Efficiency and equity consequences of decentralizationin health: an economic perspective
Abstract
A B S T R A C T – Over the recent years, decentralization has been adopted in many health systems. Thequestion however remains of whether local actors do better than the central government.We summarize the main insights from economic theory on the impact of decentraliza-tion and its empirical validation. Theory suggests that the decision to decentralize resultsfrom a trade-off between its advantages (like its capacity to cater to local tastes) and costs(like inter-regional spillovers). Empirical contributions point that decentralization results inbetter health outcomes and higher expenditures, resulting in ambiguous consequences onefficiency; equity consequences are controversial and address the relevance of redistributionmechanisms.
Full text
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
www.else ie .p / psp
O iginal
a icle
E ficiency
and
equi y
consequences
o
decen aliza ion
in
heal h:
an
economic
pe spec i e
Joana
Al esa,∗,
Susana
Pe al ab,
Julian
Pe elmana
aEscola
Nacional
de
Saúde
Pública,
Uni e sidade
No a
de
Lisboa,
Lisboa,
Po ugal
bNo a
School
o
Business
and
Economics,
Uni e sidade
No a
de
Lisboa,
Lisboa,
Po ugal
a
i
c
l
e
i
n
o
A icle
his o y:
Recei ed
17
Sep embe
2012
Accep ed
21
Janua y
2013
A ailable
online
10
July
2013
Keywo ds:
Decen aliza ion
E ficiency
Equi y
Heal h
economics
Incen i es
a
b
s
a
c
O e
he
ecen
yea s,
decen aliza ion
has
been
adop ed
in
many
heal h
sys ems.
The
ques ion
howe e
emains
o
whe he
local
ac o s
do
be e
han
he
cen al
go e nmen .
We
summa ize
he
main
insigh s
om
economic
heo y
on
he
impac
o
decen aliza-
ion
and
i s
empi ical
alida ion.
Theo y
sugges s
ha
he
decision
o
decen alize
esul s
om
a
ade-o
be ween
i s
ad an ages
(like
i s
capaci y
o
ca e
o
local
as es)
and
cos s
(like
in e - egional
spillo e s).
Empi ical
con ibu ions
poin
ha
decen aliza ion
esul s
in
be e
heal h
ou comes
and
highe
expendi u es,
esul ing
in
ambiguous
consequences
on
e ficiency;
equi y
consequences
a e
con o e sial
and
add ess
he
ele ance
o
edis ibu ion
mechanisms.
©
2012
Escola
Nacional
de
Saúde
Pública.
Published
by
Else ie
España,
S.L.
All
igh s
ese ed.
Consequências
em
eficiência
e
equidade
da
descen alizac¸ão
na
saúde:
uma
pe spe i a
económica
Pala as-cha e:
Descen alizac¸ão
Eficiência
Equidade
Economia
da
saúde
Incen i os
e
s
u
m
o
Apesa
de
mui os
sis emas
de
saúde
e em
op ado
ecen emen e
pela
descen alizac¸ão,
fica
po
escla ece
se
os
go e nos
locais
êm
um
melho
desempenho
do
que
os
cen-
ais.
Es e
a igo
suma iza
os
p incipais
esul ados
da
eo ia
económica
sob e
o
impac o
da
descen alizac¸ão
e
a
sua
alidade
empí ica.
A
decisão
de
descen aliza
esul a
duma
a bi-
agem
en e
as
an agens,
como
a
adap ac¸ão
às
p e e ências,
e
os
incon enien es,
como
as
ex e nalidades
in e - egionais.
Os
es udos
empí icos
suge em
que
a
descen alizac¸ão
pe -
mi e
ganhos
em
saúde
mas
ambém
despesas
maio es,
com
consequências
ambíguas
em
e mos
de
eficiência,
e
que
as
consequências
pa a
a
equidade,
sendo
con o e sas,
indicam
a
ele ância
dos
mecanismos
de
edis ibuic¸ão.
©
2012
Escola
Nacional
de
Saúde
Pública.
Publicado
po
Else ie
España,
S.L.
Todos
os
di ei os
ese ados.
∗Co esponding
au ho .
E-mail
add ess:
joana.al [email protected]
(J.
Al es).
0870-9025/$
–
see
on
ma e
©
2012
Escola
Nacional
de
Saúde
Pública.
Published
by
Else ie
España,
S.L.
All
igh s
ese ed.
h p://dx.doi.o g/10.1016/j. psp.2013.01.002
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
75
In oduc ion
O e
he
ecen
yea s,
decen aliza ion
has
been
one
o
he
main
e o ms
adop ed
in
many
heal h
sys ems.
S ong
deci-
sion
powe
o
sub-cen al
go e nmen s
al eady
exis ed
in
Scandina ian
coun ies
and
in
ede al
s a es
like
Canada,
Swi ze land
and
Aus alia1;
he
mos
ecen
de olu ion
p o-
cesses
ha e
been
obse ed
in
NHS
sys ems,
namely
in
Spain
and
I aly
( owa d
egions),
and
he
UK
( owa d
Wales,
Sco -
land,
No he n
I eland).
In
a
ew
wo ds,
mo e
delega ion
o
local
au ho i ies
was
expec ed
o
imp o e
se ices
h ough
a
combina ion
o
be e
knowledge
o
local
needs,
p e e -
ences
and
p o ide s’
cha ac e is ics,
highe
accoun abili y
o
policy-make s
and
e ficiency-enhancing
compe i ion
among
ju isdic ions.
These
expec ed
benefi s
a e
howe e
a
om
ob ious
and
he
majo
ques ion
o
policy-make s
and
esea che s
emains,
ha
is,
whe he
local
ac o s
eally
do
be e
han
he
cen al
go e nmen .
Pu
di e en ly,
he
issue
is
whe he
decen aliza ion
allows
o
he
p o ision
o
be e
and
equi-
able
heal h
se ices
o
all
ci izens
a
an
accep able
cos .
Oa es2was
he
fi s
one
o
sugges
ha
he
decision
o
decen-
alize
esul s
om
a
ade-o
be ween
i s
ad an ages
and
cos s.
Whe he
he
benefi s
ac ually
ou weigh
he
cos s
–
he
e ficiency
issue
–
is
ul ima ely
an
empi ical
ques ion.
The
li e a u e
has
mainly
ocused
on
o e all
e ficiency
by
es i-
ma ing
he
impac
o
decen aliza ion
on
economic
g ow h.3,4
Ano he
al e na i e
is
o
measu e
he
impac
o
decen aliza-
ion
on
sec o al
policies.5Decen aliza ion
o
heal h
policies
is
ela i ely
ecen ;
he
li e a u e
is
s ill
sca ce
and
o en
p o-
ides
con adic o y
esul s.
We
summa ize
he
main
insigh s
om
economic
heo y
on
he
impac
o
decen aliza ion
o
heal h
policies,
and
we
e iew
empi ical
s udies
ha
ha e
es ed
some
o
hese
heo e ical
assump ions.
Ins i u ional
con ex
Decen aliza ion
in
heal h
ca e
can
adop
many
di e en
o ms.
Following
he
ypology
p oposed
by
V angbaek,6ou
main
in e es
he e
is
in
de olu ion
o
poli ical
decen aliza ion,
h ough
which
powe
is
“decen alized
o
lowe -le el
poli ical
au ho i ies
such
as
egions
o
municipali ies”.
O he
o ms
o
decen aliza ion
a e
also
possible
which
a e
beyond
he
scope
o
he
p esen
pape .
De-concen a ion
and
bu eauc a-
iza ion
in ol e
ans e s
be ween
adminis a i e
le els
and
om
poli ical
o
adminis a i e
le el,
espec i ely
( hink
o
example
in
he
Po uguese
con ex ,
o
ans e
o
compe ences
o
Regional
Heal h
Au ho i ies).
Decen aliza ion
also
e e s
o
delega ion,
which
in ol es
ans e ing
powe
o
mo e
o
less
au onomous
public
o ganiza ion
managemen
(like
pub-
lic
en e p ises,
o
example
he
“hospi als
S.A”
in
Po ugal,
Founda ion
T us s
in
he
UK
o
public
insu ance
companies
in
Bisma ckian- ype
heal h
sys ems).
P i a iza ion
i sel
can
be
conside ed
as
a
o m
o
decen aliza ion.
Ou
ocus
he e
is
hus
on
he
ans e
o
poli ical
au ho i y
in
he
heal h
a ea
om
highe
o
lowe
le els
o
go e nmen
o
om
na ional
o
sub-na ional
le els.7Regional,
p o incial
o
municipal
elec ed
go e nmen s
may
hus
be
esponsible
o
planning,
o ganizing,
deli e ing
and
financing
heal h
se ices.
Mul iple
a angemen s
a e
howe e
possible
and
de olu ion
has
aken
many
di e en
o ms
ac oss
OECD
coun ies
( o
a
comple e
mapping
o
decen aliza ion
expe iences
in
Eu ope,
see
Bankauskai e
e
al.8).
The
size
o
decen alized
en i ies
is
highly
a iable,
om
small
coun ies
in
Sweden
(a e age
popula ion
31,000)
o
la ge
au onomous
egions
in
Spain
(a e -
age
popula ion
2,444,000).
The
ex en
o
compe ences
also
a ies.
The
cen al
s a e
may
keep
esponsibili ies
in
se e al
domains,
o
ins ance
he
Canadian
Fede al
s a e
defines
poli-
cies
ega ding
heal h
p e en ion
and
p omo ion,
he
Swiss
Con ede a ion
defines
he
basic
heal h
benefi
package,
and
gene ally
all
cen al
s a es
impose
a
se ies
o
mo e
o
less
s ingen
egula ions
on
quali y,
supply,
co e age,
p icing
ules
o
budge
alloca ion.
A
key
issue
which
di e en ia es
decen aliza ion
expe iences
ela es
o
he
unding
o
heal h
expendi u es
and
financial
au onomy.
Decen alized
go e n-
men s
may
ha e
he
powe
o
aise
axes,
o
ecei e
ans e s
depending
on
hei
con ibu ion
o
fiscal
e enues.
They
may
hen
be
ee
o
se
he
budge
alloca ed
o
heal h
and
i s
dis ibu ion
among
heal h
sec o s.
By
con as ,
sub-cen al
go e nmen
le els
may
be
financed
by
ans e s
based
on
isk-equaliza ion
schemes
and
benefi
om
low
au onomy
in
defining
ax
a es.
In
he
la e
case,
decen alized
go -
e nmen s
may
hus
ha e
he
poli ical
powe
o
decide
abou
alloca ion
o
esou ces
bu
do
no
con ol
he
amoun
o
a ailable
esou ces
o
heal h.
Hence
he
expendi u e
side
is
decen alized,
bu
e enue
is
no .
Fig.
1,
ex ac ed
om
Jouma d
e
al.9,
nicely
desc ibes
he
deg ee
o
decen aliza ion
in
heal h
in
OECD
coun ies
(whe e
a
0
sco e
implies
ha
cen al
go e nmen
akes
mos
o
key
decisions
while
a
6
sco e
implies
esidual
compe ences
o
he
cen al
go e nmen ).
Coun ies
ha e
he e ogeneous
sub-cen al
go e nmen s,
wi h
di e en
size,
au onomy
and
esponsibili ies
alloca ed.
The
c i e ia
o
define
his
sco e
ha e
been
es ablished
by
Pa is
e
al.10,
using
a
su ey.
They
include
in
pa icula
he
sub-cen al
go e nmen s’
au ho i y
on
se -
ing
ax
bases
and
a es,
he
budge
alloca ion
o
heal h
and
i s
dis ibu ion
be ween
heal h
sec o s,
he
financing
o
di e en
heal h
se ices
and
p ac i ione s
and
he
se ing
o
public
heal h
objec i es.
Spain,
Canada,
Finland,
Sweden
and
Swi ze land
ha e
highe
au onomy
o
sub-cen al
go e n-
men s
while
Po ugal
is
among
he
coun ies
wi h
he
lowes
heal h
decen aliza ion.
No e
ha
he
decen aliza ion
ques ion,
he
way
we
pose
i
(de olu ion
o
poli ical
decen aliza ion),
is
mo e
common
in
NHS- ype
heal h
sys ems.
These
sys ems
ha e
been
usually
cha ac e ized
by
highly
cen alized
decision-making,
wi h
one
single
insu e /paye
and
p o ide
( he
cen al
s a e),
and
li -
le
au onomy
gi en
o
p o ide s.
By
con as ,
heal h
sys ems
based
on
social
insu ance
schemes
ha e
been
usually
cha -
ac e ized,
o
a ious
ex en s,
by
mul iple
insu ance
schemes
and
some
deg ee
o
publicly
subsidized
p i a e
p o ision.
Hence,
ans e ing
powe
o
local
ins i u ions
has
ce ainly
esponded
o
a
demand
o
mo e
au onomy
in
mo e
igid
NHS-
ype
heal h
sys ems.
Spain,
Canada
and
I aly
a e
pa icula ly
ele an
cases
o
unde s and
he
empi ical
li e a u e
because
he
consequences
o
decen aliza ion
in
heal h
ha e
been
mo e
ex ensi ely
analyzed.
The
mos
decen alized
coun y
acco ding
o
76
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
6
5
4
3
2
1
0
FRA
GRC
IRL
KOR
LUX
NLD
TUR
ISL
BEL
SVK
PRT
HUN
CZE
DEU
POL
MEX
JPN
ITA
DNK
NZL
AUS
GBR
NOR
AUT
SWE
CHE
FIN
CAN
ESP
Fig.
1
–
Decen aliza ion
sco e
o
OECD
coun ies
(ex ac ed
om
Jouma d
e
al.9).
Jouma d
e
al.9is
Spain.
The
cen al
go e nmen
p o ides
he
gene al
amewo k
and
coo dina es
he
heal h
sys em
while
Au onomous
Communi ies
(AC)
p o ide
heal h
ca e
se ices
and
a e
esponsible
o
heal h
planning,
o ganiza ion
and
managemen .11 Funds
a e
cen ally
collec ed
and
alloca ed
o
egions
by
means
o
a
block
cen al
g an
ollowing
an
unadjus ed
capi a ion
o mula.11 Only
Na a a
and
he
Basque
Coun y
benefi
om
fiscal
au ho i y.
By
con as ,
he
Canadian
Con ede a ion,
composed
o
10
p o inces
and
3
e i o ies,
has
been
decen alized
since
he
19 h
cen u y.12 Heal h
unding,
adminis a ion
and
deli e ing
a e
compe ences
o
p o inces,
which
also
define
physician
financing
ules
and
hospi al
global
budge s.
Howe e
heal h
p omo ion,
p e en ion
and
p o ision
o
specific
g oups
a e
Fede al
compe encies.
P o inces
a e
esponsible
o
und-
ing
heal h
ca e
expendi u es,
based
on
p o incial
axes
and
ans e s
om
he
cen al
go e nmen .
T ans e s
depend
hemsel es
on
axes
collec ed
a
p o incial
le el;
an
equal-
iza ion
p og am
was
howe e
implemen ed
o
a oid
s ong
di e ences
in
ans e s
ela ed
o
disc epancies
in
p o inces’
e enue-gene a ing
po en ial.13
Finally,
he
20
I alian
egions
also
con ol
heal h
ca e
p o ision,
al hough
gene al
objec i es
(like
he
de ailed
lis
o
se ices
o
be
p o ided)
and
main
p inciples
o
he
heal h
sys em
a e
defined
a
he
cen al
le el.
Regions
can
de elop
egional
heal h
plans,
alloca e
esou ces
and
collec
e enues
eely,
like
se ing
use
cha ges
o
d ugs
p esc ip ions
o
eimbu se
a es
o
d ugs
and
se ices
no
co e ed
a
na ional
le el.
Regional
au ho i ies
also
enjoy
financial
au onomy
in
axa ion;
in
p ac ice,
egional
heal h
expendi u es
a e
financed
a
36.7%
by
egional
axes.14 Addi ionally,
egional
heal h
expendi u es
a e
financed
by
ans e s
om
he
cen al
s a e
(57.3%
o
expendi u es),
essen ially
om
a
so-called
Na ional
equalizing
und.
This
und
is
based
on
he
idea
ha
all
egions
mus
achie e
a
minimum
le el
o
expendi u es,
so
ha
ans e s
op
up
egions’
own
e enue
in
case
i
is
unable
o
und
his
minimum
le el.
Theo e ical
backg ound
The
deba e
on
he
ela i e
me i s
o
policy
decen aliza ion
da es
back
o
Tiebou ,15 who
pu
o wa d
a
e y
op imis ic
a gumen
o
he
op imali y
o
local
public
good
p o i-
sion,
based
on
an
analogy
be ween
compe i i e
ma ke s
and
compe ing
ju isdic ions.
Oa es,2in
con as ,
sugges s
ha
decen aliza ion
esul s
om
a
ade-o
be ween
i s
ad an ages
and
cos s.
Acco ding
o
Oa es,
he
ad an age
o
decen aliza ion
lies
in
i s
capaci y
o
ca e
o
local
as es,
be
i
because
o
be e
in o ma ion
o
simply
because
he
cen-
al
go e nmen
canno
di e en ia e
public
good
p o ision.
The
cos
o
decen aliza ion
s ems
essen ially
om
he
p es-
ence
o
in e - egional
spillo e s.
Fo
ins ance,
be e
heal h
p e en ion
in
one
municipali y
benefi s
i s
neighbo s.
How-
e e ,
i
heal h
p e en ion
is
decen alized,
each
municipali y
ails
o
ake
in o
accoun
he
benefi
o
i s
in es men
in
o he
municipali ies,
and
is
he e o e
likely
o
in es
sub-op imally,
leading
o
unde -p o ision
and/o
low
quali y
o
heal h
se -
ices.
While
Oa es’
a gumen
ha
a
cen al
go e nmen
is
unable
o
di e en ia e
he
p o ision
o
local
public
goods
acco ding
o
local
p e e ences
has
been
challenged
since,
his
main
in ui ion
ha
he
decision
o
decen alize
es s
on
a
un-
damen al
ade-o
be ween
cos s
and
ad an ages
emains.
One
o
he
main
cos s
o
decen aliza ion
which
has
been
analyzed
in
he
li e a u e
is
he
ailu e
o
exploi
economies
o
scale.16 Fo
ins ance,
collec i e
pu chasing
o
esou ces
(d ugs,
medical
de ices,
equipmen s,
e c.)
o
collec i e
admin-
is a ion
o
heal h
se ices
may
display
dec easing
a e age
cos s
and
be
hus
mo e
e ficien ly
p o ided
a
a
cen alized
le el.
The
in e - egional
spillo e s
a
he
hea
o
Oa es’s
a gu-
men ,
e e ed
he e-abo e,
a e
o
he
ho izon al
ype
( ha
is,
be ween
sub-cen al
go e nmen s
o
he
same
le el).
Ano he
so
o
spillo e s
which
may
c ea e
ine ficiencies
unde
decen-
aliza ion
is
he
e ical
ones
be ween
local
and
cen al
go e nmen s.
These
occu
i
one
go e nmen
le el’s
policy
decisions
ha e
an
impac
in
he
policy
ou comes
o
he
o he
le el.
These
phenomena
a e
highly
plausible
in
he
heal h
ca e
sec o .
I ,
o
example,
sub-cen al
go e nmen s
a e
espon-
sible
o
some
public
heal h
p e en ion
p og ams
and
he
cen al
s a e
deli e s
heal h
ca e,
local
policies
likely
a ec
se ices
use
and
cos s
a
he
cen al
le el.
Unless
sub-cen al
go e nmen s
a e
ewa ded
o
hei
ac ions,
hey
ha e
li le
incen i es
o
p o ide
an
op imal
alue
o
p e en ion,
whose
benefi s
a e
ully
enjoyed
by
he
cen al
s a e.
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
77
Recen ly,
he
li e a u e
has
pu
o wa d
poli ical
economy
a gumen s
in
a o
o
decen aliza ion.
Fo
ins ance,
i
may
be
ha
he
o e s,
who
a e
impe ec ly
in o med
abou
he
economy,
use
he
policies
pu
in
place
by
he
neighbo ing
poli icians
o
impose
discipline
on
hei
own
local
ep e-
sen a i e.
In
such
a
wo ld
o
oppo unis ic
policy
make s,
decen aliza ion
may
be
a
s ong
mechanism
o
p e en
co -
up ion
o
p omo e
p o ision
o
public
se ices.17 Ano he
a gumen
is
ha
decen aliza ion
o ces
local
ju isdic ions
o
compe e
o
mobile
esou ces
( o
ins ance,
labo
o
capi al)
and
pu s
a
downwa d
p essu e
on
axa ion,
hus
p omp ing
a
mo e
e ficien
use
o
sca ce
esou ces
(see
Wilson18 o
a
su -
ey
o
he
ax
compe i ion
li e a u e).
Besley
and
Sma 19 show
ha
his
need
no
be
he
case,
since
highe
discipline
o
local
poli icians
comes
a
he
cos
o
a
wo se
selec ion.
Indeed,
i
is
only
when
an
oppo unis ic
poli ician
misbeha es
ha
he
o e s
a e
gi en
he
chance
o
ous
he
om
powe
and
eplace
he
wi h
a
po en ially
bene olen
poli ician.
Poli ical
accoun -
abili y
esul s
om
a
ade-o
be ween
hese
wo
mechanisms
(discipline
and
selec ion),
and
u he
decen aliza ion
has
opposi e
e ec s
in
bo h.
Wha
can
be
said
abou
he
likelihood
o
local
go e n-
men s
being
cap u ed
by
special
in e es s?
Ba dhan
and
Mookhe jee20 show
ha
he e
is
no
heo e ical
eason
o
sup-
pose
ha
local
go e nmen s
a e
mo e
p one
o
his
han
cen al
ones.
They
pu
o wa d
a
heo e ical
model
o
p oba-
bilis ic
o ing
whe e
cap u e
by
special
in e es s
esul s
om
a
combina ion
o
o e
awa eness,
in e es
g oup
cohesi eness,
elec o al
unce ain y
and
compe i ion,
dis ic
he e ogenei y
and
he
elec o al
sys em.
Thei
main
conclusion
is
ha
he
ex en
o
cap u e
a
he
local
le el
is
mos
likely
con ex -
specific
and
needs
o
be
assessed
empi ically.
A
ew
au ho s
ha e
looked
a
he
applicabili y
o
hese
a gu-
men s
o
he
specific
case
o
heal h
decen aliza ion.
Le aggi
and
Smi h21 claim
ha
sub-cen al
go e nmen s
a e
be e
in o med
abou
he
cons ain s
o
local
supply
and
abou
a i-
a ions
in
demand.
Va ia ions
in
demand
a e
ob iously
ela ed
fi s
o
di e ences
in
local
needs,
o
ins ance
egions
expe i-
ence
disc epancies
in
he
p e alence
o
diseases,
in
beha io al
and
non-beha io al
isk
ac o s
(including
in
pa icula
age-
ing
o
social
de e minan s),
pe haps
also
in
he
e ec i eness
o
specific
in e en ions.
Hence
p io i ies
a e
likely
o
di e
acco ding
o
bu den
o
disease
o
cos -e ec i eness
c i e ia.
Also
local
p e e ences
shape
a ia ions
in
demand,
leading
o
di e en
p io i ies
and
esou ce
alloca ion
c i e ia.
We
may
hink
ha
some
egions
ha e
a
highe
conce n
o
inequali ies
in
heal h
and
social
de e minan s,
while
o he s
pu
a
g ea e
emphasis
on
p o ide
choice
and
esponsi eness
o
pa ien s’
p e e ences.
These
di e ences
in
needs
and
p e e ences
a e
e y
clea
i
we
compa e
Eu opean
coun ies.
Disc epancies
in
local
supply
a e
ela ed
o
he
a ailabili y
and
densi y
o
physicians,
nu ses,
equipmen s,
bu
also
o
di e ences
in
local
p ices
and
p ac ices.
Cen al
go e nmen s
ha e
ied
o
accoun
o
hese
di e ences
in
financing
sub-cen al
go e n-
men s,
h ough
mo e
o
less
sophis ica ed
isk-adjus men
schemes,
bu
may
no
be
able
o
add ess
hese
issues
o
locally
adjus ed
planning
o
o ganiza ion.
Sub-cen al
go e nmen s
may
be
be e
equipped
o
espond
o
local
p e e ences
and
p io i ies,
o
coo dina e
p o ide s’
ac ions
and
o
iden i y
he
sou ces
o
ine ficiency.
Hence,
decen aliza ion
is
expec ed
o
enhance
quali y
and
esponsi eness
o
ca e
while
educing
cos s;
his
heo e ical
iew
is
a
he
o igin
o
mos
decen al-
iza ion
p ocesses.
As
ega ds
he
po en ial
incen i es
ha
a o
e ficiency
and/o
quali y
in
heal h
ca e
p o ision,
bo h
he
ax
compe-
i ion
and
he
ya ds ick
compe i ion
mechanisms
a e
likely
o
a ise
in
heal h
policies,
hus
leading
sub-cen al
go e n-
men s
o
compe e
wi h
each
o he
o
p o ide
high-quali y
se ices
a
low
use
cha ges
o
financed
h ough
lowe
axes.
Fi s ,
be e
and
mo e
e ficien
se ices
con ibu e
o
a ac
mobile
ci izens,
hence
inc ease
sub-cen al
go e nmen s’
economic
ac i i y
and
fiscal
e enues.
Second,
compe i ion
occu s
h ough
ci izens
benchma king
hei
demands
on
he
basis
o
neighbo
sub-cen al
go e nmen s’
pe o mance.22 I
sub-cen al
go e nmen s
a e
poli ically
accoun able,
a
local
poli ician
p o iding
poo e
se ices
han
his
neighbo s’
coun-
e pa s
would
likely
ail
e-elec ion.
In
a
ew
wo ds,
e ficien
p o ision
con ibu es
o
a ac i eness
and
ci izens’
sa is ac-
ion.
I
is
a
well-known
ac
ha
decen aliza ion
may
ham-
pe
edis ibu ion
(see,
C eme
e
al.23 o
a
su ey
o
he
li e a u e),
when
axes
a e
locally
collec ed.
A eas
wi h
high
heal h
needs
a e
in
gene al
he
poo e
ones
( he e
is
la ge
e idence
o
he
ela ionship
be ween
po e y
and
poo
heal h
s a us,
see
Ma mo
Re iew24);
unde
fiscal
decen al-
iza ion,
unde p i ileged
a eas
cap u e
lowe
esou ces
om
axa ion
and
benefi
om
a
lowe
capaci y
o
in es
in
high-
quali y
se ices
and
e ficien
p o ision.
Decen aliza ion
hen
p oduces
di e ences
in
he
quali y
and
a ailabili y
o
se -
ices,
cha ges
and
ou comes.
Inc easing
inequi y
is
one
o
he
majo
h ea s
o
decen aliza ion,
e en
i
good
isk-sha ing
ag eemen s
po en ially
mi iga e
his
e ec .
In
o he
wo ds,
he
iabili y
o
decen aliza ion
depends
on
he
exis ence
o
solida i y
mechanisms
be ween
decen alized
au ho i-
ies.
The
e e se
is
also
possible,
i.e.,
sub-cen al
go e nmen s
only
compe ing
on
high-quali y
se ices,
leading
o
o e -
p o ision
o
se ices
and
high
heal h
expendi u es
( ace
o
he
op).
This
si ua ion
is
mo e
likely
o
occu
when
fiscal
decen aliza ion
is
low
and
sub-cen al
go e nmen s
ace
so
budge
cons ain s.
This
may
he
case,
o
example,
i
he
cen al
go e nmen
sys ema ically
bails
ou
sub-cen al
go -
e nmen s
which
a e
excessi ely
indeb ed
o
i
ans e s
o
sub-cen al
go e nmen s
a e
based
on
pas
expendi u es.
This
b ings
us
o
he
undamen al
ques ion
o
he
und-
ing
o
local
public
goods.
Indeed,
a ious
a angemen s
a e
possible
ega ding
he
deg ee
o
decen aliza ion
o
bo h
he
expendi u e
and
he
e enue
unc ions.25 A
conside able
sha e
o
local
go e nmen
unding
comes
om
cen al
go -
e nmen
ans e s.
These
ans e
schemes
mus
be
ca e ully
designed
and
es
on
clea
ules
(depend
on
a iables
which
a e
no
easy
o
local
go e nmen s
o
manipula e)
so
as
o
a oid
mo al
haza d
leading
o
excess
expendi u es.
Indeed,
as
he
cen al
go e nmen
is
unable
o
dis inguish
he
sou ces
o
high
spending
be ween
highe
needs
and
ine ficiencies,
sub-
cen al
go e nmen s
ha e
an
incen i e
o
hide
hei
ue
needs
o
ob ain
highe
financing
om
he
cen al
au ho i y.
To
sum
up,
he
heo e ical
li e a u e
does
no
p o ide
a
defini i e
answe
abou
he
impac
o
decen aliza ion
on
heal h
policies.
The
high
conce n
in
many
coun ies
o
equi y
78
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
Table
1
–
Li e a u e
e iew:
impac
o
decen aliza ion.
1s
au ho Coun y Objec i e Pe iod
o
he
da a
Decen aliza ion
indica o Main
esul s
Zhong (2010) Canada
Impac
o
he
in oduc ion
o
a
highe
deg ee
o
decen aliza ion
on
equi y
in
heal h
ca e
u iliza ion
1994/95
1998/99
2000/01
In oduc ion
o
Heal h
and
Social
T ans e
in
1996/97,
i.e.,
la ge
block
unding
o
p o inces
Mos
inequi y
in
heal h
ca e
use
is
explained
by
a ia ions
wi hin
p o inces.
An
inc ease
in
decen aliza ion
lead
o
lowe
o e all
and
wi hin-p o ince
inequi y
in
GP
and
hospi al
se ices
and
lowe
be ween-p o ince
a ia ion
in
GP,
hospi al
and
specialis
se ices
Sche fle (2006
S.) USA/Cali o nia
Impac
o
highe
decen aliza ion
o
men al
heal h
ca e
se ices
on
uninsu ed
heal h
spending
1989–2000
Decen aliza ion
o
heal h,
men al
heal h,
and
social
se ices
(g ea e
financial
flexibili y
and
esponsibili y)
om
he
Cali o nia
s a e
o
coun ies
(p og am
ealignmen
in oduced
he
abili y
o
ans e
unds
be ween
p og ams
–
men al
heal h,
social
se ices
and
gene al
heal h)
Coun y
spending
on
men al
heal h
se ices
dec eased
a e
p og am
ealignmen .
Coun ies
wi h
abo e
median
heal h
e enues
a e
less
likely
o
make
a
ans e
o
ano he
p og am.
Wi h
ealignmen ,
coun ies
showed
he
same
le el
o
commi men
and
suppo
wi h
men al
heal h
se ices,
since
di e ed
unds
wen
mainly
o
men al
heal h
se ices
wi hin
he
ealignmen
p og ams
package
Jiménez-Rubio (2011A) 20
OECD
coun ies
Impac
o
decen aliza ion
on
heal h
s a us
(in an
mo ali y)
1970–2001
Decen aliza ion
measu ed
as:
(i)
Sha e
o
au onomous
ax
e enue
o
SCG
o e
he
CG
ax
e enue.
(ii)
Sha e
o
o al
(au onomous
o
no )
SCG
ax
o e
he
o al
e enue
Nega i e
and
significan
ela ionship
be ween
fiscal
decen aliza ion
and
in an
mo ali y,
i
a
subs an ial
deg ee
o
au onomy
in
he
sou ces
o
e enue
is
de ol ed
o
sub-cen al
go e nmen s
P ie o (2012) Spain
Es ima e
de e minan s
o
egional
heal h
ca e
expendi u e,
in
pa icula
he
deg ee
o
egional
au onomy
1992–2005
Regions
clus e ed
acco ding
o
he
deg ee
o
au onomy:
(i)
Expendi u e
and
high
fiscal
au onomy.
(ii)
Expendi u e
and
low
fiscal
au onomy.
(iii)
No
au onomy
GDP/capi a
has
mo e
influence
on
HCE
in
egions
wi h
highe
ax
au onomy,
maybe
due
o
equaliza ion
e o s
o
Spanish
NHS.
Decen aliza ion
wi hou
inequali y
is
possible
wi h
fiscal
equaliza ion.
I
high
di e ences
in
weal h
and
weak
equaliza ion
among
egions,
decen aliza ion
leads
o
inequali y
Bo dignon (2009
B.) I aly
Impac
o
egions’
bailou
expec a ions
on
unding
by
CG;
impac
o
egions’
bailou
expec a ions
on
HCE
(mid-90s),
de e minan s
o
egions’
bailou
expec a ions
1990–1999
Regions’
bailou
expec a ions
based
on
ex e nal
cons ain s
in
he
mid-90s:
coun ies’
e alua ion
o
en e ing
he
EURO
(1997),
index
o
public
budge
igh ness,
Maas ich
ules
(1994)
HCE
is
influenced
by
egional
financing,
which
is
in
u n
influenced
by
bailou
expec a ions
om
he
CG.
Lowe
expec a ions
o
bailing
ou
gene a e
a
s onge
link
be ween
ex-an e
financing
and
egional
HCE.
Riche
and
mo e
au onomous
egions
ha e
lowe
expec a ions
o
CG
in e en ion,
hence
hey
a e
mo e
financially
esponsible.
The
CG
is
also
mo e
p one
o
cu
financing
o
egions
unde
lowe
bailou
expec a ions
because
he
commi men
o
cu
HCE
will
be
aken
se iously
F ed iksson (2008
F.) Sweden Es ima e
c oss-coun ies
di e ences
in
suppo ing
he
“Pa ien
Choice
Recommenda ion”,
i.e.,
coun ies’
willingness
o
a o
o
no
he
pa ien ’s
choice
o
specialis
2001–2005
Recommenda ion
by
he
CG
can
be
mo e
o
less
suppo ed
by
coun ies,
which
ha e
la ge
compe ences
in
he
heal h
sec o
High
a ia ion
on
suppo
o
he
Pa ien
Choice
Recommenda ion
among
coun y
councils.
Abili y
o
choose
heal hca e
p o ide
a ies
acco ding
o
coun y,
c ea ing
inequali ies
in
access
o
ca e
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
79
Table
1
(Con inued)
1s
au ho
Coun y
Objec i e
Pe iod
o
he
da a
Decen aliza ion
indica o
Main
esul s
López-Casasno as (2007
L.-C.) 110
egions
in
8
OECD
coun ies
Measu e
he
de e minan s
o
HCE
accoun ing
o
egional
e ec s
1997
No
indica o
o
decen aliza ion
bu
accoun
o
possible
egional
di e ences
in
HCE
An
inc ease
on
he
pe cen age
o
people
wi h
mo e
han
65
yea s
and
in
he
pe cen age
o
public
heal h
expendi u es
inc eases
HCE.
Be ween
coun ies
a iabili y
o
expendi u es
is
smalle
han
wi hin
coun y
a ia ions.
Income
elas ici y
o
HCE
is
low
and
inc eases
wi h
a ia ion
in
HCE
Cos a-Fon (2007
C.-F.) Spain
Measu e
spa ial
spillo e s
in
HCE
(in e -ju isdic ion
in e ac ions
in
defining
HCE)
2003
Compa e
egions
wi h
(i)
fiscal
and
poli ical
au onomy
(Na a a
and
Basque
coun y);
(ii)
poli ical
au onomy
(Ca alonia,
Galicia,
Andalucía,
Cana ias
and
Valencia);
(iii)
no
au onomy
E idence
o
spa ial
co ela ion.
Highe
HCE
when
egions
ha e
highe
poli ical
decen aliza ion
and
e en
highe
i
hey
ha e
also
fiscal
decen aliza ion.
Highe
HCE
i
egion
and
CG
a e
uled
by
he
same
pa y
Cos a-i-Fon (2005
C.-i.-F.) Spain
Measu e
income- ela ed
in a- egional
inequali ies
in
sel - epo ed
heal h,
and
i s
ela ion
wi h
he
deg ee
o
decen aliza ion
1997
Compa e
egions
wi h
(i)
fiscal
and
poli ical
au onomy
(Na a a
and
Basque
coun y);
(ii)
poli ical
au onomy
(Ca alonia,
Galicia,
Andalucía,
Cana ias
and
Valencia);
(iii)
no
au onomy
O e all
inequali y
is
low.
De olu ion
does
no
a o
in e - egion
inequali y.
By
con as
de olu ion
seems
o
a o
p o-equi y
policies:
inequali y
is
he
highes
in
egions
wi hou
au onomy
(INSALUD)
C i elli (2006
C.) Swi ze land
Measu e
de e minan s
o
public
HCE
accoun ing
o
egional
(can ons)
e ec s
1996–2001
No
indica o
o
decen aliza ion
bu
accoun
o
possible
egional
di e ences
in
HCE
Di e ences
in
HCE
ela ed
o
usual
ac o s
bu
in
pa icula
o
physician
densi y.
S ong
di e ences
in
HCE
ac oss
can ons
explained
in
pa
by
di e ences
in
supply
Cos a-Fon (2009
C.-F.) Spain
Impac
o
decen aliza ion
on
HCE,
in e dependency
be ween
neighbo
municipali ies
and
measu e
income
e ec
a
egional
le el
1995–2002
Indica o s
o
decen aliza ion
deg ee
and
i s
e olu ion
ac oss
ime,
owa d
cu en
si ua ion:
(i)
fiscal
and
poli ical
au onomy
(Na a a
and
Basque
coun y);
(ii)
poli ical
au onomy
(Ca alonia,
Galicia,
Andalucía,
Cana ias,
Valencia);
(iii)
no
au onomy
Poli ical
decen aliza ion
inc eases
HCE
in
he
sho
e m
bu
p oduces
cos
sa ings
in
he
long
un
(sunk
cos
o
decen aliza ion
and
expe ience
e ec s).
Highe
expendi u es
a e
ela ed
o
highe
fiscal
au onomy.
Exis ence
o
spa ial
co ela ion.
Low
impac
o
income
on
HCE,
hence
decen aliza ion
is
no
associa ed
o
s ong
egional
inequali ies
Can a e o (2005
C.) Spain
Measu e
de e minan s
o
public
HCE
accoun ing
o
egional
e ec s
1993–1999
Compa e
egions
wi h
(i)
fiscal
and
poli ical
au onomy
(Na a a
and
Basque
coun y);
(ii)
poli ical
au onomy
(Ca alonia,
Galicia,
Andalucía,
Cana ias,
Valencia);
(iii)
no
au onomy
Regional
income
and
supply
ha e
low
impac
on
HCE,
while
ageing
is
mo e
ele an .
Hence
lowe
income- ela ed
inequali y
is
expec ed
Jiménez-Rubio (2011) Canada
Impac
o
fiscal
decen aliza ion
on
heal h
ou comes
(in an
mo ali y)
1979–1995
Fiscal
decen aliza ion
is
measu ed
as
he
p opo ion
o
sub-na ional
heal h
spending
o e
o al
heal h
expendi u e
Significan
and
subs an ially
lowe
in an
mo ali y
ela ed
o
highe
fiscal
decen aliza ion
(1%
inc ease
in
fiscal
decen aliza ion
dec eases
in an
mo ali y
by
10%)
Giannoni (2002) I aly
Measu e
de e minan s
o
HCE
accoun ing
o
egional
e ec s,
and
impac
o
1992
cos -con ainmen
e o ms
1980–1995
No
indica o
o
decen aliza ion
bu
accoun
o
possible
egional
di e ences
in
HCE
S ong
impac
o
egional
income
on
HCE
and
ele an
a ea
clus e
di e ences.
Cos -con ainmen
e o ms
did
no
al e
mos
egional
di e ences
in
HCE
and
wo sened
he
si ua ion
o
some
egions
below
he
a e age
80
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
Table
1
(Con inued)
1s
au ho
Coun y
Objec i e
Pe iod
o
he
da a
Decen aliza ion
indica o
Main
esul s
Can a e o (2008
C.) Spain
Impac
o
fiscal
decen aliza ion
on
heal h
ou comes
(in an
mo ali y
and
li e
expec ancy)
1992–2003
Fiscal
decen aliza ion
is
measu ed
as
he
p opo ion
o
sub-na ional
heal h
spending
o e
o al
heal h
expendi u e
Decen aliza ion
significan ly
educes
in an
mo ali y
and
inc eases
li e
expec ancy
Di
Ma eo (1998) Canada
Measu e
de e minan s
o
HCE
accoun ing
o
egional
(p o inces)
e ec s
1965–1991
No
indica o
o
decen aliza ion
bu
accoun
o
possible
egional
di e ences
in
HCE
HCE
highly
ela ed
o
pe
capi a
income
(bu
elas ici y
below
1),
ageing
and
ede al
ans e s.
Reflec s
p o incial
HCE
eflec s
ans e s
om
CG
which
howe e
do
no
co ec
egional
income
disc epancies
Jiménez-Rubio (2008
J.-R.) Canada
Measu e
income- ela ed
inequali ies
in
heal h
and
heal h
ca e
use
ac oss
egions
(p o inces)
2001
No
indica o
o
decen aliza ion
bu
measu e
egional
di e ences
in
inequali ies
Heal h
ca e
use
inequali ies
a e
d i en
by
be ween-p o inces
a ia ion.
Inequali ies
in
heal h
a e
mos ly
d i en
by
wi hin-p o inces
a ia ion.
Di e en
esul s
o
heal h
and
heal h
ca e
inequali ies
seem
o
sugges
o he
heal h
de e minan s
a he
han
heal h
ca e
access
a e
essen ial
o
explain
inequali ies
HCE,
heal h
ca e
expendi u es;
CG,
cen al
go e nmen ;
SCG,
sub-cen al
go e nmen ;
GP,
gene al
p ac ice;
AC,
au onomous
communi ies.
in
heal h
is
a
majo
a gumen
agains
decen aliza ion,
unless
s ong
solida i y
mechanisms
a e
pu
in
place.
The
con a-
dic o y
expec a ions
abou
e ficiency
equi e
ca e ul
empi ical
alida ion.
In
hese
imes
o
ad e se
economic
ci cums ances,
e ficiency
is
mo e
han
e e
a
majo
issue
in
decision-making,
and
s ong
e idence
is
necessa y
be o e
ad oca ing
o
decen-
aliza ion
in
heal h.
Empi ical
e idence
In
o de
o
ge
an
exhaus i e
o e iew
on
empi ical
e idence,
we
pe o med
a
sys ema ic
li e a u e
e iew.
We
conduc ed
a
compu e ized
li e a u e
sea ch
in
PubMed
(Na ional
Cen e
o
Bio echnology
In o ma ion,
Be hesda,
Ma yland)
and
Google
Schola ,
supplemen ed
by
a
sea ch
o
quo ed
e e ences.
Tex
keywo ds
used
in
he
sea ch
included
decen aliza ion,
ede -
alism,
heal h,
heal h
ca e,
equi y,
and
e ficiency.
We
es ic ed
ou
analysis
o
hose
pe o med
in
OECD
coun ies
o
allow
o
ele an
compa isons,
ha
is,
in
a
compa able
con ex .
We
only
included
empi ical
s udies
abou
he
impac
o
decen-
aliza ion,
hence
excluding
heo e ical
s udies,
s udies
abou
legisla ion
o
abou
poli ical
o
o ganiza ional
aspec s
o
decen aliza ion,
edi o ials
and
li e a u e
e iews.
The
sea ch
was
limi ed
o
English-language
a icles
published
om
1995
o
July
2012,
when
decen aliza ion
in
heal h
has
become
a
el-
e an
and
applied
policy
op ion
in
OECD
coun ies.
Ou
sea ch
allowed
collec ing
17
pape s,
whose
main
cha ac e is ics
and
esul s
a e
displayed
in
Table
1.
S udies
we e
di ided
acco ding
o
dis inc
analyzes,
based
on
he
mos
ele an
decen aliza-
ion
ou comes,
namely
inequali ies
in
heal h
and
heal h
ca e,
heal h
expendi u es
and
heal h
ou comes.
Inequali y
in
heal h
and
heal h
ca e
use
The
consequences
o
decen aliza ion
on
inequali y
we e
he
main
issue
analyzed
by
Zhong13,
Jiménez-Rubio
e
al.1,
Cos a-
i-Fon 26 and
F ede iksson
and
Winblad27.
Zhong13 ound
ha
in
Canada
inequali ies
in
heal h
ca e
use
(o e all,
wi hin
p o inces
and
be ween
p o inces)
dec eased
a e
decen al-
iza ion.
Mos
inequali y
was
explained
by
wi hin-p o inces
a ia ion,
while
be ween-p o inces
a ia ions
did
no
much
con ibu e
o
inequali y.13 By
con as ,
Jiménez-Rubio
e
al.1
also
using
da a
om
Canadian
p o inces
obse ed
ha
income- ela ed
inequali ies
in
heal h
ca e
use
esul ed
om
be ween-p o inces
a ia ions
while
income- ela ed
inequali-
ies
in
heal h
we e
ela ed
o
wi hin- egion
a ia ions.
While
Zhong13 used
an
o e all
inequi y
measu e,
Jiménez-Rubio
e
al.1used
an
income- ela ed
one,
which
possibly
explains
he
disc epan
esul s.
The
esul s
by
Jiménez-Rubio
e
al.1
ce ainly
ques ion
he
equaliza ion
scheme
ac oss
p o inces;
he
edis ibu ion
o
unds
om
iche
o
poo e
p o inces
does
no
seem
o
a oid
inequi y
in
heal h
ca e
use
ela ed
o
be ween-p o ince
a ia ions
( o
example,
heal h
ca e
use
is
lowe
in
Quebec
ha
has
a
lowe - han-a e age
income
pe
capi a).
As
he
au ho
emphasizes,
ad oca es
o
decen al-
iza ion
would
howe e
conside
hese
di e ences
as
ela ed
o
di e en
p e e ences,
hence
legi ima e.
Income- ela ed
inequi y
in
heal h
was
mos ly
ela ed
o
di e ences
be ween
ich
and
poo
people
wi hin
p o inces,
he e
ques ioning
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
81
he
e o s
by
p o inces
o
educe
inequali ies
(would
a
cen al
go e nmen
be
mo e
commi ed
o
educe
inequi y
in
heal h?).
Fo
Spain,
Cos a-i-Fon 26 ound
a
small
o e all,
wi hin
and
be ween
egions
income- ela ed
inequali y
in
heal h,
al hough
somewha
highe
in
egions
wi h
lowe
au onomy.
The
au ho
explains
his
di e ence
by
he
g ea e
ole
o
he
p i a e
sec o
in
hose
egions.
He
also
emphasizes
ha
decen aliza ion
may
ha e
a o ed
equi y
due
o
a
high
commi men
o
egions
o
achie e
his
objec i e,
which
co esponds
o
a
high
ci izens’
conce n.
Finally,
F ed iks-
son
and
Winblad27 s udied
a
specific
e o m
owa d
mo e
p o ide
choice
in
Sweden,
whe e
heal h
ca e
use
is
highly
decen alized
a
he
coun y
le el.
The
au ho
showed
ha
while
some
coun ies
suppo ed
he
e o m
h ough
a o ing
choice,
o he s
implemen ed
adminis a i e
ba ie s
agains
i .
The e o e,
au onomy
on
decision
making
( ega ding
p o ide
choice)
esul ed
on
inequali ies
be ween
people
li ing
in
di e en
Swedish
coun ies.
Heal h
expendi u es
Inequali y
ela ed
o
decen aliza ion
was
also
obse ed
indi-
ec ly,
h ough
examining
he
impac
o
egional
GDP
pe
capi a
on
heal h
ca e
expendi u es.
Indeed,
a
g ea e
impac
o
egional
income
on
heal h
ca e
expendi u e
would
mean
ha
iche
egions
spend
significan ly
mo e
on
heal h
han
poo
ones,
hence
c ea ing
inequali ies
in
heal h
ca e
use.
Al hough
all
s udies
using
egional
da a
conclude
ha
heal h
ca e
expendi u es
do
no
a y
much
wi h
income
( he
elas ici y
wi h
espec
o
income
is
below
one),
esul s
a e
con as ed
and
a y
be ween
coun ies.
Fo
Spain,
P ie o
and
Lago-Pe ˜
nas28
ound
a
posi i e
ela ion
be ween
income
and
public
heal h
expendi u es
only
in
egions
wi h
highe
fiscal
au onomy.
Regions
wi hou
fiscal
au onomy
benefi ed
om
na ional
equaliza ion
e o s
and
he e o e
managed
o
achie e
high
le els
o
heal h
ca e
expendi u es
despi e
o
lowe
income.
The
low
impac
o
income
on
heal h
ca e
expendi u es
in
Spain
was
confi med
by
Can a e o,
Cos a-i-Fon
and
Moscone.29,11
The
same
conclusion
could
be
d awn
o
Canada,21 al hough
he
impac
o
income
on
heal h
expendi u es
was
g ea e
han
in
Spain.
Using
a
specific
indica o ,
Di
Ma eo
and
Di
Ma eo30
showed
he
posi i e
impac
o
ede al
ans e s
on
heal h
ca e
expendi u e.
Tha
is,
ede al
ans e s
ce ainly
educe
he
impac
o
egional
income
on
expendi u es
bu
do
no
ully
co ec
egional
income
disc epancies.
The
pic u e
is
some-
wha
di e en
in
I aly,
whe e
egional
income
had
a
s ong
impac
on
heal h
ca e
expendi u es.31 The
au ho s
showed
also
ha
cos -con ainmen
measu es
ha e
no
al e ed
c oss-
egional
di e ences
in
heal h
ca e
expendi u es
and
ha e
e en
wo sened
hem
in
some
cases.
Fo
Swi ze land,
C i elli
e
al.32 obse ed
a
la ge
impac
o
physician
supply,
which
is
highly
a iable
ac oss
egions,
on
heal h
ca e
expendi u es.
Hence
c oss- egional
di e ences
in
heal h
ca e
supply
may
also
be
a
sou ce
o
inequali y
ac oss
egions.
Finally,
using
da a
om
110
egions
o
8
OECD
coun ies,
Lopez-Casasno as
and
Saez33 showed
a
low
elas ici y
o
heal h
ca e
expendi u e
wi h
espec
o
income.
Howe e ,
he
impac
o
income
was
highe
ac oss
egions
in
coun ies
wi h
highe
in e - egional
income
inequali ies,
leading
in
u n
o
highe
disc epancies
in
heal h
ca e
expendi u es.
Sche fle
and
Smi h34,
Cos a-i-Fon
and
Moscone11,
Cos a-
i-Fon
and
Pons-No ell22 add essed
he
impac
o
decen al-
iza ion
on
heal h
expendi u es,
while
Bo dignon
and
Tu a i35
examined
he
impac
o
he
in e ac ion
be ween
cen al
and
sub-cen al
go e nmen s.
Sche fle
and
Smi h34 showed
o
Cali o nia
ha
spending
on
uninsu ed
pa ien s
dec eased
wi h
he
highe
coun ies’
au onomy
in
alloca ing
unds
ac oss
heal h
and
social
p og ams.
Cos a-Fon
and
Pons-No ell22
obse ed
ha
Spanish
egions
wi h
bo h
poli ical
and
fis-
cal
au onomy
had
highe
heal h
expendi u es
as
compa ed
o
egions
wi hou
au onomy
o
wi h
poli ical
au onomy
only.
Cos a-i-Fon
and
Moscone11 obse ed
howe e
ha
decen aliza ion
inc eased
expendi u es
only
in
he
sho
un,
p oducing
sa ings
in
he
long
un.
They
a gued
ha
fi s ly
decen aliza ion
inc eased
heal h
ca e
cos s
due
o
sunk
cos s;
hen
his
e ec
was
minimized
in
he
long
un
by
an
expe ience
cu e,
ha
is,
lea ning-by-doing
allowed
educing
cos s.
Rega ding
assump ions
om
economic
he-
o y,
Cos a-i-Fon
and
Moscone11 also
showed
he
exis ence
o
spa ial
co ela ion
o
heal h
ca e
expendi u es
be ween
neighbo
ju isdic ions,
sus aining
he
assump ion
o
ho izon-
al
spillo e s.
This
esul
may
hence
be
due
o
compe i ion
be ween
neighbo
ju isdic ions
o
inc ease
a ac i eness
o
due
o
poli icians
being
judged
based
on
he
neighbo
coun-
e pa s’
ac ions.
Finally,
Bo dignon
and
Tu a i35 used
1990s
I alian
adjus -
men
p ocess
imposed
by
an
ex e nal
en i y
( he
Eu opean
Union)
o
unde s and
he
ole
o
expec a ions
and
he
s a e-
gic
in e ac ions
be ween
he
cen al
s a e
and
egions,
using
da a
on
heal h
ca e
expendi u es.
Regions
expec ing
ha
he
cen al
go e nmen
will
in e ene
in
case
o
deb
so ened
hei
budge
cons ain s
and
he
ela ionship
be ween
he
unding
hey
ecei e
om
he
cen al
go e nmen
and
heal h
expendi u es
was
lowe .
By
con as ,
du ing
ex e nal
adjus -
men ,
he
cen al
go e nmen
was
pe cei ed
as
mo e
s ingen
and
he
link
be ween
unding
and
expendi u es
became
s onge .
And
he
cen al
go e nmen
was
also
mo e
p one
o
cu
unding
o
egions
because
i
knew
i s
commi men
no
o
bailou
would
be
aken
mo e
se iously.
In e es ingly,
au ho s
showed
ha
mo e
au onomy
lead
o
mo e
finan-
cially
esponsible
policies
because
expec a ions
o
bailou
we e
low.
This
pape
hus
emphasized
ha
ans e
mecha-
nisms,
which
a e
always
p esen
in
decen alized
coun ies,
c ea ed
incen i es
ha
may
be
mo e
o
less
de imen al
o
e ficiency
depending
on
how
hey
a e
designed
and
pu
in
p ac ice.
Heal h
esul s
Resul s
abou
he
impac
o
decen aliza ion
on
heal h
ou comes
a e
sca ce
bu
consis en
ac oss
s udies.
Jiménez-
Rubio12,36 and
Can a e o
and
Pascual37 ag ee
on
iden i ying
a
posi i e
ela ionship
be ween
fiscal
decen aliza ion
and
heal h
esul s,
measu ed
h ough
in an
mo ali y.
Can a e o
and
Pascual37 used
an
al e na i e
measu e
o
heal h
s a us
(li e
expec ancy)
which
confi med
he
posi i e
con ibu ion
o
decen aliza ion
o
heal h
s a us.
82
e
p
o
s
a
ú
d
e
p
ú
b
l
i
c
a
.
2
0
1
3;3
1(1):74–83
Conclusion
Al hough
he
economics
li e a u e
has
la gely
discussed
he
impac
o
decen aliza ion,
empi ical
esul s
emain
sca ce,
in
pa icula
ega ding
decen aliza ion
in
heal h.
Ne e he-
less,
decen aliza ion
o
heal h
policies
and
unding
has
been
la gely
implemen ed
in
a ious
coun ies
and
is
gene ally
con-
side ed
as
a
ele an
op ion
o
imp o e
e ficiency
in
heal h
ca e.7
Ou
e iew
o
he
li e a u e
allows
d aw
p elimina y
conclusions,
based
on
ela i ely
ecen
decen aliza ion
expe-
iences,
in
pa icula
in
Canada,
I aly,
Spain
and
Swi ze land.
Fi s
and
o emos ,
de olu ion
o
poli ical
(and
fiscal)
au ho -
i y
o
sub-go e nmen
le els
seems
o
inc ease
heal h
ca e
expendi u es
bu
also
imp o e
heal h
ou comes,
mainly
mea-
su ed
h ough
in an
mo ali y
a e.
Acco ding
o
economic
heo y,
decen aliza ion
may
os e
compe i ion
be ween
ju isdic ions
o
inc ease
a ac i eness,
and
inc ease
p essu e
on
local
go e nmen s
because
ci izens
e alua e
hei
pe o -
mance
based
on
neighbo
coun e pa s’
ac ions.
The
empi ical
li e a u e
sugges s
ha
enhanced
compe i ion
p omp s
local
decision-make s
o
inc ease
heal h
ca e
expendi u es
( ace
o
he
op)
and
no
he
e e se
in
o de
o
dec ease
axes
( ace
o
he
bo om),
wi h
a
a o able
impac
on
heal h.
The
case
o
Spain
is
pa icula ly
enligh ening,
which
shows
ha
he
inc ease
in
heal h
ca e
expendi u es
has
been
he
highes
in
egions
wi h
fiscal
au onomy.
This
esul
may
eflec
he
high
people’s
demand
o
high-quali y
heal h
se ices,
and
he
highe
esponsi eness
o
local
au ho i ies
o
his
p e e ence.
The
highe
heal h
ca e
expendi u es
unde
decen aliza ion
may
howe e
eflec
also
highe
cos s,
ela ed
o
ins ance
o
duplica ion
o
inpu s
( wo
neighbo
egions
o e ing
simila
se ices
which
could
be
sha ed),
diseconomies
o
scale,
o
he
sunk
cos s
associa ed
o
implemen a ion
o
a
local
heal h
p o-
ision
scheme.
To
some
ex en ,
ad e se
incen i es
may
play
some
ole
i
sub-cen al
go e nmen s
a e
no
ully
financially
esponsible
o
hei
expendi u es.
In
pa icula ,
mo al
haz-
a d
may
exis
i
sub-cen al
go e nmen s
expec
hei
deb s
being
co e ed
by
he
cen al
s a e
(bailou )
o
i
hei
budge
depends
mo e
la gely
om
ans e s
om
he
cen al
s a e.
In
a
ew
wo ds,
decen aliza ion
does
no
appea
a
fi s
sigh
as
a
means
o
con ol
o
educe
heal h
ca e
expendi u es,
bu
as
an
incen i e
o
p o ide
be e
and
possibly
mo e
expensi e
se ices.
The
e ficiency
consequences
a e
hus
ambiguous,
as
i
is
unclea
whe he
addi ional
benefi s
–
measu ed
h ough
a
e y
educed
numbe
o
indica o s
–
a e
wo h
addi ional
cos s.
Rega ding
equi y,
empi ical
esul s
a e
ambiguous
and
ce ainly
ela ed
o
he
specific
coun ies’
con ex .
Inequi y
in
heal h
and
heal h
ca e
is
low
in
Spain,
whe e
heal h
expendi u es
a e
also
poo ly
ela ed
o
he
egion
GDP/capi a.
Addi ionally,
inequi y
in
heal h
ca e
seems
o
be
lowe
in
mo e
au onomous
egions,
so
ha
decen aliza ion
may
ha e
a o ed
equi y.
In
Canada,
income- ela ed
inequi y
in
heal h
ca e
is
ela ed
o
be ween-p o inces
inequali ies.
In
I aly
and
Swi ze land,
he e
is
a
s ong
ela ionship
be ween
egions
( esp.
can ons)
income
and
heal h
ca e
expendi u es,
esul -
ing
in
a
la ge
he e ogenei y
in
heal h
ca e
expendi u es.
These
esul s
may
ce ainly
be
ela ed
o
lowe
equalizing
mechanisms
in
hese
wo
coun ies
coupled
wi h
a
highe
fiscal
au onomy.
No e
also
ha
di e en
decisions
ac oss
egions,
namely
abou
physicians
o
equipmen
supply,
also
po en ially
c ea e
di e ences/inequi y
in
heal h
ca e
deli e y.
To
conclude,
solida i y
mechanisms
ac oss
sub-cen al
au ho i ies
a e
ele an
o
a oid
he
eme gence
o
la ge
inequali ies
ac oss
egions
in
heal h
ca e
deli e y
and
expend-
i u es.
Howe e ,
edis ibu ion
o
unds
also
educes
ju is-
dic ions’
financial
esponsibili y,
wi h
possible
de imen al
consequences
on
expendi u es
and
ambiguous
consequences
on
e ficiency.
This
las
aspec
is
o
pa icula
impo ance
i
ju isdic ions
compe e
wi h
each
o he
o
p o iding
high-
quali y
se ices
and
no
h ough
lowe ing
ax
a es.
Funding
This
in es iga ion
was
unded
by
he
Fundac¸ão
pa a
a
Ciência
e
a
Tecnologia
h ough
he
p ojec
PTDC/EGE-
ECO/104094/2008.
Conflic s
o
in e es
The
au ho s
decla e
ha
he e
a e
no
conflic s
o
in e es s.
e
e
e
n
c
e
s
1.
Jiménez-Rubio
D,
Smi h
P,
Doo slae
E.
Equi y
in
heal h
and
heal h
ca e
in
a
decen alised
con ex :
e idence
om
Canada.
Heal h
Econ.
2008;17:377–92.
2.
Oa es
W.
Fiscal
ede alism.
New
Yo k:
Ha cou
B ace
Jo ano ich;
1972.
3.
Iimi
A.
Decen aliza ion
and
economic
g ow h
e isi ed:
an
empi ical
no e.
J
U ban
Econ.
2005;57:449–61.
4.
Rod iguez-Pose
A,
Ezcu a
R.
Does
decen aliza ion
ma e
o
egional
dispa i ies?:
a
c oss-coun y
analysis.
J
Econ
Geog .
2010;10:619–44.
5.
Galiani
S,
Ge le
P,
Scha g odsky
E.
School
decen aliza ion:
helping
he
good
ge
be e ,
bu
lea ing
he
poo
behind.
J
Public
Econ.
2008;92:2106–20.
6.
V angbaek
K.
Towa ds
a
ypology
o
decen aliza ion
in
heal h
ca e.
In:
Sal man
RB,
Bankauskai e
V,
V angbaek
K,
edi o s.
Decen aliza ion
in
heal h
ca e.
Maidenhead:
McG aw-Hill.
Open
Uni e si y
P ess;
2007
(Eu opean
Obse a o y
on
Heal h
Sys ems
and
Policies
Se ies).
p.
44–62.
7.
Bankauskai e
V,
Sal man
RB.
Cen al
issues
in
he
decen aliza ion
deba e.
In:
Sal man
RB,
Bankauskai e
V,
V angbaek
K,
edi o s.
Decen aliza ion
in
heal h
ca e.
Maidenhead:
McG aw-Hill.
Open
Uni e si y
P ess;
2007
(Eu opean
Obse a o y
on
Heal h
Sys ems
and
Policies
Se ies).
p.
9–21.
8.
Bankauskai e
V,
Dubois
HFW,
Sal man
RB.
Pa e ns
o
decen aliza ion
ac oss
Eu opean
heal h
sys ems.
In:
Sal man
RB,
Bankauskai e
V,
V angbaek
K,
edi o s.
Decen aliza ion
in
heal h
ca e.
Maidenhead:
McG aw-Hill.
Open
Uni e si y
P ess;
2007
(Eu opean
Obse a o y
on
Heal h
Sys ems
and
Policies
Se ies).
p.
23–43.
9.
Jouma d
I,
And é
C,
Nicq
C.
Heal h
ca e
sys ems:
e ficiency
and
ins i u ions.
Pa is:
OECD
Publishing;
2010
(OECD
Economics
Depa men .
Wo king
Pape s;
769).
10.
Pa is
V,
De aux
M,
Wei
L.
Heal h
sys ems
ins i u ional
cha ac e is ics:
a
su ey
o
29
OECD
coun ies.
Pa is:
OECD