DECENTRALIZATION AND POVERTY REDUCTION India and Indonesia Democratic Decentralization Relative Grad

1
DECENTRALIZATION AND POVERTY REDUCTIONIndia and
Indonesia Democratic DecentralizationRelative
Gradualism versus Big Bang OECD Paris
29/09/2004

• William McCarten
• World Bank
• with inputs from Kai Kaiser,Bert Hofman,
  Vikram Chand

2
OECD Questions

• Present empirical findings from case studies
  focusing on the impact of decentralization on (i)
  participation,(ii) empowerment,(iii) access to
  services and (iv) local governance -control of
  corruption.Theme levels versus ?s
• Identify common patterns and determinants
  (political, institutional, fiscal formulas)
  under which decentralization policies are most
  likely to be pro-poor. Political rules for
  decision making
• Explore in which regulatory areas of public
  service delivery are the potential benefits of
  decentralization for the poor most likely to
  occur.
• Propose policy measures to improve existing donor
  policies promoting decentralization and local
  governance.

3
A conceptual framework

• A simple, but highly focused, framework for
  examining decentralization through three
  challenges
• (1) moving government closer to the people with
  increased accountability to local preferences
• (2) establishing a sustainable, higher level
  fiscal framework for local government and
• (3) improving local service delivery and better
  access for poor (ergo avoid elite capture)

4
A little decentralization political theory

• Timothy Besley (1997) argues that
  decentralization enhances the prospects for
  poverty reduction only when it leads to
  fundamentally (i) new institutions, (ii) changes
  political structures, (iii) improved governance,
  or (iv) changed attitudes towards the poor.
• James Manor (2004) advises that poor become more
  powerful politically
• (i) when the ruling party in the political system
  is pro-poor or follows redistributive policies
  (e.g. West Bengal since 1970s ) ,
• (ii) when rival parties and elites in the
  political system and at lower levels, compete to
  appeal to poor voters (e.g. Last Indian general
  election- produced a government committed in its
  Common Program to being more inclusive and to
  make decentralization work for rural poor.)
• (iii) when pro-poor coalitions have strength
  within civil society. Hence organizational
  training for marginalized groups is used as a
  tactic.
• v) when poor people are given exclusive or
  predominant influence within decentralized
  bodies. (West Bengal and parts of Kerala ) ,
• (v) when poor and excluded groups gain sufficient
  seats on decentralized bodies to make it
  necessary for leaders to build alliances with
  their representatives. (Power-sharing
  arrangements to ensure stability in heterogeneous
  societies)

5
Social Capital Glue for Effective
Decentralization

• Paul Seabright (1997) suggested that when local
  community activities involving wide coordination
  and co-operation, are successfully implemented
  there is an increased likelihood of future
  co-operation and a cumulative process of
  augmenting social capital.
• Note Kerala started the process of
  decentralization with high social capital
• Indonesia may still be starved of non government
  institutions that cut across religious and ethnic
  boundaries (See Andrew Rosser IDS Working paper,
  Indonesia the Politics of Inclusion (2004))
• What are potential bridging social capital
  /glue creating strategies?

6
Indonesia

• Decentralization is response to heavy handed
  centralist ways of New Order Suharto Government.
  
• Some hierarchical relationships disappears
• Hence Big Bang in 1999-2001 (administrative,
  political and fiscal )
• Transfer responsibilities, 2 million of 3.4
  million central civil servants transferred
• 400 local Government 30 provinces
• 16,000 facilities handed over to the regions.
• Brand new intergovernmental fiscal system
• New political system, new accountability
  arrangements
• No harm clause resulted in less equalization
  in transition
• DAU , General Transfer Political Bargaining
  historical legacy
• Law 22/1999 of Regional Governance shifted the
  balance from a largely top-down form of
  accountability to local political accountability.
• Election of the head of region, and the annual
  accountability speech.
• The head is elected by and accountable to the
  local parliaments (DPRD). Law 22/1999 Article
  16(2) stipulates that the regional head of the
  executive (bupati/walikota) and legislative act
  in partnership (kemitraan).

7
Development spending increased
Central and regional development spending,
percent of GDP
8
Unequal transfers
9
Indonesia Challenges

• Wage bill to formula element major service
  disruption, but wage setting is done nationally
• Minimum standards are vital, but might be a
  Trojan horse for recentralization without
  transparent definitions and feasible timetables
• Provinces intentionally forgotten, but came
  back from dead in final design
• DAU to regions and districts
• Development framework for budgetary on leading
  and coordinate donors in decentralization
• Local government highly fiscal dependence and
  this works against accountability and own tax
  effort
• Indonesia pre-Big Bang had an overstaffed,
  unresponsive civil service.
• Yogyakarta, a province with too many civil
  servants, began to experiment in 2002 with
  civil service reform by reviewing functions,
  organization an staffing. It has offered
  flexible severance packages

10
Peoples Perception of Services after One Year
11
Corruptionretail or speed money
12
Indonesia Unfinished Agenda

• Incentive for regions to improve finances
• Lump sum components of transfers
• Capacity building
• Natural resource components of revenue sharing
  with revenue assignments leads to inequality in
  own revenue
• Enhancing accountability
• Incomplete definitions of assignments and
  political responsibilities

13
Directions for Indonesia reforms

• The center should devolve more resources
• Preferably in the form of tax base
• The DAU should become more equalizing
• And governance should improve
• Through more clarity in local functions
• .better rules on financial management and
  procurement.
• ..better rules for local politics..
• And DAKs (program grants ) with incentives to
  perform such as minimum service standards

14
India 3rd Tier Decentralization

• The impacts of varying a few institutional
  components in one state can be compared to
  conditions in other states
• Indian rural decentralization process has led to
  both successes and failures.
• The success stories, including qualified
  successes, of which I argue Madhya Pradesh is
  one, are often misunderstood

15
Implementation of Indian decentralization by
states

• Great variations among states
• Regular, competitive, elections with high
  participation but (differing approaches to
  permitting political party affiliation.)
• Lagging fiscal decentralization
• Lagging administrative decentralization
• Limited accountability downwards (MP continuing
  experimentation,Kerala 33 capital budget
  devolution)
• Problems of inclusion (SC/ST, OBC )
• In most states insufficient information,little
  transparency, little monitoring, risk of
  corruption

16
Design of decentralization

• Constitutional Amendments (73d, 74th and
  ESA-tribal)
• Political Decentralization and Accountability
  Self-government 3 tier system
• Gram Sabha ? town call meeting with quorum
• State Acts
• Administrative Decentralization functions
  (potential 29 functions ) and roles
• State Finance Commissions
• Fiscal decentralization transfers and own
  revenues
• Neglected

17
Plot of Road Access Children in School (NFHS
Sample ) Circa 1992
18
Awareness Political Incentives

• In 1994 Madhya Pradesh Government mobilised
  newly PRIs energies by organising a massive
  survey on education access .
• Canvassed PRI member in their constituencies.
  Who was not attending school? From the results,
  local elected representatives were encouraged
  to prepare a Village Education Register. This
  became the basis for further planning in the
  field of education .
• A Human Development Report published in1995
  showed how far the state had to travel in social
  indicators indicators. The Chief Minister
  regarded it as a record of the state's failures
  that had to be set right.
• Subsequent HDRs then serve as a useful monitoring
  tools. Opposition parties too have used it to put
  pressure on the government. The HDR has been an
  important element in development debates in MP
• European Union donor funding for DPEP and ADB
  program loan sequenced after political
  decentralization and back-stopped flexible wage
  policy .

19
Rajiv Gandhi missions Parallel agencies and
competitor to main departments Innovative
substitute for civil service reform

• While the PRIs were created to articulate local
  demands,government resources and second track
  was brought under special Rajiv Gandhi missions
  led by especially selected and committed
  officers.
• These missions were given a degree of autonomy in
  decision making and they reported almost
  directly to the CM.
• Once local PRI were articulated and the needs
  verified the mission released funds to the
  panchayat. The teacher is paid by the sarpanch .
  This is an attempt to create local
  accountability. 

20
A little more decentralization theory

• Without strong accountability mechanism and
  accessible networks to draw upon expert know
  how from state gov. decentralization may lead
  to efficiency and effective losses.
• Richard Bird notes that decentralization must
  find ways of coping with the challenge and
  opportunities of double information asymmetry
  when managing knowledge-based services.
• Higher-level governments may not know what is
  needed, while the local government may not know
  how to do it.
• Madhya Pradesh established the Rajiv Gandhi
  missions partly to cope with this double
  information asymmetry problem.

21
MP Education expansion strategy

• Reducing physical distance to schools has been
  the objective of EGS
• Creation of about 28,000 schools consisting of
  (usually) one or (at most) two teachers and at
  least 25 pupils in tribal areas, 40 elsewhere.
  The creation of an EGS school is conditional on
  the reception by the government of a request made
  by parents of prospective pupils through the gram
  panchayat, thus making the reach of the school
  supply responsive to parents or local
  sarpanches demands
• Participation in DPEP and the experimentation of
  new teaching methods in alternative schools (now
  merged with EGS ones) have also contributed to
  the extension of the public sector schooling .
• Dalit hamlets did not have easy access to
  existing schools social obstacles compounding
  physical ones.

22
Absenteeism Teachers2003 Survey Reported in
2004 WDR
23
Absenteeism Medical Staff
24
Literacy rate outcomes by state
25
Madhya Pradesh Implication

• A success given the initial conditions and great
  risk of caste, tribal and gender exclusion by
  local elites
• Madhya Pradersh tried to use direct democracy
  forum ( gram sabha ) to hold corruption in check
  and ensure accountability of the heads (sarpanch)
  of gram panchayat. Results not very successful.
  
• A continuous process of innovation which has yet
  to achieve an equilibrium
• Madhya Pradesh model is likely more
  appropriate for north Indian state than more
  heavily praised Kerala or West Bengal models
• (Many experts believe that West Bengal story
  cant be replicated) Crook, R. C. Sverrisson,
  A.S. (2001) Decentralization and Poverty
  Alleviation in Developing Countries A
  Comparative Analysis or, is West Bengal Unique?,
  IDS Working Paper 130 (June) 1-60 .

26
Shared patterns in India and Indonesia

• Civil service was corrupt or unresponsive
• Civil service was not accountable but
  decentralization may have an impacts.
• Lack of clarity on assignments of function in
  both countries
• Allocations for efficiency gains yet to be
  achieved
• Monitor performance of service delivery and
  publicize this to permit yardstick competition
  and decentralized inter-jurisdictional
  demonstration effect (Pierre Salmon concept )
• Donors have important role to play in design of
  monitoring systems. UNDP Report/ PRSP strategies
  / MDG international experience

27
When is Decentralization Pro Poor?

• Benefits of decentralization for the poor most
  likely to occur-
• Answer Primary education and primary health
  access and quality can be improved for
  disadvantages groups. Kerala, India has
  emphasized community development and self help
  projects for poor but sustainability requires
  good governance
• Few benefits or difficult benefits in
  participatory natural resource management
  (Example Central Government should design
  forestry management regulations but local
  government might take day to day management
  responsibility. )
• Corruption, including speed money, might be
  squeezed if accountability to people is
  developed However too much was expected from
  the direct democracy experiment in MP and
  people campaign for decentralization in Kerala
• Delegate democracy mechanism accountability need
  to be build up as well as participatory
  democracy and donors can play vital role in
  capacity building. Seat reservation,
  supermajority or decision making rules
  requiring accommodation
• Monopoly of traditional centralized civil service
  is contested

28
Accountability and Resource Efficiency Issues

• Community participation can be enhanced by
  contribution of buildings, teacher-parent
  committees, and accountability to local councils.
• Clarify well defined minimum standards
  monitoring in Indonesia
• Concentrate on observable, easy-to-monitor
  targets
• More local revenue needed to build accountability
• Soedjito and Kerstan paper advocate create an
  incentive system for local governments, but
  incentive that can be manipulated should be
  avoided as should formulas that lend themselves
  to creating moral hazard opportunities
• Decentralized civil service reform can play a
  vital role in accountability and poverty
  reduction, if Central Governments devolves
  discretionary freedom over civil service reform.
  Yogyakarta has taken initiative, but Indonesian
  Government has not approved. With lower wage bill
  there are more resources for non wage services.
  Use parallel small agencies with esprit , such as
  Gandhi Missions of MP , to solve knowledge
  problem and expand service delivery to the poor
  .
• Potential scope for better coordinating between
  multilaterals and bilateral donors, based on
  comparative advantage. Some "pass through" of on
  lending terms might reflect social goals, such
  as reaching MDG targets.

29
Policy measures to improve existing donor policies

• Training for disadvantaged groups who win
  elections to local governments
• Encourage monitoring and public dissemination
  of results Requiring the government to state its
  policy and spending intentions clearly is a vital
  precondition to holding it accountable
• In identifying remedies, delivering credible
  policies that are sustained, effective, and
  adequately financed goes to the heart of the
  political process
• Does the cabinet have clearly articulated policy
  priorities?
• Are the domestic stakeholderscivil society,
  business community, public interest groups, labor
  unions, farmers associations, and other interest
  groupsconsulted on policy?
• Benchmark access to services stratified by groups
  income levels.