Decentralizing Central Targets: Decentralizing Central Targets: Empirical Evidence from Italy Empirical Evidence from Italy M. A. Antonelli M. A. Antonelli Università “La Sapienza”- Università “La Sapienza”- Roma Roma V. Grembi V. Grembi Università Cattolica - Università Cattolica - Milano Milano
23
Embed
Decentralizing Central Targets: Empirical Evidence from Italy
Decentralizing Central Targets: Empirical Evidence from Italy. M. A. Antonelli Università “La Sapienza”- Roma V. Grembi Università Cattolica - Milano. Paper Aims. Targeting local services (quality or quantity) Ex ante justification Ex post justification Crèmer e Palfrey (2000 e 2002). - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Decentralizing Central Targets:Decentralizing Central Targets:Empirical Evidence from ItalyEmpirical Evidence from Italy
M. A. AntonelliM. A. Antonelli
Università “La Sapienza”- Roma Università “La Sapienza”- Roma
V. GrembiV. Grembi
Università Cattolica - MilanoUniversità Cattolica - Milano
Paper AimsPaper Aims
• Targeting local services (quality or quantity)
E
x
a
n
t
e
j
u
s
t
i
f
i
c
a
t
i
o
n
E
x
p
o
s
t
j
u
s
t
i
f
i
c
a
t
i
o
n
Crèmer e Palfrey (2000 e 2002)
Basic expenditure decision model in two settings unitary settingdecentralized setting
In both settings the policy maker has to decide how to
split budget among a bunch of good;
Check with Italian unbalanced panel (2002-2007) with
data at the municipal level on public childcare
OutlinesOutlines
The modelThe modelunitary setting (1)unitary setting (1)
• Government: 1 level Government: 1 level • Two services Two services xx andand yy • Quality is given. Quality is given. • The pThe policy maker maximize an utility function which olicy maker maximize an utility function which
is additive respect to the expenditure compis additive respect to the expenditure components (x onents (x and y) subject to a budget constraintand y) subject to a budget constraint
• Each component is weighed by a parameterEach component is weighed by a parameter (θ) which (θ) which stands forstands for political visibility one service political visibility one service
Unitary setting (2)Unitary setting (2)
• Maximization problem:Maximization problem:
• θ is the weigh which is associated to each θ is the weigh which is associated to each service from the politician perspeservice from the politician perspectivective;;
• T T = total revenues= total revenues;;• andand represent expenditure level (funded by T)represent expenditure level (funded by T)
maxx
T,y
T
UL ln xT (1 ) ln yT
T xT yT
xT
yT
0;1
Unitary setting (3)Unitary setting (3)
Solutions for the expenditure levelSolutions for the expenditure level
andand
Funds are allocated according toFunds are allocated according to θ θ
xT T
yT (1 )T
Decentralized Setting Decentralized Setting
• Government: two levelsGovernment: two levels• Local government is in charge for the allocation of 2 Local government is in charge for the allocation of 2
services (services (x x andand y y).).• The governThe government set and expenditure target ment set and expenditure target
(quantitative terms) on x in order to assure a (quantitative terms) on x in order to assure a minimum level of the serviceminimum level of the service
Decentralized SettingDecentralized Setting
Capital expenditures for theCapital expenditures for the supply expenditure are supply expenditure are funded by the central levelfunded by the central level;;
The local level is in charge for the current The local level is in charge for the current expenditureexpenditure a supply expansion means and induced a supply expansion means and induced local expenditurelocal expenditure;;
The local level gets a central grant for its service The local level gets a central grant for its service management expenditures (based on several given management expenditures (based on several given parameters)parameters)
Decentralized targetDecentralized target
• The current expenditures for x canThe current expenditures for x can be covered be covered:: (local funding)(local funding) (central funding)(central funding)
• The target can be The target can be partially funded by with partially funded by with which represents target implementation due to the which represents target implementation due to the central grantcentral grant
• represents the institutional characteristics which represents the institutional characteristics which rules the intergovernmental relationships among rules the intergovernmental relationships among levels (impact on the politician decision to match the levels (impact on the politician decision to match the target)target)
xT
x B
x B
0;1
Maximization ProblemMaximization Problem
OneOne--perioperiodd model (myopthic politician) model (myopthic politician) XXE E and Yand YEE are the total expenditure of the local are the total expenditure of the local
government for running local services x and ygovernment for running local services x and yCurrent expenditure for x are given byCurrent expenditure for x are given by
subject to: T+B= subject to: T+B= XXE E +Y+YEE
XE
XE xT x B
maxXE ,YE
U ln(XE xT ) (1 ) ln(YE yE )
Maximization problemMaximization problem
andand are the local expenditure levels set in are the local expenditure levels set in the unitary the unitary settingsetting;;
The maximization is on the expenditure The maximization is on the expenditure variations characterizing the decentralized variations characterizing the decentralized setting compared to the unitary settingsetting compared to the unitary setting
The expenditure levels areThe expenditure levels are
xT
yT
XE
* xT (T B xT yT )
xT
* xT (T B xT yT ) x B
YE
* yT (1 )(T B xT yT )
Statica comparataStatica comparata
Value of parameters
Sign of the lo cal funding expenditure
( *Tx )
Sign of the induced expenditure
( Bx )
Sign of the total expenditure
( EX )
=0 *Tx <0 Bx >0 EX =0 the t otal current expenditure is invariant.
The spending policy is characterized by a reallocation of resources from the local funding expenditure to the induced expenditure (claim 2)
<1 *Tx <0 Bx >0 EX >0 the total current expenditure increases. The
induced expenditure increases while the local funding expenditure decreases. The spending policy is characterized by a reallocation of resources from the local funding expenditure to the induced expenditure (claim 2)
1<
<dB
xd B
*Tx <0 Bx >0 EX >0 the total current expenditure increases. The
induced expenditure increases while the local funding expenditure decreases. The spending policy is characterized by a reallocation of resources from the local funding expenditure to the induced expenditure (claim 3)
>1 *Tx >0 Bx >0 TX >0 the t otal current expenditure increases.
Both expenditure components increase
= *Tx >0 Bx =0 TX >0 the total current expenditure increases but
the induced expenditure’s component is equal to zero. Therefore, the entire central grant is used to increase the lo cal funding expenditure. (This case occurs when 0 ).
Empirical PartEmpirical Part
• A supply expansion policy on a targeted service within a multilevel institutional framework depends both on local preferences (political visibility of the service) and on the institutional framework
• Attempt to assess how the local decisions on local service supply react to these elements
• A unique attempt to implement minimum level in Italy: childcare 2010 Lisbon target
Italian institutional frameworkItalian institutional frameworkFinanziamenti Livelli di Governo Funzioni
Stanziamenti "una tantum" per l'estensione dell'offerta
Stato Definizione dei LEP
Stanziamenti regionali per l'estensione dell'offerta e la gestione del servizio
RegioniFunzione di
programmazione
Ripartizione delle risorse regionali tra gli enti di livello inferiore
Province (o ambiti territoriali)
Formazione degli educatori e/o funzioni di
programmazione
Copertura dei costi di gestione del servizio con compartecipazione al costo da parte degli utenti
Comuni (o unioni)
Gestione del servizio
Italian Case StudyItalian Case Study
• Political visibility or local preferences Del Boca et al. 2005. Cultural differences in the
demand for the service (e.g. Southern/Northern European States)
Central role of the family in providing social services and financial support to its members
The value families attach to the care of young children
• Instrumental Variable approach• Hausman Taylor estimator
• HT is as IV estimator which relies for endogenous time invariant covariates on the mean of the exogenous time-varying covariates and for the endogenous time-varying their own averages
(1.54) (-2.20) (1.54) (1.06) (1.54) (0.98)Macro Area Effect No No No No Yes Yes Yes Yes No No No No
Regional Effects No No No No No No No No Yes Yes Yes YesConstant 0.25*** 0.05* 0.34*** 0.21*** 0.25*** 0.16*** 0.34*** 0.26*** 0.25*** 0.33*** 0.34*** 0.20***