Transcript Decentralization

Design and Management of
Decentralized Expenditures and
Transfers
Underpinnings for successful decentralization
EHTISHAM AHMAD
NOVEMBER 2012
Key questions
2
 Why do countries decentralize?
 Keeping the country together—political economy objectives
 Improving service delivery; generating sustainable growth
 Tradeoffs between local discretion and minimum standards
 Subsidiarity principles versus political economy of inter-jurisdictional
competition?
 Expenditure management and accountability
 How to prevent “game-play” and leakages?
 What are the underpinnings for transparency, coordination and efficiency?
 How to design coordinated budget frameworks, information systems and
tracking of cash?
 How to prevent PPPs from degenerating into postponement of liabilities?
 Designing and managing transfer systems
 How to ensure effective equalization or generating disincentives through
transfers?
 How can performance based transfers be used without overriding accountability?
 Scope for financing investment needs?
Spending and accountability
3
Making decentralization work: Subsidiarity and
competition?
4
 Can one achieve accountability solely through spending
devolution?
 Is there a clearly defined role for the “level at which
services are delivered”?


Externalities in provision: climate change; health care; national
standards
Do voters care unless they pay, at the margin, for clearly defined
services?
 Accountability needed for effective decentralization


But can voters compare and punish badly performing local
governments (yardstick competition)?
Role of information on budgets, use of funds and outcomes available
to electorates and governments
Modified subsidiarity principles
5
Centralization
needs
• Cooperation
• Coordination
• Harmonization
Factors
• Economies of
scale
• Spillovers
• Congestion
Political
economy
• Equalization
• Interpersonal
equity
Supranational (EU)
• Rules;
regulations
• Common
policies
Provincial
•
•
Financing at margin
Accountability
Decentralization
• Devolution
• Participatory
democracy
• Control
• Accountability
and effective
provision
Political economy
• Yardstick
competition
Local
•
•
Financing at margin
Accountability
Adapted from Dafflon, in Ahmad and Brosio, Handbook of Fiscal Federalism
Whose responsibilities?
6
 Major problems in poorly designed assignments:
 Lack of clarity in assignments (overlapping functional and economic
responsibilities) or excessive earmarking complicate accountability


See Dafflon (2006; Handbook of Fiscal Federalism);
Assignments without:


Adequate financing; or
Ability to implement
 Attempts to “unload” responsibilities in Big Bang, especially
without own-source revenues and adequate financing—may
backfire:




Nigeria: local governments stopped paying primary teachers
Pakistan: inadequate financing for devolved responsibilities—functions
either not carried out, or performed badly
Indonesia: effectiveness of service delivery—incentives for LGs?
Bolivia and Colombia (until recently); overlapping functions
Can Minimum Standards offset Assignment
problems?
7
 Minimum standards represent central objectives
superimposed on local preferences

Should be financed by special purpose transfers
 In a decentralized framework, could use these as part
of the “social contract” leading to local autonomy


Rationalization in terms of “outcomes”: e.g., years of
schooling; immunization standards
Sanctions if not met through shared and equalization funds
 But not detailed input norms (like Soviet Gosplan)
 These are contrary to effective decentralization and reduce
local accountability
Expenditure management and
accountability
8
INFORMATION GENERATION AND
INSTITUTIONS
Incentive structures and accountability
9
 With incomplete information:

Incentives to hide spending and liabilities
Off-budget operations
 SoEs
 PPP may be a problem without proper accounting for liabilities


Weak prudential management of systemic private liabilities
Mexican roads (1990s);
 Spain and Irish property development
 Well within limits set by rules (Spain and Ireland formally
compliant with Maastricht)
 Irrational private spending at local level condoned by local power
elites (Spain, Portugal) comes home to roost

 In both cases, liabilities can be shifted to future
governments; or higher levels—accountability not
guaranteed
Game-play on spending, debt and liabilities?
10
 Hiding spending, assets and liabilities

Reduce deficits in cash (C)or financial assets (F), without affecting all
recognized liabilities (R) or extended net worth based on future flows
(E)




Selling non-financial assets in R, for cash in F
Assuming future pension liabilities in E, for cash and financial
assets in F
Securitization C of future revenue streams F (common in Latin
American local governments
Treating borrowing F as revenue C (several US States)
 Solution: maintain parallel reporting on C, F, R and E
 New accounting and reporting standards for PPPs
(IPSAS32);

Plus the design of contracts, involving firm-level commitments
Rights and obligations associated with all future
cash flows, E
11
All currently
recognized assets
and liabilities, R
Financial assets
and liabilities, F
Cash
FF
C
Need for standardized information:
Who does what and what outcomes?
12
 Clarify functions (UN-COFOG)
 Economic components (GFSM2001)
 Wages, O&M; Capital, etc…
 But distinguish between decentralized and deconcentrated operations
 Outcomes can be added onto the Chart of Accounts
 But distinguish between own-account and deconcentrated
operations on both spending and revenue sides
 Programs and sub-programs as components
 Budget, accounting and reporting critical for “yardstick
competition”
 Focus on outcomes (performance budgeting) useful, but
insufficient without the component building blocks
Spending assignments and Information Requirements
Functions
Economic
WagesClassification
Other
Capital
Current
O&M
Etc…
Central
Departments/
Municipal
Real outcome indicators
Education
a. Primary
cccc
ccccc
cccc M
ccccc
ccccc M
b. Secondary
cccc
ccccc
Cccc
ccccc
Ccccc
a. Basic
Preventive/Clinics
cccc
ccccc
cccc M
cccc
ccccc M
Water
cccc
ccccc
cccc M
cccc
ccccc M
Sanitation
cccc
ccccc
cccc M
cccc
ccccc M
cM
Heath care
Note: “c” represents a central assignment, and M is municipal.
In the cases above, the Municipal government would be responsible for the construction of a school room, or a clinic, without full
responsibility for the primary education or basic preventive care function or sub-function. This makes it easy for the local government to
evade responsibility or accountability for outcomes—as it is quite easy to blame another level for the absence of teachers or health care
personnel.
13
Preventing game-play
14
 Need harmonized medium-term budget frameworks,
for all levels of government
 Standardized framework (GFSM2001), with multiple
criteria, for accounting and reporting
 Follow the cash (TSAs at each level—or
correspondent arrangements)
Follow the cash: why the resistance to TSAs?
15
 Regardless of budgeting model, TSA is a critical element
of good management



Track sources and uses of funds;
Separate “pots” complicate cash management; more importantly
facilitate rent seeking;
International best practice is appropriate
 Even if there is a Central TSA, what about sub-national
entities?

Should there be subnational TSAs? Chinese example
 Role of donor agencies:


Use of correspondent accounts within a TSA?
Zero-balance arrangements?
TSA with donors/local government
16
Chart 1, Local government and
donor use of correspondent
accounts linked to TSA
MOF
Informa on
flo s
Payment
request
C/A 2 w
C/A 1
Payee
Payment Order
Payment request
ZBA
Payment
Designing and Implementing
Transfer Systems
17
Transfers to ensure accountability?
18
 Need for own-revenues at the margin
 Important for incentives
 Does not always require sub-national tax administrations
 Avoid debilitating ‘gap-filling’ transfers
 Minimize earmarked transfers,
 address monitoring and reporting capabilities
 Combination of specific purpose and general
(equalization) grants

See Ehtisham Ahmad (ed.), Financing Decentralized
Expenditures (Elgar, 1997)
Design of transfers
19
 Equalization principles (mainly for current
spending)


Revenue capacities only (Canadian model), or
Also expenditure responsibilities (Australian model)





Cost of service provision
Modeling
Institutional design (grants commission)
Information flows
Need standardized factors—actuals lead to “gap filling”
 Minimum standards in areas of local
competence?

Inimical to decentralization



Could be undertaken directly by center
Financial or physical inputs (e.g., wages or capital); or outputs (e.g.,
number of years of school); or outcomes (e.g., literacy rates).
Selected performance based transfers, especially for capital
spending?
Equalization Frameworks and Institutions
20
 Need to be designed in incentive-compatible manner
 For sub-national entities not to play games:
 Full information should be made public on the
formulations and data used to make the assessments

Independent agency with representation by the
provinces/subnationals
Commonwealth Grants Commission formulations
 Indian Finance Commission


Or government body, under the direction of the Ministry of
Finance?
 Issues of credibility and trust
21
Managing performance-based transfers: need for
information flows and sanctions
22
 Increasing emphasis on “performance-based” transfers,
including by international agencies
 Difficulty if the object of the transfer falls in areas of LG
competence



Diversion of resources; not just transfers, but own revenues
Dilution of accountability
Potential distortions, if outputs rather than outcomes targeted

(standardized tests in US “No Child Left Behind”, while deterioration
in standards
 Could be useful to address externalities—including for
investments

Contract federalism (Spahn, 2006; OECD, 2011)
 Repeated games and credibility of sanctions
23
Central government conditions or objectives
24
 Typical problem: supra-national, or donor country
objectives, sub-national administrations


Capital transfers (DAK in Indonesia)
Standard solution: conditional transfers, often with matching
grant provisions
often ineffective: the poorest regions unable to meet matching
requirements
 With weak information on the uses of funds, scope for diversion of
resources

 Extreme example (infrastructure):

China: grain silos, central government responsibility,
financing channeled through provinces and counties,
 county-level administration


How to ensure that central funds will be used to build silos?
Possible solution
25
 Use local monitoring and sanctions for efficient
outcomes
1. Use standardized information on the sources and
actual uses of funds


Permits the use of inter-jurisdictional competition to sanction
non-performance
And if performance below minimum standard—use of central
sanctions
2. Plus use of performance-based transfer to meet
infrastructure gaps in a cost-effective manner