Transcript Thornton Matheson

Fiscal Federalism and Inequality
in Latin America
Thornton Matheson
International Monetary Fund
Presentation for the LAC Tax Policy Forum
Panama, September 17, 2010
Overview
•
•
•
•
•
Theory of fiscal federalism
Practice of fiscal federalism in Latin America
General reform goals
Local Government (LG) revenue instruments
Design of interjurisdictional redistribution
Fiscal Federalism: Theory
• Where regional preferences differ, devolution
of fiscal authority can improve allocative
efficiency and welfare.
• To achieve this, LGs must be accountable to
local constituents
– Raise own revenues at the margin
– Significant discretion over spending
Fiscal Devolution: Constraints
• Interregional factor mobility
• Local administrative resources
• Macroeconomic stabilization
– LG soft budget constraints
– Excessive devolution
• Regional tax base disparities*
– Economic activity often concentrated in a few
large cities
– Natural resource wealth also unevenly distributed
Fiscal Federalism and
Interpersonal Inequality
• LGs often largely responsible for social welfare
spending (health, education) critical to
reducing inequality
• LGs with poorer constituencies often have
smaller PC tax bases
– Necessitates interjurisdictional transfers
Fiscal Federalism in Latin America
• Expenditure: Relatively low level of devolution
• Revenues: Low level of own revenues
Table 1:
Region
LG Share of Expenditure
LG Share of Revenues
OECD
30.6
22.7
Latin America
19.4
11.7
Source: http://www1.worldbank.org/publicsector/decentralization/fiscalindicators.htm
Data include shared revenues, but exclude transfers
Fiscal Federalism in Latin America
• Mismatch between revenues and
expenditures
• Heavy reliance on transfers and revenue
sharing
– Most major tax bases assigned to center
– Transfers/revenue sharing often pro-cyclical
• Extensive earmarking of transfers
– Many LGs have little discretion over spending
– Vertically overlapping spending responsibilities
FF in LAC: Tax Assignment
• Most major bases assigned to center:
–
–
–
–
–
–
CIT: national
PIT: national (except Brazil and Mexico)
Asset: national
VAT: national (except Brazil)
Sales: national
Excises: national/state
• States and local tax bases more limited:
– Vehicles: state
– Property: state/municipal
– Business taxes: municipal
• Natural Resources: shared
Source: Amhad & Brosio (2008), 7 largest LAC countries
FF in Latin America: Consequences
• Revenue dependency and earmarking
undermine LG accountability
– Poor incentives to develop own revenue sources
– Poor incentives for efficient service provision
• Procyclical revenues and revenue/expenditure
mismatch encourage LG deficit financing
– Soft budget constraints can threaten
macroeconomic stability
FF in Latin America: Reform Goals
• Develop LG own revenue sources
– Low cyclicality
– Progressive
– Evenly distributed across jurisdictions
• Constrain LG deficit spending
• Make transfer system transparent, fair and
reliable
– Reduce earmarking
Revenues: Property Tax
• Advantages:
– Immobile base
– Progressive
– Low cyclicality
• Disadvantages:
– Administration complex
– Interregional base disparities
Revenues: Property Tax
• Chief source of own revenue for LGs in LAC
• However, productivity relatively low:
– OECD average: 2.6% GDP
– Developing countries: 0.6% GDP
•
•
•
•
Mexico: 0.2%
Brazil: 0.5%
Argentina: 0.4%
Colombia: 1%
– Problem: Low tax rates, poor cadastres, administrative
capacity, lack of political will to enforce
Revenues: Consumption Taxes
• VAT, sales taxes, destination-based excises
– E.g., Brazil (ICMS), Colombia (alcohol, cigarettes, gasoline),
Argentina (turnover)
• Advantages:
– Low cyclicality relative to income taxes
– Geographically dispersed
– Destination basis generally avoids tax competition (except
cross-border shopping)
• Disadvantages:
– Subnational VAT complex, distortive
– Sales taxes cascade, avoidable, narrow bases
– Origin-based excises more efficient
Revenues: Income-type VAT
• Advantages:
– Broader base than CIT, buoyant revenue source
– Benefit charge for local businesses
– Alternative to commercial property tax
• Disadvantages:
– Geographic concentration
– Tax competition (origin-based)
– Cyclicality (but less so than CIT)
• Examples:
– IRAP (Italy, regional) raises 2.5% of GDP
– IETU (Mexico, federal) has some similarities
Revenues: PIT Surcharge
• CG sets base, LGs allowed to set own rate (perhaps
within limits)
• Advantages:
– Administrable (CG can collect)
– Progressive
• Disadvantages:
– Cyclicality
– Base mobility
• Some CGs share PIT revenue with LGs (e.g., Brazil,
Mexico).
– Unlike surcharge, this is not LG own revenue.
– US: Many states impose own PITs on federal PIT base.
Revenues: Natural Resource Taxes
• Theory dictates these should be CG taxes:
– Uneven regional tax base
– Potentially complex
– Volatile revenues
• However, they are often shared with LGs.
– Need for infrastructure, environmental cleanup
– Local population ownership claims
• Examples:
– Bolivia: Departments receive about 2/3 of gas rents and 1/3
hydrocarbon income tax (IDH). Municipalities receive ¼ of IDH.
– Colombia: 48% oil royalties go to departments, 13% to municipalities,
8% ports.
• Possible alternative: Production excises with redistribution
– May also create stabilization fund to counter revenue volatility
Interjurisdictional Redistribution
• Given disparities in LG tax bases and
administrative constraints, LAC countries will
likely remain fairly reliant on revenue sharing and
transfers.
• Greater decentralization itself leads to more
transfers
– OECD: elasticity of total transfers with respect to LG
own revenue share = 1.5
• Redistribution therefore needs to be well
designed to promote LG accountability and hard
budget constraints.
Transfer Design
• Transfers should not be based on historical LG revenue
or spending patterns
– This creates incentive for LGs to undertax and/or
overspend.
• Transfers should be based on the difference between
ex ante LG revenue capacity and spending
requirements
– Revenue capacity: Apply average tax rate to average base,
or macroeconomic measure of GRP, income PC
• Transfers could be varied anti-cyclically
– In practice, transfers are often pro-cyclical
– “Rainy day funds”