Lecture 5. Fiscal decentralization
and the incentives of
• In Eastern Europe and the Soviet Union in
particular, very little thought was given to reform
of government administration and how to make it
support reform instead of opposing it. Yeltsin
team managed by decrees from above but gave
little thought to incentives of bureaucrats.
• In China, despite the absence of drastic political
change (the power of the communist party
remained unchallenged), various aspects of
government reform played in the hand
(consciously or not) of market reform.
• Fiscal contracting system between 1980 and 1993
(crucial first period of reforms).
• Provinces signed five year contracts with central
government to give over either a fixed sum (and keep
the rest) or give a share of revenues.
• Guangdong gave a fixed amount per year during 1980-
1987 and then during 1988-1993 an amount that
increased by 9% a year. Many governments retained
100 % of local revenue at the margin.
• These contracts had interesting incentive properties if
they were credible (no ex post renegotiation over
revenues or subsidies).
• After 1994, a new tax system was put in place
giving local government rights over local taxes
(taxes from enterprises other than central
government enterprises, personal income tax
and business sales taxes). Provincial
government kept 25% of VAT. Also important
category of extra-budgetary revenues which are
not shared with central government. There has
thus always been since the reforms a fiscal
arrangement making local government residual
claimants over taxes. Potentially, provincial
government thus had a stake in the
development of the tax base and economic
growth, assuming that a higher tax base benefits
A simple model (Jin et al. 2004)
• Call Y(e) value created by local businesses and e the
reform and pro-business policies and regulations
pursued by local government, dY/de >0. There is
however a cost associated to e, c(e) (convex) since
status quo is not pro-business.
• Call y(Y(e)) the tax revenue generated, dy/dY >0, y
concave positive function of e.
• Local government maximizes zvy(Y(e))-R- c(e) +T
where v is local share of revenues, z is marginal
retention rate of local revenues, R is lump sum
remittances to central government and T is transfers
received from central government.
• First order condition:
• First order condition:
Increase in z
• The more the local government is residual
claimant of increase in revenues, the more
effort it will make towards development of
• Local government has better knowledge
than central government about what helps
• Incentive aspects of fiscal arrangements
important in transition economies.
• We see that the discrepancy between ex
ante contracts and ex post implementation
was not extremely large and declined over
• Suggests credibility.
• Commitment problem more related to
subsidization rather than rent-grabbing,
i.e. soft budget constraints problem more
serious than predatory behavior since
more provinces spend more than specified
in the contract rather than the opposite.
• Increasing correlation over time between
budgetary expenditures and revenue.
Suggests both that provinces were
residual claimants over their revenue and
that they were fiscally responsible.
• Strong contrast with findings of
Zhuravskaya (2000) for Russia.
Chinese and Russian fiscal
• In contrast to the long term fiscal
arrangements in China, in Russia under
Eltsin annual negotiations on tax-sharing
and transfers. Creates ratchet effect and
soft budget constraints. Variation in local
revenues met one for one by variation in
transfers, (Zhuravskaya, 1999)
• Berkowitz and Li (1999) have analyzed
“overgrazing” aspect of Russian taxation.
• Qian-Roland (1998): Competition between local
governments for foreign capital creates
incentives to overinvest in infrastructure and has
the unintended consequence of hardening
• Federalism does not necessarily harden budget
constraints, it may soften them. Governments
may strategically distort composition of
expenditures in order to attract more subsidies.