA CORSI ,CELINE KAUFFMANN, IDA M CDONNELL,
HOLGER OSTERRIEDER, NICOLAS PINAUDand LUCIA WEGNER
Decentralisation has been advocated by donors and development
agencies as an instrument to ensure broader participation of citizens as
well as to improve local governance leading to poverty reduction from
the bottom up. On the basis of a comprehensive review of nine case
studies documented in the literature, this study questions this assumption.
According to our findings, a clear link between decentralisation and a
reduction in poverty cannot be established. Two important policy lessons
emerge from this study. First, in an environment where the central state
barely fulfils basic functions and is not interested in giving power and
resources to local tiers of government, decentralisation should not be a
priority for donors as it could be rather counter-productive. Second, in
countries that fulfil basic prerequisites, decentralisation could be a
powerful tool for poverty reduction.
La decentralisation a ete mise en avant par les partenaires au
developpement comme un instrument permettant d’assurer la
participation des citoyens a la definition des politiques publiques
ainsi que l’amelioration de la gouvernance locale, debouchant ainsi sur
une approche participative de la reduction de la pauvrete. Sur la base
d’une revue systematique de neuf etudes de cas de reference sur le sujet,
cet article remet en cause cette hypothese. Nous concluons en effet qu’il
est impossible d’etablir un lien clair entre decentralisation et reduction
de la pauvrete. Des lors, ce travail aboutit a deux recommandations
importantes. Tout d’abord, dans un contexte ou l’Etat central parvient a
The European Journal of Development Research, Vol.17, No.4, December 2005, pp.626–648ISSN 0957-8811 print/ISSN 1743-9728 online
DOI: 10.1080/09578810500367649 q 2005 Taylor & Francis
Johannes Jutting is Senior Economist at the OECD Development Centre, Paris; CelineKauffmann, Nicolas Pinaud and Lucia Wegner are Economists at the OECD DevelopmentCentre, Paris; Holger Osterrieder is Programme Officer at the United Nations DevelopmentProgramme Mauritania; and Elena Corsi is Executive Assistant at the Europe ResearchCenter, Harvard Business School. Email: [email protected].
peine a remplir ses missions essentielles et est reticent a confier des
responsabilites et des ressources aux echelons locaux de gouvernement,
la decentralisation ne saurait faire figure de priorite pour les bailleurs.
Par contre, dans des pays qui remplissent les conditions minimales pour
avancer sur la voie de la decentralisation, cette derniere peut constituer
un puissant instrument de reduction de la pauvrete.
INTRODUCTION
Decentralisation has been a major concern of developing countries, the international
development community and researchers for two decades. The debate has mainly
centred on two principal questions: first, what are the main driving forces and reasons
for decentralisation and how can one maximise its overall benefits [e.g.Oates, 1972;
Manor, 1999; Fukasaku and Mello, 1999; Dethier, 2000]? Second, what is the
impact of decentralisation on selected variables of interest like ‘corruption’ [e.g.
Fisman and Gatti, 2002], ‘government responsiveness to local needs’ [e.g. Faguet,
2004], ‘delivery of public services’ [e.g. Litvack and Seddon, 1999; Lieberman,
2002] and ‘political stability’ [e.g. World Bank, 2000]? Interestingly, and despite its
being a ‘hot topic’ for some time now, it is only recently that decentralisation’s
relationship to poverty has received attention [e.g. Crook and Manor, 1998; Von
Braun and Grote, 2002; Crook, 2003; Vedeld, 2003; Asfaw et al., 2004]. The
economic literature on poverty has long ignored the potential importance of
decentralisation in achieving poverty reduction objectives like promoting
opportunities, empowerment, participation, security, and rights for the poor and
excluded at the local level [Von Braun and Grote, 2002].
The objective of this paper is to help fill this gap and identify the determinants
of pro-poor decentralisation processes. Up to the present, the study by Von Braun
and Grote [2002] seems to be the most advanced and in-depth treatment of the
impact of decentralisation on poverty. Based on a rigorous review of the literature
and cross-country comparisons, the authors conclude that decentralisation serves
the poor, but only under specific conditions. They recommend that these
conditions should be analysed within a framework that tackles political, fiscal and
administrative decentralisation simultaneously, while also taking into account
country-specific conditions and different types of decentralisation policies. This
study follows these suggestions by reviewing various case studies addressing the
impact of decentralisation on poverty under a common framework in order to
distil patterns of pro-poor decentralisation policies.
The implementation of decentralisation is usually motivated by the following
two arguments, which are highly relevant for poverty reduction:
. Decentralisation can lead to an increase in efficiency. Decentralisation can be
a tool for central authorities to better identify people’s needs and preferences.
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 627
Central state authorities usually lack the ‘time and place knowledge’ [Hayek,
cited in Ostrom et al., 1993] to implement policies and programmes that
reflect people’s ‘real’ needs and preferences. If properly managed,
decentralisation is seen as a way to improve allocative efficiency [Oates,
1972; Musgrave, 1983].
. Decentralisation can lead to improved governance. Decentralisation
enhances accountability and monitoring of government officials and
decision-makers. Unchecked authority and inadequate incentives encourage
‘rent seeking behaviour’ among government officials. Decentralisation
undermines these opportunities by creating institutional arrangements that
formalise the relationship between citizens and public servants. Political
decentralisation and especially the election of local officials by citizens,
together with a strong legal framework, can create local accountability,
thereby fostering officials’ legitimacy, bolstering citizens’ involvement and
interest in politics and deepening the democratic nature of the country’s
institutions [Crook and Manor, 1998; Manor, 1999; Blair, 2000].
Both arguments are highly relevant for poverty reduction concerns. Increased
possibilities for participation, improved access to services and a more efficient
provision of public goods at the local level are major components of most anti-
poverty programmes. Nevertheless, the linkages might not be so straightforward.
As for poverty itself, decentralisation is a multifaceted concept: its outcome
depends to a large extent on the form and type of decentralisation that one looks
at. Two basic concepts are distinguished: deconcentration and devolution. The
first aims at transferring responsibilities to field and subordinate units of
government where field units basically remain under the hierarchical authority of
central state authorities and have no distinct legal existence from the central state.
In contrast, devolution refers to a transfer of competencies from the central to
distinct legal entities, for example area-wide regional or functional authorities,
non-governmental and private organisations. These legal entities have a wide
autonomy in decision making.
Addressing the issue of pro-poor decentralisation is timely as donors are
increasingly looking for tools to strengthen the poverty focus of their
programmes and policies. With respect to decentralisation and poverty a recent
report from the Development Assistance Committee (DAC) concluded that ‘there
is an urgent need to examine more systematically the conditions under which
decentralisation benefit[s] the poorest section of the population within local
governments [. . .]’ [OECD, 2003].
This paper is organised as follows: the following section introduces a
framework on the linkages between decentralisation and poverty. In the third
section the outcomes of the literature review are reported and characteristics of
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH628
success and failure highlighted. The determinants of pro-poor decentralisation
processes are presented and discussed in the fourth section, while the final section
concludes by presenting some policy implications for donors.
DECENTRALISATION AND POVERTY OUTCOMES: A SIMPLE
FRAMEWORK
The relationship between decentralisation and poverty is quite complex and
depends on the specific definition of both terms. For the purpose of this study,
decentralisation is defined broadly as embracing the transfer of power and
resources from higher tiers to lower tiers of government. The scope and type of
the decentralisation process is addressed by looking at the actual implementation:
whereas devolution is the most far-reaching form of decentralisation,
deconcentration basically only includes the transfer of administrative power.
Poverty is defined in its multidimensional sense focusing particularly on
empowerment and on the access to social services. Both aspects are supposed to
be directly influenced by decentralisation while an income effect would only
materialise at a later stage (Fig. 1).1 Empowerment is measured by looking at
‘effective participation possibilities for all groups of the society’ as well as ‘local
government’s capacity to implement pro-poor policies’. This distinction stems
from the consideration that the participation of the poor in decision-making does
not necessarily imply that local governments will have the power and resources to
implement pro-poor policies or that the voice of the poor will be heard [Mosse,
2001]. Local governments’ capacity to decide and implement pro-poor policies
largely depends on the design of the decentralisation process and it is related to
factors such as local governments’ human and fiscal resources, type of functions
devolved and the extent of elite capture [Bardhan and Mookherjee, 2000].
In theory, an effective devolution of power should increase the incentives and the
capacity of the poor to actively participate in decision-making, to decide and lobby
for their interests [Manor, 1999], bringing about their ‘empowerment’ as well as pro-
poor policies. As the capacity of citizens to monitor local officials and politicians is
higher in a decentralised system, there are opportunities for an increase in
transparency and thus for a reduction in corruption and an overall improvement in
local governance [Manor, 1999; Blair, 2000; Crook and Sverrisson, 2001].
Decentralised forms of government can equally bring about efficiency gains,
especially in service delivery as local officials are supposed to have a better
knowledge of local needs and preferences [Hayek in Ostrom et al., 1993].
Two sets of conditions are crucial determinants for the realisation of the
described positive effects of decentralisation on poverty outcomes as mapped out
in Fig. 1 [Jutting et al., 2004]: (1) country specificities (size of the country, level
of economic development, degree of democracy and capacities) and; (2) the
design of the decentralisation process itself. The latter is dependent on the
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 629
interaction of political factors (commitment to a real devolution process, policy
coherence, donors’ involvement); administrative factors (anti-corruption
measures, division of functions, building local capacity); and fiscal factors
(type and amount of resources devolved).
DECENTRALISATION AND POVERTY: LESSONS FROM SELECTED
COUNTRY EXPERIENCES
The following evaluation is based on reported evidence in the literature on how
decentralisation affected poverty in terms of access to services and
empowerment. The countries selected are thus countries where decentralisation
has taken place and for which evidence on the impact of the reform on poverty
has been collected. Particular attention is given to those countries considered as
having achieved decentralisation. Countries in which decentralisation is an
FIGURE 1
DECENTRALISATION AND POVERTY: BACKGROUND AND PROCESS CONDITIONS
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH630
TA
BL
E1
DE
CE
NT
RA
LIS
AT
ION
AN
DP
OV
ER
TY
OU
TC
OM
ES
INS
EL
EC
TE
DC
OU
NT
RIE
S
Cou
ntr
yA
uth
ors
Acc
ess
tose
rvic
esE
mp
ow
erm
ent
To
tal
sco
reP
arti
cip
atio
nin
loca
lin
stit
uti
on
sL
oca
lg
ov
ernm
ents
’(L
G)
cap
acit
yto
dec
ide/
imple
men
tp
ro-p
oo
rp
oli
cies
Boli
via
Alt
man
&L
alan
der
,2
00
3;
Fag
uet
,2
00
3,
20
04
;C
IE-
SIN
,2
00
3;
Jett
e,2
00
5.
–In
crea
sed
acce
ssto
hea
lth
and
educa
tion;
–L
arg
ein
ves
tmen
tb
yL
Gs
intr
ansp
ort
,w
ater
infr
astr
uct
ure
,ed
uca
tion
and
sanit
atio
n.
–B
road
par
tici
pat
ion
of
the
po
or
inlo
cal
inst
itu
tio
ns
(e.g
.comites
devigilian-
cia)
and
loca
lel
ec-
tio
ns.
–L
arg
ed
evo
luti
on
of
po
wer
s,re
sou
rces
and
resp
on
sib
ilit
ies
toL
G;
–L
Gis
resp
on
sib
lefo
rp
ub
lic
serv
ices
,th
atis
,o
wn
ersh
ipo
fin
fras
tru
ctu
rean
dre
spo
nsi
bil
ity
tom
ain
tain
,eq
uip
,ex
pan
dan
dad
min
-is
ter
faci
liti
es.
–G
oo
dre
spo
nsi
ven
ess
tolo
cal
nee
ds.
–In
som
em
un
icip
alit
ies
elit
eca
ptu
reis
repo
rted
.
***
Ind
ia(K
eral
a)N
aray
ana
&K
uru
p,
20
00;
Nar
ayan
a2
00
5;
Hel
ler,
20
01.
–In
crea
sed
acce
ssto
pri
mar
yh
ealt
hca
re.
–B
road
par
tici
pat
ion
of
the
po
or
and
low
erca
stes
inlo
cal
inst
i-tu
tion
san
dlo
cal
elec
tio
ns.
–L
arg
ed
evo
luti
on
of
po
wer
s,re
sou
rces
and
resp
on
sib
ilit
ies
toL
G;
–D
iscr
etio
nar
ypow
erover
spen
din
gan
dst
aff;
–G
oo
dre
spo
nsi
ven
ess
tolo
cal
nee
ds.
***
TA
BL
E1
–Continued
Cou
ntr
yA
uth
ors
Acc
ess
tose
rvic
esE
mp
ow
erm
ent
To
tal
sco
reP
arti
cip
atio
nin
loca
lin
stit
uti
on
sL
oca
lg
ov
ernm
ents
’(L
G)
capac
ity
tod
ecid
e/im
ple
men
tp
ro-p
oo
rp
oli
cies
Ind
ia(W
est
Ben
gal
)
Cro
ok
&S
ver
riss
on
,2
00
1;
Mat
hew
&M
athew
,2
00
3;
Vo
nB
rau
n&
Gro
te,
20
02.
–In
crea
sed
acce
ssan
dq
ual
ity
of
bas
icse
rvic
es.
–B
road
par
tici
pat
ion
of
the
po
or
and
low
erca
stes
inlo
cal
inst
i-tu
tion
san
dlo
cal
elec
tio
ns.
–L
arg
ed
evo
luti
on
of
po
wer
s,re
sou
rces
and
resp
on
sib
ilit
ies
toL
G;
–D
iscr
etio
nar
ypow
erover
spen
din
gan
dst
aff;
–Y
etli
mit
edd
ecis
ion
-mak
-in
gau
ton
om
yin
the
hea
lth
sect
or
and
po
ver
tyal
lev
ia-
tio
np
rog
ram
mes
;–
Go
od
resp
on
siv
enes
sto
loca
ln
eed
s.
***
Ph
ilip
pin
esR
oo
d,
19
98;
Bir
d&
Ro
dri
-g
uez
,1
99
9;
Tan
zi,
20
00
;A
ng
eles
&M
agno
,2
00
4.
–A
cces
sto
serv
ices
slig
htl
yin
crea
sed,
wit
hre
gio
nal
dis
par
itie
s;–
Wid
eco
rru
pti
on
inth
eed
uca
tion
syst
em,
par
-ti
ally
faci
lita
ted
by
dec
entr
alis
atio
n;
–D
ecli
ne
inth
ep
rop
ort
ion
of
stu
den
tsco
mp
leti
ng
pri
mar
y-l
evel
educa
tion.
–B
road
par
tici
pat
ion
of
the
po
or
inlo
cal
inst
itu
tion
san
dlo
cal
elec
tio
ns;
–R
egio
nal
dif
fere
nce
sin
par
tici
pat
ion
and
stro
ng
elit
eca
ptu
reo
flo
cal
po
liti
cs.
–L
arg
ed
evo
luti
on
of
po
wer
s,re
sou
rces
and
resp
on
sib
ilit
ies
toL
G;
–F
un
ctio
ns
such
ash
ealt
hca
re,
infr
astr
uct
ure
,so
cial
serv
ices
,en
vir
on
men
tan
dn
atura
lre
sou
rces
man
age-
men
td
evo
lved
toL
G;
–L
arg
ein
ves
tmen
tin
soci
alan
dec
on
om
icse
rvic
esan
din
serv
ice
del
iver
y;
–P
arti
cipat
ory
met
hods
imp
rove
loca
ld
ecis
ion
s.–
Eli
teca
ptu
re,
the
pre
sen
ceo
flo
cal
bo
sses
and
fam
ily
po
licy
lim
its
LG
resp
on
-si
ven
ess
toth
ep
oo
r.
**
Gh
ana
Vo
nB
rau
n&
Gro
te,
20
02
,A
san
te&
Ay
ee,
20
04,
Cro
ok
,2
00
3
–In
crea
sed
qu
alit
y/q
uan
tity
of
road
s;–
Acc
ess
tohea
lth
and
edu
cati
on
par
tial
lyin
crea
sed
,yet
elit
eca
ptu
reo
fsu
bsi
die
s.
–B
road
par
tici
pat
ion
of
the
po
or
inlo
cal
inst
itu
tio
ns.
–L
arg
ed
evo
luti
on
of
po
wer
s,re
sou
rces
and
resp
on
sib
ilit
ies
toL
G;
–Im
pro
vin
glo
cal
hu
man
and
tech
nic
alca
pac
ity;
–L
ow
,y
etim
pro
vin
g,
resp
on
siv
enes
so
fL
Gto
loca
ln
eed
s.
**
So
uth
Afr
ica
Hel
ler,
2001;
Fri
edm
an&
Kih
ato
,2
00
4;
Cam
ero
n,
20
02.
–B
ette
rin
fras
tru
ctu
re;
–In
crea
sin
gre
gio
nal
dis
-p
arit
ies
inh
ealt
han
ded
uca
tio
n.
–Im
pro
ved
par
tici
-p
atio
nin
elec
tio
ns
for
the
po
or
–d
ue
also
tod
emo
cra-
tisa
tio
n.
–L
arg
ed
evo
luti
on
of
po
wer
san
dre
spo
nsi
bil
itie
s,y
etst
ron
gce
ntr
alco
ntr
ol;
–L
ow
erca
pac
ity
inru
ral
area
s;–
Pro
ble
ms
inin
terp
reti
ng
the
outc
om
esof
par
tici
pat
ory
met
ho
ds.
**
Ug
and
aF
ran
cis
&Ja
mes
,2
00
3;
Sco
tt-H
erri
dg
e,2
00
2;
Wo
rk,
20
02;
Fo
ster
&M
iju
mb
i,2
00
2;
PD
M2
00
3;
Sch
ott
on
,2
00
4.
–R
educe
dvac
cinat
ions;
–D
eter
iora
ting
infr
astr
uc-
ture
and
red
uce
dd
eliv
ery
of
serv
ices
just
ified
maj
or
centr
alst
ate
con
tro
lan
dd
irec
tin
terv
enti
on
;–
Imp
rov
emen
tsin
serv
ice
del
iver
yan
dp
ov
erty
red
uct
ion
mai
nly
du
eto
cen
tral
stat
ein
terv
enti
on
Incr
ease
dp
oli
tica
lp
ar-
tici
pat
ion
yet
elit
eca
p-
ture
of
loca
lp
oli
tics
;
–L
oca
lco
ntr
ol
ov
erlo
cal
staf
fan
dbro
addel
egat
ion
of
po
wer
s;–
Dir
ect
centr
alst
ate
inte
r-v
enti
on
inso
me
sect
ors
and
lim
ited
loca
lco
ntr
ol
ov
erfi
nan
cial
reso
urc
esre
du
ces
LG
’sd
ecis
ion-m
akin
gca
pac
ity;
–L
imit
edlo
cal
hu
man
and
tech
nic
alca
pac
ity;
–L
oca
ld
ecis
ions
are
scar
cely
tak
enin
toco
nsi
der
atio
n.
*
TA
BL
E1
–Continued
Cou
ntr
yA
uth
ors
Acc
ess
tose
rvic
esE
mp
ow
erm
ent
To
tal
sco
reP
arti
cipat
ion
inlo
cal
inst
itu
tion
sL
oca
lg
ov
ernm
ents
’(L
G)
cap
acit
yto
dec
ide/
imple
men
tp
ro-p
oo
rp
oli
cies
Gu
inea
Boss
uy
t&
Go
uld
,2
00
0;
Lo
quai
,2
00
4;
PD
M,
20
03
.–
Wea
klo
cal
serv
ice
del
iv-
ery
capac
ity;
–S
om
eim
pro
vem
ents
inac
cess
tobas
icse
rvic
es.
–In
suffi
cien
to
pp
ort
u-
nit
ies
for
loca
lp
ar-
tici
pat
ion
.
–S
carc
elo
cal
po
wer
san
dre
sou
rces
;–
Lac
kof
hum
anca
pac
ity;
–S
tro
ng
cen
tral
stat
eco
ntr
ol
ov
erlo
cal
po
liti
cs;–
Apriori
con
tro
lo
ver
loca
lp
oli
cies
.
–
Ind
ia(M
adh
ya
Pra
des
h)
McC
arte
n&
Vy
asu
lu,2
00
4;
Joh
nso
n,
20
03;
Nar
ayan
a,2
00
5.
–A
cces
sto
edu
cati
on
slig
htl
yin
crea
sed
.–
Lim
ited
par
tici
pat
ion
of
the
po
or,
wo
men
and
soci
ally
excl
ud
edin
loca
lin
stit
uti
on
s;–
Incr
ease
dpar
tici
-p
atio
nin
use
rg
rou
ps.
–V
ery
lim
ited
po
wer
san
dre
spo
nsi
bil
itie
sd
eleg
ated
toL
Gs;
–N
ore
ald
ecis
ion-m
akin
gre
spo
nsi
bil
ity
for
loca
lg
ov
ernm
ents
;–
LG
acts
asag
ent
of
cen
tral
go
ver
nm
ent
–
Definitionofcategories
:—
***:decentralisationwithanoverallpositive
impact
onpovertyreported
.—
**:decentralisationwithsomepositive
impactsonpovertyreported
.—
*:decentralisationwithmixed
impact
—-:
dec
entr
alis
atio
nw
ith
no
posi
tive
(or
neg
ativ
e)im
pac
ton
pover
tyre
port
ed.Source:
bas
edo
nJu
ttin
get
al.
,2
00
4.
TA
BL
E2
DE
TE
RM
INA
NT
SF
OR
SU
CC
ES
SF
UL
PR
O-P
OO
RD
EC
EN
TR
AL
ISA
TIO
N
Countr
yS
core
(fro
mT
able
1)
Poli
tica
lfa
ctors
Adm
inis
trat
ive
fact
ors
Fis
cal
fact
ors
Boli
via
***
–D
ecen
tral
isat
ion
by
des
ign;
–P
oli
cyco
her
ence
;–
NG
Oin
volv
emen
tuse
of
par
tici
pat
ory
met
hods.
–S
taff
trai
nin
gan
dce
ntr
alst
ate
support
;–
Cle
ardiv
isio
nof
funct
ions,
yet
stro
ng
centr
alre
gula
tion;
–A
nti
-corr
upti
on
mea
sure
s.
–S
table
tran
sfer
s(c
ondit
ional
and
unco
ndit
ional
)an
dlo
cal
tax-r
aisi
ng
capac
ity;
–Y
et,
incr
easi
ng
condit
ional
tran
sfer
san
dce
ntr
alst
ate
contr
ol
over
expen
dit
ure
s.In
dia
(Ker
-al
a)***
–D
ecen
tral
isat
ion
by
des
ign;
–P
oli
cyco
her
ence
;–
Sea
tsre
serv
edin
loca
lbodie
sfo
rw
om
enan
dlo
wer
cast
es,
use
of
par
tici
pat
ory
met
hods,
tran
spar
ency
,in
stit
uti
ons
of
dir
ect
dem
oc-
racy
(pet
itio
ns,
refe
rendum
).
–S
trong
supp
ort
from
cent
ral
gover
nm
ent;
–S
taff
trai
nin
g;
–P
eopl
e’s
Cam
pai
gn
for
Dec
entr
alis
edP
lan-
nin
g
–S
table
tran
sfer
san
dlo
cal
tax-
rais
ing
capac
ity.
India
(Wes
tB
engal
)
***
–D
ecen
tral
isat
ion
by
des
ign;
–P
oli
cyco
her
ence
;–
Sea
tsre
serv
edin
loca
lbodie
sfo
rw
om
enan
dlo
wer
cast
es,use
of
par
tici
pat
ory
met
hods,
inst
ituti
ons
of
dir
ect
dem
ocr
acy
(pet
itio
ns,
refe
rendum
).
–C
lear
div
isio
nof
funct
ions;
–Y
et,
centr
ally
defi
ned
dev
elopm
ent
pro
gra
mm
esan
dce
ntr
alst
ate
contr
ol
over
hea
lth.
–S
table
tran
sfer
san
dlo
cal
tax-
rais
ing
capac
ity.
Phil
ippin
es***
–D
ecen
tral
isat
ion
by
des
ign;
–P
oli
cyco
her
ence
;–
NG
O’s
par
tici
pat
ion
indec
isio
ns-
mak
ing
and
monit
ori
ng;
use
of
par
tici
pat
ory
met
hods,
inst
rum
ents
of
dir
ect
dem
ocr
acy
inpla
ce(p
etit
ions,
refe
rendum
)
–S
taff
trai
nin
g-
som
ere
gio
ns
stil
lla
ckca
pac
ity;
–Q
uit
ecl
ear
div
isio
nof
funct
ions
-st
ill
stro
ng
pre
sence
of
centr
alst
ate
inhea
lth
expen
dit
ure
s;–
Eli
teca
ptu
re;
–A
nti
-corr
upti
on
mea
sure
s(w
atch
dog
gro
ups)
.
–S
table
and
pre
dic
table
tran
s-fe
rs:
fixed
shar
eof
gro
ssre
ven
ues
(inte
rnal
reven
ue
allo
tmen
t);
–T
ransf
ers
not
bas
edon
spec
ific
regio
nal
nee
ds
and
uneq
ual
dis
trib
uti
on;
–L
Gpow
erto
crea
teow
nre
ven
ue
and
levy
taxes
,fe
esan
doth
ers;
–L
arge
fisc
alau
tonom
y,
yet
incr
easi
ng
dep
enden
ceon
tran
sfer
s.
TA
BL
E2
–Continued
Countr
yS
core
(fro
mT
able
1)
Poli
tica
lfa
ctors
Adm
inis
trat
ive
fact
ors
Fis
cal
fact
ors
Ghan
a**
–D
ecen
tral
isat
ion
by
des
ign;
–S
om
epoli
cyin
coher
ence
;–
Par
tici
pat
ory
pla
nnin
gpro
cedure
s,in
stit
u-
tional
ised
rela
tionsh
ips
bet
wee
nel
ecte
dan
del
ecto
rate
.
–C
entr
alst
ate
inte
rfer
esin
loca
lpoli
tics
and
appoin
ts30%
of
loca
las
sem
bly
mem
ber
s;–
Eli
teca
ptu
re.
–P
oli
cyin
coher
ence
lim
its
loca
lre
sourc
es.
South
Afr
ica
**
–D
ecen
tral
isat
ion
by
des
ign;
–P
arti
cipat
ory
met
hods,
NG
Oin
volv
emen
t,
acti
ve
pre
ss;
Cen
tral
ised
stru
cture
of
poli
tica
lpar
tyre
duce
sdow
nw
ard
acco
unt-
abil
ity.
–S
taff
trai
nin
g;
–E
xce
ssiv
ere
gula
tion
of
loca
lfu
nct
ions
and
shar
edar
eas
of
auth
ori
tyli
mit
slo
cal
auto
nom
y;
–A
nti
-corr
upti
on
mea
sure
s.
–S
table
and
pre
dic
table
tran
s-fe
rs,
yet
lim
ited
;
–H
igh
LG
reli
ance
on
loca
lta
xes
,yet
som
epro
ble
ms
of
loca
lta
xev
asio
n.
Ugan
da
*–
Dec
entr
alis
atio
npar
tial
lyby
def
ault
and
par
tial
lyby
des
ign;
–In
crea
sing
centr
alst
ate
com
mit
men
t;–
Som
epoli
cyin
coher
ence
,al
soby
donors
;
–S
eats
rese
rved
inlo
calbodie
sfo
rw
om
enan
dad
opti
on
of
par
tici
pat
ory
met
hods.
–S
om
etr
ainin
gfo
rlo
cal
staf
f;–
Rec
ent
effo
rts
from
the
gover
nm
ent
toin
crea
seL
Gca
pac
ity;
–Q
uit
ecl
ear
div
isio
nof
funct
ions,
yet
stro
ng
centr
alst
ate
inte
rfer
ence
by
sect
or;
–S
trong
elit
eca
ptu
rew
ith
rece
nt
effo
rts
tore
duce
it.
–L
oca
lre
sourc
esst
emm
ing
most
lyfr
om
tied
tran
sfer
s;–
Dec
entr
alis
atio
na
chal
lenge
tolo
cal
tax
coll
ecti
on.
Guin
ea–
–D
ecen
tral
isat
ion
by
def
ault
;–
Cen
tral
opposi
tion;
–P
oli
cyin
coher
ence
,al
soby
donors
;–
No
acco
unta
bil
ity
and
no
bott
om
up
dec
isio
nm
akin
gpro
cess
es;
–In
form
atio
non
LG
acti
vit
ies
lim
ited
by
languag
epro
ble
ms.
–N
oca
pac
ity
buil
din
gef
fort
s;–
No
clea
rdiv
isio
nof
funct
ions
and
lack
of
know
ledge
on
who
does
what
;S
trong
centr
alst
ate
contr
ol
(apriori
contr
ol)
.
–L
imit
edre
sourc
es;
–L
imit
edan
dra
ndom
centr
al
tran
sfer
s;–
Loca
lta
xev
asio
n.
India
(Mad
hya
Pra
des
h)
––
Dec
entr
alis
atio
nby
def
ault
;–
Res
ista
nce
from
bure
aucr
acy;
Par
tici
pat
ory
mea
sure
sin
pla
cean
dse
ats
rese
rved
inlo
cal
bodie
sfo
rw
om
enan
dlo
wer
cast
es.
–S
taff
trai
nin
g;
–U
ncl
ear
div
isio
nof
funct
ions;
–S
trong
elit
eca
ptu
re.
–L
imit
edre
sourc
es.
incomplete process have also been selected. A more comprehensive overview of
these and other country examples can be found in Jutting et al. [2004]
The selected case studies are evaluated in terms of the achievement of
decentralisation under the two areas of poverty outcomes identified in our
framework (Figure 1). There are four different categories: i) positive impact;
ii) somewhat positive impact; iii) mixed impact; and iv) no/negative impact. For
each category a score ranging from three stars (positive) to no star
(decentralisation with no positive (or negative) impact on poverty reported)
has been given (see footnote to Table 1). The summing up of the individual scores
for the two areas of poverty outcomes then leads to the overall classification
(column 6 in Table 1). While it would be useful to report hard evidence about
how many poor people have actually benefited in terms of improved access to
services such information is not available in a systematic comparable manner.
Table 1 should be interpreted with caution for the following reasons: first,
mainly peer-reviewed articles were selected to guarantee a minimum quality
standard for the studies included with the disadvantage that a lot of grey
literature, which is generally more recent, had to be excluded. In some cases the
decentralisation process has changed considerably since the publication of these
articles potentially leading to quite different poverty outcomes. Second, it is
possible that not all of the peer-reviewed articles available have been found so the
reported evidence might not give the full picture. Third, the use of different
methodologies/conceptual and analytical frameworks by the selected authors
partially limits cross-country comparisons. Despite these caveats, the table
provides very interesting general patterns of success or failures of decentralisa-
tion for poverty reduction.
That the positive performers (Bolivia and the Indian states of West Bengal
and Kerala) also adopted a comprehensive decentralisation approach,
concurrently undertaking political, fiscal and administrative decentralisation
might explain their relative success. The process went beyond deconcentration to
a real delegation of power to lower tiers of government, with support from central
government and the adoption of parallel policies to overcome eventual side
effects of decentralisation like increasing regional disparities.
The somewhat positive performers (South Africa, Philippines and Ghana) are
characterised by relatively successful decentralisation programmes, with an
identifiable impact on poverty. Local government’s autonomy is partially limited
by the central state’s detailed definition of local functions and procedures (South
Africa) or the presence of some central state appointees in the governing local
bodies (Ghana). This has reduced the capacity of local government to implement
locally decided policies, and had a negative impact on participation
(empowerment). In the case of the Philippines, the widespread corruption and
an unequal distribution of central state transfers limited local government
capacity to realise pro-poor policies. Nevertheless, in these countries,
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 637
achievements which can be ascribed to decentralisation have been made. In
Ghana the construction of new roads and the provision of subsidises for the poor
have increased their access to services, even if part of these subsidies are captured
by the rich. Political participation of the poor has increased and recently local
governments seem to be responding more to local needs [Asante and Ayee, 2004].
South Africa reported some progress in housing, infrastructure and participation
even if the degree of progress achieved has been questioned [Heller, 2001].
The mixed and negative performers are characterised by uncompleted
decentralisation processes (Uganda where fiscal decentralisation lags behind),
limited delegation of powers (Madhya Pradesh) or failed decentralisation (Guinea
with no impact on poverty or possibly an adverse effect on it). Local government
mainly acts as an agent of central authorities, with no decision-making power and
scarce resources. When it does have power, the paucity of human and fiscal resources
and the lack of autonomy brought a negative impact on service delivery, as in the
case of Uganda’s health sector [Scott-Herridge, 2002].
The comparison of these nine case studies suggests that there is a link between
decentralisation and poverty; nonetheless this is quite complex. Furthermore, although
it appears that the chances of pro-poor decentralisation increase with the level of a
country’s overall development – all negative performers are least developed countries
(LLDCs) while most of the positive performers are middle-income countries –
important exceptions, such as the Indian states of West Bengal and Kerala, remain.
The study by Jutting et al. [2004] on a larger sample of countries suggests that country
characteristics are less important than the design of the process when looking at the
implementation and impact of decentralisation. Consequently, and following the
framework presented in Figure 1, the next section deals in more detail with the
different process conditions of decentralisation in the case study countries.
WHAT DETERMINES PRO-POOR OUTCOMES OF DECENTRALISATION
POLICIES?
In the following we compare the different experiences within and between the
groups of performers highlighting similarities and differences in the process of
decentralisation followed. In line with the framework presented earlier, the focus
will be on political, administrative and fiscal factors. Table 2 presents a summary
of the main findings.
Political Factors: Commitment, Policy Coherence, Transparency and
Participation Commitment
The first variable analysed in Table 2 (column 2) is the country’s commitment
to decentralisation, that is, if the decentralisation reform has been adopted
by design or by default. Decentralisation by design refers to countries that
decentralised believing that effective decentralisation could serve their policy
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH638
objective, be this poverty reduction, efficiency or widening democracy. All
positive and rather positive performers believed in the benefits of decentrali-
sation and were committed to delegating powers and responsibilities. Where
governments were committed, the process was better planned, resources and
functions were delegated and human local capacity was taken care of. As local
government received more powers and resources its ability to have a positive
impact on poverty increased. The most prominent case in this respect is probably
West Bengal where pro-poor decentralisation was possible thanks to the
commitment of the pro-poor communist party [Crook and Sverrisson, 2001].
On the other hand, decentralisation for the negative performers has been the
result of international pressure (Guinea) or was seen as a means to consolidate
central power against ethnic or political rivalry, yet with no commitment from the
central government to effectively empower local government, delegate functions
or achieve pro-poor outcomes: decentralisation by default. It is of particular
importance that commitment concerns the central bureaucracy and not only the
political leadership (Madhya Pradesh); see also Ribot [2002].
Internal Policy Coherence
Most countries witnessed some form of policy incoherence because the central state
adopted policies that partially undermined the power of local authorities by
indirectly diminishing resources or functions. A typical example is the adoption of
sectoral policies and the use of sectoral funds by the central government to finance
interventions in health, water, sanitation and infrastructure while simultaneously
devolving these tasks to local government. This is less of a problem in the positive or
rather positive countries where they are surmountable limitations (e.g. Ghana);
however, they are real obstacles to pro-poor decentralisation in the others where such
incoherence hides an unwillingness to decentralise. Policy incoherence was reported
in Table 2 only when it had a particular relevance.
For many of the positive and rather positive performers policy coherence
implies also that decentralisation is part of a broader agenda of reforms
undertaken by the government. Most of them adopted prior and/or parallel to
decentralisation, economic liberalisation and democratisation reforms (e.g.
Ghana, South Africa, Philippines). In India (West Bengal) decentralisation
reforms have been linked to a comprehensive and unusually successful land
reform programme. These accompanying measures can help to counter the
adverse effects of decentralisation in terms of elite capture and corruption.
Transparent and Participative Process
In all the countries reviewed the decentralisation process has been democratic
(i.e. delegation of powers to locally elected institutions). Local government is
elected, even if in some cases local democracy suffers from a direct central state
interference in politics and a lack of democracy at the national level (Guinea) or
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 639
is weakened by the presence of appointees from the central government in their
governing bodies (Ghana).
Contrary to the negative performers, positive performers have sufficient
information flows, thus citizens are informed about the activities of local
institutions and are more able to participate and to monitor local officials. For
instance, the media is very active in South Africa; it issues regular statements
about local government activities in India and Bolivia. It is also noteworthy that
in positive performing countries, it is the local government that provides
substantial information on decentralisation and public policy issues, which seems
to be a result of greater local capacity. In some of the negative performers, the
information is provided almost exclusively from the central government in an
effort to hold local government accountable and to assert its role in the political
process (e.g. Uganda).
Besides providing for local elections, positive and rather positive countries
have also created democratic procedures at the local level to allow for people’s
involvement in public affairs by other, institutionalised, means. This has
stimulated broader and more extensive involvement of previously excluded
segments of the population. Positively performing countries provide for the use of
participatory planning procedures, for example to determine the level of revenue
to be transferred to governments (the Gram Sansad performs this function in
India). India, since 1993, reserves seats in the local governments for specific
groups of the population (women and lower castes) and provides for the
institutionalisation of village assemblies charged with monitoring, and presenting
policy suggestions to, municipalities as well as decision-making at the local level.
In all negatively performing countries participation of the poor in local
institutions is low and elite capture high. This tends to reflect a top-down culture
of politics, and distrust towards the elected communal representatives. In spite of
efforts to provide adequate information on local government activities and the
introduction of some participatory institutions, citizens’ opportunities to
participate remain insufficient.
A strong civil society and the adoption of measures to include it in the process
of decentralisation seem to have further reinforced pro-poor outcomes of
decentralisation processes. Local civil society plays an important role for a
transparent and participative process, insofar as it can exercise pressure on the
governments and control their actions (e.g. South Africa). Yet this depends on
the organisational structure of civil society, its composition, its ‘civic-ness’, its
strength, independence of the state and its capacity to represent and promote the
interests of the poor and socially excluded [Narayana, 2005]. For example, in
Ghana elected assemblies are bound to encourage and support community self-
help groups. At the same time, these associations are largely dominated by the
elites, thus the provision does not empower the poor and the socially excluded
[Crook, 2003]. In South Africa civil society is facing serious financial problems.
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH640
The intervention of the state to support some groups is not always beneficial as it
erodes their independence [Friedman and Kihato, 2004]. On the other hand,
countries that are less successful in reaching pro-poor outcomes through
decentralisation are characterised by a weak, often elite-dominated civil society
(e.g. Uganda).
Administrative Factors: Countering Corruption, Capacity Building and the Role
of the Central State
Elite capture and corruption
Table 2 underlines the emergence of elite capture and corruption when these
factors constituted an obstacle to pro-poor outcomes. In reality, there is no
example of the non-occurrence of elite capture or corruption across all nine case
studies. Community-based participatory approaches are not a panacea. Never-
theless, positive performers tend to be less marked by elite capture and corruption
than negative performers. At least, in these cases, there is an awareness of the
need for action and the countries concerned are putting in place measures to
counter corruption and to monitor administrative behaviour (e.g. inspectors and
social audit in India; investigating commissions appointed by the provinces in
South Africa; vigilance committees in Bolivia). In negatively performing
countries, decentralisation is often seen as part of a patrimonial agenda aimed at
preserving the monopoly of power and ensuring control over resources.
Capacity building
Most countries in the process of decentralisation experience some form of
interference from central government. This can be direct (e.g Guinea in local
politics) or indirect as a consequence of policy incoherence. Central governments
tend to legitimise their interference in local policies by pointing to a lack of capacity
in local government. In reality, this argument hides mainly their unwillingness to
delegate power. While all positive performers are characterised by strong local
human capacity, often it does not precede the decentralisation reforms. Table 2
shows that all positive performers invested in building local capacity. At the same
time, the literature gives evidence that local government capacity can increase as a
consequence of decentralisation [Rondinelli et al., 1989]. The devolution of powers
can generate ‘learning-by-doing’ processes which can help local government to
build up capacity through practice. Common to negative performers is the
considerable lack of pre-existing local human capacity and very little support from
the central state in favour of training during reform.
Central/local powers
A key issue which emerges from the review of the case studies is the role of the
central state in a decentralised system. While local government can be an
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 641
important means for creating enabling environments for poverty reduction,
the overall responsibility for implementing pro-poor policies remains with the
central state. This is even more essential in environments with high inequalities
based on traditional social institutions, e.g. gender or caste. Recent research on
India suggests that giving power to local tiers of government is not sufficient to
increase the participation of marginalised groups, in particular women
[Narayana, 2005]. The central state has to ensure that existing social inequalities
are taken into account and are not re-enforced once decentralisation has started.
Indeed this seems to have been the case in Uganda where judicial reform
favoured local councils that often discriminate against women. To have a pro-
poor impact, decentralisation should be accompanied by complementary
measures such as investment in education or the promotion of land reform.
Whereas it is undisputed that for public goods with inter-jurisdictional spill-over
effects (e.g. vaccination), the central state should continue to take the lead in their
delivery with local governments playing a subsidiary role, the debate on the role
of the state is still not settled when it comes to those goods that are not
exclusively public such as health, education, water and sanitation.
While the case studies demonstrate that positively performing countries had a
clearer division of functions between the centre and the periphery, it does not
shed light on which functions should be delegated and to what extent.
Nevertheless, it seems that even in a decentralised system, the central
government has an important role to play.
Fiscal: Stable and Type of Resources
The last column of Table 2 shows that decentralisation has been successful, in
terms of pro-poor objectives, only when local governments disposed of sufficient
resources to fulfil their new tasks. A study on India found a significant correlation
between the health conditions of the rural poor, measured by the infant mortality
rate in rural areas, and fiscal decentralisation [Asfaw et al., 2004]. Lack of
resources and freedom in allocating them seem to have been a general problem
for those countries in which decentralisation did not bring about pro-poor
outcomes.
Basically, there are three ways in which local authorities can obtain resources:
through transfers from the central governments, by raising their own taxes and
from donors’ contributions. However, there are no data in the case studies for the
latter. In many of the positively performing countries, resources at the local level
have come both from transfers and local taxes (e.g. Bolivia). Independent and
substantial tax-raising powers seem to be a major criterion for successful pro-
poor decentralisation, e.g. China [Von Braun and Grote, 2002]. However, these
tend to increase regional inequalities from an economic and social point of view
as happened in China. Needs-based transfers from the central government are
helpful in targeting the poor [Bird and Rodriguez, 1999]. Transfers in general, as
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH642
long as they are transparent, stable and predictable, have a positive impact on
decentralisation as they can alleviate regional inequalities, enable local
government to achieve bottom-up, pro-poor policies while at the same time
grant that main services are delivered (e.g. Kerala). In Bolivia 20 per cent of the
budget is automatically allocated to local governments each year. However,
transfers should not be tied by the central government and constitute the only
funding of local policies as they will leave local governments with no decision-
making autonomy, e.g. Uganda [Scott-Herridge 2002]. An important element for
successful poverty-focused decentralisation is whether local governments have
the power to allocate resources. Limited or no ability to decide on expenditure is
an impediment to pro-poor decentralisation as it leads to poor local
responsiveness to local needs. While the central government can still follow
successful pro-poor policies (as is the case of Uganda), the benefits which might
stem from bringing governments closer to the people are not fully exploited.
All of the negative performers are characterised by extremely limited
financial resources, that is to say, limited (or tied) transfers and limited or no local
tax-raising powers. Although it is certainly true that unrestricted power to decide
on expenditures or arbitrary decisions bears the risk of increased elite capture and
corruption and can be a threat to macroeconomic stability, e.g. Philippines
[Tanzi, 2000], freedom to decide how to spend resources is essential for effective
pro-poor decentralisation. Many authors underline the importance of the structure
of the transfer system and call for the establishment of monitoring instruments
[Bird and Rodriguez, 1999; Crook, 2003; Crook and Sverrisson, 2001].
The last column of Table 2 also highlights a common problem in local tax-raising
capacity, which is more evident in negatively performing countries due to the
phenomena of local tax evasion. This can be ascribed to a lack of tax-raising
legitimacy of local governments (e.g. Guinea and South Africa to a certain extent). In
Uganda, tax evasion is aggravated by the unwillingness of local authorities to impose
taxes (mainly those on the rich) and the recent reliance on private collectors that tend
to capture part of the resources [Francis and James, 2003].
CONCLUSIONS
This paper assesses the conditions under which decentralisation most likely leads
to pro-poor outcomes. Despite reviewing the literature for nine, regionally
dispersed, countries it is nearly impossible to find hard, measurable evidence of
the impact of decentralisation on poverty. It is hence difficult to establish a clear
link between decentralisation and concrete/measurable improvements in access
to services and empowerment. This is an important area for future research.
Nevertheless, reviewing the reported outcomes in the nine countries, and
linking them back to inputs and process, provides interesting patterns of pro-poor
decentralisation processes. It appears that the theoretical benefits of
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 643
decentralisation for poverty reduction – increased efficiency and improved
governance for more demand-oriented social service provision – are far from
assured. On the contrary, while positive experiences in terms of improved access
to services and empowerment have been reported in some countries – notably
Bolivia and in some Indian states – the picture is rather gloomy in most African
countries. Reforms in African countries have been limited to administrative
deconcentration – with very little or no positive impact on poverty reduction.
The political economy of reform processes in countries characterised by weak
governance systems and capacities often makes a full delegation of powers and
resources unlikely. Hence, key factors of success for pro-poor decentralisation
seem to be a real devolution of power and resources while establishing
accountability systems. As decentralisation involves a redistribution of power
and can generate opposition from those who will be negatively affected by it, it
can only be effective and generate pro-poor outcomes if there is a strong
commitment from central government and political leaders: ownership and
objectives of the reforms are decisive. While there is potential for poverty
reduction through democratic decentralisation, there are also some hidden
dangers such as reinforcing local elites. Finally, the review of case studies
suggests that in a decentralised system the central state retains an important role
in leading and monitoring the reform process. Two important policy lessons for
donors can be taken from these findings:
1. Donors should be more aware of the political economy of decentralisation as a
change process. Decentralisation is a political process that will not produce
gains on all sides: as with all reforms it creates winners and losers. Hence, it
will be important to ensure that the winners become reform ‘champions’ and
that the losers have a forum where their complaints will be dealt with. Various
forms of compensation should be considered. Donors should promote reforms
on several levels within the government, including at the intermediate levels
that have an administrative and political role to play. Otherwise reforms will
not produce sustainable results. While planning entities might be more willing
to engage in decentralisation, the ministries of finance and/or internal affairs
may be more reluctant when it comes to managing the process.
2. Donors should emphasise the instrumental character of decentralisation in
creating an enabling environment for poverty reduction at local levels.
Decentralisation should be promoted as an important instrument for poverty
reduction, not as an end in itself. As the link between decentralisation and
poverty reduction is not straightforward, donors should work to straighten this
link by putting pressure on governments for pro-poor outcomes, constantly
monitoring the process and adopting flexible, impact-oriented (learning-by-
doing) and country-specific policies. In order to ensure effective monitoring,
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH644
the adoption of service delivery standards and pro-poor desired outcomes is
required.
In all, the basic conclusion from this study is that to achieve pro-poor
decentralisation, commitment from the national elite and the support of the
population is fundamental, while sound donor support could substantially help to
achieve the task. Where there is no support for decentralisation or the state is
collapsing, donors should redirect their efforts to build this support, or, in the
second case, focus first on state-building, as the role of the central state in the
decentralisation process is crucial.
ACKNOWLEDGEMENTS
The authors would like to thank three anonymous referees for their detailed and
most helpful comments on earlier versions of this paper. All authors write in their
individual capacities.
NO TE
1. In a more comprehensive review by Jutting et al. [2004] including 19 case study countries, none ofthe studies analysed income effects, while the majority looked at participation and access/use ofsocial services. A more comprehensive framework including the income dimension has beendeveloped by Steiner [2005].
REFERENCES
Altman, David and Rickard Lalander, 2003, ‘Bolivia’s Popular Participation Law: An UndemocraticDemocratisation Process?’, in Axel Hadenius (ed.), Decentralisation and DemocraticGovernance: Experiences from India, Bolivia and South Africa, Stockholm: Almqvist andWiksell International, pp.63–104, , www.egdi. gov.se/pdf/study/study2003_3.pdf . .
Angeles, Leonora C., and Francisco A. Magno, 2004, ‘The Philippines: Decentralisation, LocalGovernments and Citizens Action’, in Oxhorn et al., pp.211–265.
Asante, Felix A. and Joseph R.A. Ayee, 2004, ‘Decentralisation and Poverty Reduction’, paperpresented at the international conference on ‘Ghana at the Half Century’, July 18–20, Institute ofStatistical, Social and Economic Research (ISSER), University of Ghana and Cornell University,Accra, Ghana, ,www.isser.org/Decentralisation_Asante_Ayee.pdf .
Asfaw, Abay, Klaus Frohberg, K.S. James and Johannes Jutting, 2004, ‘Modelling the Impact ofFiscal Decentralisation on Health Outcomes: Empirical Evidence from India’, Discussion Paperson Development Policy 87, Zentrum fur Entwicklungsforschung (ZEF), Bonn: University ofBonn.
Bardhan, Pranab K. and Dilip Mookherjee, 2000, ‘Capture and Governance at Local and NationalLevels’, American Economic Review, Vol.90, No.2, pp.135–139.
Bird, Richard and E.R. Rodriguez, 1999, ‘Decentralisation and Poverty Alleviation, InternationalExperience and the Case of the Philippines’, Public Administration and Development, Vol.19,pp.299–319.
Blair, Harry, 2000, ‘Participation and Accountability at the Periphery: Democratic Local Governancein Six Countries’, World Development, Vol.28, No.1, pp.21–39.
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 645
Bossuyt, Jean and Jeremy Gould, 2000, ‘Decentralisation and Poverty Reduction: Elaborating the
Linkages’, Policy Management Brief, No.12, ECDPM, Maastricht, The Netherlands,
,www.oneworld.org/ecdpm/pmb/b12_gb.htm . .
Cameron, Robert, 2002, ‘Central-Local Financial Relations in South Africa’, Local Government
Studies, Vol.28, No.3, pp.113–134.
CIESIN, 2003, ‘Case Study: Bolivia in’, The Online Sourcebook on Decentralisation and Local
Development, realised by the Centre for International Earth Science Information Network, with
the collaboration of SDS, FAO, UND, GTZ, WB, New York: Columbia University,
,www.ciesin.org/decentralisation/English/CaseStudies/Bolivia.html . .
Crook, Richard, 2003, ‘Decentralisation and Poverty Reduction in Africa: The Politics of Local-
Central Relations’, Public Administration and Development, Vol.23, pp.77–88.
Crook, Richard and James Manor, 1998, Democracy and Decentralisation in South Asia and West
Africa, Cambridge: Cambridge University Press.
Crook, R. and A.S. Sverrisson, 2001, ‘Decentralisation and Poverty-Alleviation in Developing
Countries: A Comparative Analysis or, is West Bengal Unique?’, in IDS Working Paper, No. 130
Brighton: Institute of Development Studies.
Dethier, Jean Jacques (ed.), 2000, Governance, Decentralisation and Reform in China, India and
Russia, Dordrecht: Kluwer Academic Publishers.
Faguet, Jean-Paul, 2003, ‘Decentralisation and Local Government in Bolivia: An Overview from the
Bottom Up’, in Crisis State Programmer, Working Papers Series, No. 1 (May), London: London
School of Economics, , www.crisisstates.com/download/wp/WP29JPF.pdf . .
Faguet, Jean-Paul, 2004, ‘Does Decentralisation Increase Responsiveness to Local Needs? Evidence
from Bolivia’, Journal of Public Economics, Vol.88, pp.867–894.
Fisman, Raymond and Roberta Gatti, 2002, ‘Decentralisation and Corruption: Evidence Across
Countries’, Governance Policy Research Working Papers, No.2290, Washington, DC: The
World Bank, , http://econ.worldbank.org/docs/1048.pdf . .
Foster, Mick and Peter Mijumbi, 2002, ‘How, When and Why Does Poverty Get Budget Priority?
Poverty Reduction Strategy and Public Expenditure in Uganda’, Working Paper, No.163,
London: Overseas Development Institute (ODI).
Francis, Paul and Robert James, 2003, ‘Balancing Rural Poverty Reduction and Citizen Participation:
The Contradictions of Uganda’s Decentralisation Programme’, World Development, Vol.31,
No.2, pp.325–337.
Friedman, Steven and Caroline Kihato, (2004) ‘South Africa’s Double Reforms: Decentralisation and
the Transition from Apartheid’, in Oxhorn et al. (2004), pp.141–189.
Fukasaku, Kiichiro and Luiz De Mello, 1999, Fiscal Decentralisation in Emerging Economies, Paris:
OECD Development Centre Seminars.
Heller, Patrik, 2001, ‘Moving the State: The Politics of Democratic Decentralisation in Kerala, South
Africa, and Porto Alegre’, Politics and Society, Vol.29, No.1, pp.131–163.
Jette, Christian, 2005, Democratic Decentralisation and Poverty Reduction: The Bolivian Case, Oslo:
United Nations Development Programme, Oslo Governance Centre.
Johnson, Craig, 2003, ‘Decentralisation in India. Poverty, Politics and Panchayati Raj’, in ODI
Working Papers, No.199, London: Overseas Development Institute (ODI).
Jutting, Johannes, Celine Kauffmann, Ida McDonnell, Holger Osterrieder, Nicolas Pinaud and Lucia
Wegner, 2004, ‘Decentralisation and Poverty in Developing Countries: Exploring the Impact’,
Working Paper, No. 236, Paris: OECD Development Centre.
Lieberman, S.S, 2002, Decentralization and Health in the Philippines and Indonesia: An Interim
Report, Washington, DC: The World Bank, , http://www1.worldbank.org/publicsector/
LearningProgram/Decentralization/decenhealth.pdf . .
Litvack, Jennie and Martin Ravallion, 2000, ‘Decentralisation, Equity and Service Provision’,
Seminar Series on Decentralisation: Working Paper, No.2, Washington, DC: The World Bank.
Litvack, Jennie and Jessica Seddon, 1999, ‘Decentralisation Briefing Notes’, World Bank Institute
Working Papers, No.37142. Washington, DC: The World Bank, , www.worldbank.org/wbi/
publications/wbi37142.pdf . .
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH646
Loquai, Christiane, 2004, ‘Decentralisation and Poverty Reduction in a Non Performing State’,presentation prepared for the conference on ‘Decentralisation and Poverty’, OECD DCD/DEVworkshop, Paris, 29–30 Sept. 2004.
Manor, James, 1999, The Political Economy of Democratic Decentralisation, Washington, DC: TheWorld Bank.
Mathew, George and Anand Mathew, 2003, ‘India: Decentralisation and Local Governance: HowClientelism and Accountability Work’, in Axel Hadenius (ed.), Decentralisation and DemocraticGovernance: Experiences from India, Bolivia and South Africa, Stockholm: Almqvist andWiksell International, pp.13–62, , www.egdi.gov.se/pdf/study/study2003_3.pdf . .
McCarten, William and Vinod Vyasulu, 2004, ‘Democratic Decentralisation and Poverty Reductionin Madhya Pradesh: Searching for an Institutional Equilibrium’, Development in Practice,Vol.14, No.6, pp.733–740.
Mosse, David, 2001, ‘‘‘People’s Knowledge’’, Participation and Patronage: Operations andRepresentations in Rural Developmen’, in Bill Cooke and Uma Kothari (eds.), Participation:The New Thiranny?, London: Zed Books, pp.16–35.
Musgrave, Richard A., 1983, ‘Who Should Tax, Where and What?’, in Charles E. Mc Lure (ed.), TaxAssignment in Federal Countries, Canberra: Centre for Research on Federal Financial Relations.
Narayana, Dalampadi, 2005, ‘Institutional Change and its Impact on the Poor and Excluded: theIndian Decentralisation Experience’, Working Paper No.242, Paris: OECD Development Centre.
Narayana, Dalampadi and K.K.H. Kurup, 2000, Ensuring Access and Accountability: The Challengesof Decentralizing the Health Care Sector in Kerala, Kerala: Centre for Development Studies(CDS).
Oates, Wallace, 1972, Fiscal Federalism, New York: Harcourt Brace.OECD-DAC, 2003, ‘Synthesis Study on Supporting Decentralisation and Local Governance- Lessons
Learned, Good Practices and Emerging Issues’ DAC Network on Development Evaluation,DCD/DAC/EV (2003)3/REV1, Paris, 20 Oct.
Ostrom, Elinor, Sharon Huckfeldt, Charles Schweik and Mary Beth Wertime, 1993, ‘A RelationalArchive for Natural Resources Governance and Management’, International Forestry Resourcesand Institutions (IFRI) Working Paper, No.D93I-25, Workshop in Political Theory and PolicyAnalysis, Bloomington, IN: Indiana University.
Oxhorn, Philip, Joseph S. Tulchin and Andrew D. Selee (eds.), 2004, Decentralisation, DemocraticGovernance and Civil Society in Comparative Perspective: Africa, Asia and Latin America,Washington, DC: Woodrow Wilson International Centre for Scholars.
PDM, 2003, Etat de la decentralisation en Afrique, Observatoire de la Decentralisation,Paris/Cotonou: Partenariat pour le Developpement Municipal (PDM) et Karthala.
Ribot, Jesse, 2002, ‘African Decentralisation. Local Actors, Powers and Accountability’, Democracy,Governance and Human Rights, Paper, No.8, Geneva: United Nations Research Institute forSocial Development (UNRISD).
Rodinelli, Dennis A., James McCollough and Ronald W. Johnson, 1989, ‘Analysing DecentralizationPolicies in Developing Countries: A Political Economy Framework’, Development and Change,Vol.20, pp.57–87.
Rood, Steven, 1998, ‘Decentralisation, Democracy and Development’, in David Timberman, NicholasPlatt, Paul D. Hutchcroft, Emmanuel S. de Dios, Solita Collas-Monsod, Steven Rood, Jose T.Almonte, Mona Lisa Yuchengco and Rene P. Ciria-Cruz, The Philippines: New Directionsin Domestic Policy and Foreign Relations, Columbia International Affairs Online:,www.ciaonet.org/book/ass01/ass01_d.html. . .
Scott-Herridge, R., 2002, ‘Decentralisation- Does it Deliver Good Governance and ImprovedServices? The Experience of Uganda’, African Studies Centre Occasional Paper Series, No.10Coventry: Coventry University.
Shotton, Roger (ed.), 2004, Local Government Initiative: Pro-Poor Infrastructure and ServiceDelivery in Rural Asia. A Synthesis of Case Studies, New York: United Nations CapitalDevelopment Fund (UNCDF).
Steiner S., 2005, ‘Decentralisation and Poverty Reduction: A Conceptual Framework for theEconomic Impact’, German Overseas Institute, Working Paper Series, No.3, Hamburg.
DECENTRALISATI ON I N DEVELOPI NG COUNTRI ES 647
Tanzi, Vito, 2000, ‘Some Politically Incorrect Remarks on Decentralisation and Public Finance’, inDethier (ed.), pp.47–64.
Vedeld, Trond, 2003, ‘Democratic Decentralisation and Poverty Reduction – Exploring theLinkages’, Forum for Development Studies, Vol.2, pp.159–203.
Von Braun, Joachim and Ulrike Grote, 2002, ‘Does Decentralisation Serve the Poor?’, in EtishamAhmad and Vito Tanzi (eds), Managing Fiscal Decentralisation, London: Routledge,pp.92–119.
Work Robertson, 2002, Overview of Decentralisation Worldwide: A Stepping Stone to ImprovedGovernance and Human Development, paper presented at the 2nd international conference ondecentralisation, ‘Federalism: The Future of Decentralizing States?’, organised by the UNDP,Manila, Philippines, 25–27 July.
World Bank, 2000, World Development Report 2000/2001: Attacking Poverty, Oxford: OxfordUniversity Press.
THE E UROPEAN J OURNAL OF DEVELOPMEN T R ESEARCH648