Sei sulla pagina 1di 24

Economics of Governance (2007) 8: 339362

DOI 10.1007/s10101-007-0032-5
O R I G I NA L PA P E R

Complementarity, fiscal decentralization


and economic growth
Nobuo Akai Yukihiro Nishimura
Masayo Sakata

Received: 18 November 2004 / Accepted: 7 November 2006 / Published online: 23 February 2007
Springer-Verlag 2007

Abstract Theories of the voluntary provision of public goods and development economics have clarified that complementarity in the production process
is a crucial ingredient to understanding how alternative economic environments
affect economic performance. This paper examines how the structures of intraand inter-regional complementarity affect the relationship between economic
growth and fiscal decentralization. We provide a theory that describes how
fiscal decentralization affects economic growth under various structures of regional complementarity. Our empirical analysis, based on a panel data set of

The previous version of the paper was presented at the 59th Congress of the International
Institute of Public Finance (University of Economics in Prague, Prague), the 2003 Fall Meeting of
the Japanese Economic Association (Meiji University, Tokyo), the 60th Annual Meeting of the
Japanese Institute of Public Finance (Kansai University, Osaka), and in seminars at Yokohama
National University and the University of California, Irvine. The authors acknowledge the
comments and discussions by people including Timothy Goodspeed, Kiyoshi Mitsui, Motohiro
Sato, Etsuro Shioji, Tsunao Okumura, and Craig Parsons. We are also grateful for the comments
by the Editor (Amihai Glazer) and two anonymous referees. The usual disclaimer applies.
Nishimura acknowledges the financial support from JSPS (Japan Society for the Promotion of
Science) Postdoctoral Fellowships for Research Abroad.
N. Akai
School of Business Administration, University of Hyogo,
8-2-1 Gakuen-Nishimachi, Nishiku, Kobe 651-2197, Japan
Y. Nishimura (B)
Department of Economics, Yokohama National University and Queens University,
Dunning Hall, Room 209, Kingston, Ontario K7L 3N6 Canada
e-mail: ynishimu@ynu.ac.jp
M. Sakata
Department of Politics, Economics and Law, Osaka International University,
3-50-1, Sugi, Hirakata, Osaka 573-0192, Japan

340

N. Akai et al.

the fifty states of the United States over the period of 19921997, supports our
theoretical specification of the production function. Also, we observe a humpshaped relationship between fiscal decentralization and economic growth that
is consistent with our theoretical result. Our analysis also shows that the optimal degree of fiscal decentralization conducive to economic growth is higher
than the average of the data in some cases, and hence further decentralization
is recommended for economic growth.
Keywords Complementarity Fiscal decentralization Economic growth
JEL classification O40 H77
1 Introduction
One role of the public sector is to create a public organization to promote economic growth. It is well known that the average share of central government
spending in total government expenditure is typically much higher in developing countries than in developed countries. This fact, combined with the argument in many areas of economic theory in favor of decentralized organizations,
has recently encouraged the progress of fiscal decentralization in developing
countries. The problem of centralized versus decentralized governmental organization is also a relevant economic issue in developed countries. For example,
in Japan, there are discussions on the reform of taxes and intergovernmental
grants towards a more decentralized system.
A number of studies attempt to quantify the impact of decentralization by
relating some measure of decentralization to economic growth (for example,
Davoodi and Zou 1998; Zhang and Zou 1998; Xie, Zou and Davoodi 1999; Ebel
and Yilmaz 2002; Lin and Liu 2000; Akai and Sakata 2002). However, underlying
macroeconomic structures that can explain possible relationship between these
variables are not fully examined. Theoretical works are scarce. Recent work by
Sato and Yamashige (2005) introduces a political economy model with a complex principal-agent nature of the government to explain the evolution of fiscal
decentralization and economic development. Extending Barros (1990) economic growth model with a public good, theoretical models of Davoodi and Zou
(1998) and Xie, Zou and Davoodi (1999) demonstrate a theoretically optimal
degree of fiscal decentralization. However, empirical implications of their studies are either not obvious or very difficult to test. This paper presents a theoretical link between fiscal decentralization and economic growth by incorporating a
macroeconomic model that crystallizes the economic structure to explain whether fiscal decentralization is conducive to economic growth or not. We also
empirically examine the validity of our specification of the macroeconomic
structure as well as that of the derived proposition.
A key concept to analyze the mechanism of the effect of fiscal decentralization on economic growth is complementarity. Since policy makers have
limited foresight to find the correct policy to promote growth or individual
welfare, there arise successful and unsuccessful jurisdictions, and also, within

Complementarity, fiscal decentralization and economic growth

341

jurisdictions, there are successful and unsuccessful projects. Public services in


each jurisdiction have a spillover effect on the national economy. The question is how effectively a high quality public service in one jurisdiction can
make up for a poorly performed outcome in another. The less important is the
contribution of successful jurisdictions, the greater the complementarity. An
example of high complementarity between jurisdictions is that of the former
Soviet Union and the East European countries, where planners located industries across regions to maintain the comparative advantage of each region.
An example of low complementarity (high substitutability) is that of China,
where public projects have an experimental nature (Montinola et al. 1996).
The notion of complementarity was first discussed by Hirshleifer (1983) in
the public good provision problem, extended theoretically and empirically by
Cornes (1993), Conybeare et al. (1994) and others (see Cornes and Sandler
1996). Contributions in monopolistic competition (see Matsuyama 1995 for an
overview) and development economics (see, e.g., Bnabou 1996) emphasize its
importance on economic development. This paper examines how the structures
of intra-regional complementarity (referred to as local complementarity) and
inter-regional complementarity (referred to as global complementarity) affect
the relationship between economic growth and fiscal decentralization.
We consider a simple two-period OLG model where a country consists of J
jurisdictions, which need public services. There is one bureaucrat in the central
government, and one local bureaucrat in every jurisdiction. The outcomes of
public services provided by the central and local governments depend on the
ability of bureaucrats. We show the following. If the degree of local complementarity is lower than global complementarity, the relationship between fiscal
decentralization and expected economic growth is hump-shaped, and there is a
unique interior optimum that is consistent with the expected growth maximization. An intuition is the following. The risk that the bureaucrat in charge of the
public input may have low productivity can reduce economic growth. The question is how a bad project is made up with the good project in the aggregation
process. If the local complementarity is sufficiently lower than global complementarity, risk-sharing between local governments and the central government
becomes more desirable.
We then examine these theoretical results using a panel data set of the
fifty states of the United States over the period of 19921997.1 We use annual
1 Davoodi and Zou (1998) use cross-country data for 46 countries over the period of 19701989,

in which the cultural, historical, and institutional differences between countries are substantial. Given other determinants of growth, they found a negative relationship between fiscal
decentralization and economic growth. However, Davoodi and Zou (1998, p. 254) recognize
that the cross-country variation among developing countries is one of the reasons for the negative relationship. Ebel and Yilmaz (2002, footnote 16) noted that overrepresentation of the
degree of fiscal decentralization in the aggregate figure for developing countries might be
the reason for the negative relationship in Davoodi-Zous study. Ebel and Yilmaz also replicated Davoodi-Zous regression model for the six European countries (the Czech Republic,
Estonia, Hungary, Latvia, Lithuania, and Poland) where cultural and historical differences are
much smaller, in which they derive a positive relationship between fiscal decentralization and
economic growth. A similar contrast can be found between Xie, Zou and Davoodi (1999)

342

N. Akai et al.

data and the average of the sample period. We first show that our theoretical
specification of the production function is supported by the regression where we
derive statistically significant values of global and local complementarity. Also,
we observe a hump-shaped relationship between fiscal decentralization and
economic growth that is consistent with our theoretical result. An intriguing
issue is whether the optimal level of fiscal decentralization derived by our
econometric model is higher or lower than that of the actual data. In some cases,
the optimal degree of fiscal decentralization conducive to economic growth
is higher than the average of the data, and hence further decentralization is
recommended for economic growth.
To our knowledge, this is the first work which discusses the hump-shaped
relationship between fiscal decentralization and economic growth theoretically,
with consistent empirical analyses. Thiessens (2003) empirical work shows the
hump-shaped relationship between fiscal decentralization and economic growth
in the group of high-income OECD countries. The differences between his study
and this paper are as follows. First, we provide a theoretical model that explains
a relationship between fiscal decentralization and economic growth. Second, we
estimate the production function consistent with theoretical hypothesis. And
finally, we examine whether the current level of fiscal decentralization is optimal
or not.
The rest of this paper proceeds as follows. In Sect. 2 we provide an economic
growth model with complementarity in the public expenditure in an overlapping
generations setting, and examine expected GDP growth in relation to fiscal
decentralization. In Sect. 3 we empirically verify our theoretical result using US
cross-section data. Section 4 concludes. The proof of the proposition is given in
Appendix A. In Appendix B we provide the data sources.
2 A model with complementarity and economic growth
2.1 Firms and households
We consider an extension of the standard two-period OLG model with many
jurisdictions, adopted in Nishimura (2006). A country consists of J identical
jurisdictions. In each jurisdiction there are many identical firms and households.
Footnote 1 continued
and Akai and Sakata (2002). Xie, Zou and Davoodi (1999), using time series data for the United
States from 1948 to 1994, find that fiscal decentralization may be detrimental to economic growth.
Akai and Sakata (2002) point out that Xie, Zou and Davoodi (1999) used the data from a period
of high economic growth in the United States. During the early stages of economic development,
a high level of government expenditure was required to provide public investment that generates
large externalities, so that the analysis that includes such a period overestimates the contribution
of the central government. Akai and Sakata (2002) also argue that the estimation has to be done
with data which embody less cultural and historical differences and have similar development stage.
Taking account of these factors, Akai and Sakata (2002) find that fiscal decentralization contributes
to economic growth. For the same reason as in Akai and Sakata (2002), we use the data from the
period of 19921997.

Complementarity, fiscal decentralization and economic growth

343

Firms use capital K, labor L and the public good G to produce the output Y.
The production function takes the familiar form studied by Barro (1990) and
others:
1

Yt = AKt

(Gt Lt )

(1)

where Yt is the national output, Kt is capital, Lt is labor, and Gt is the public


good supplied in period t. As specified below, Gt is a function of public services provided by the central government and the local governments. Examples
include health, literacy and the average IQ level of the citizens, which depend
on public expenditures such as health and education. Following a convention in
growth theory, we assume that the public good G serves as a labor-augmenting
technology.
There are identical households who live for two periods. When agents are
young, they supply inelastically one unit of labor and receive the competitive
wage level wt , which is taxed at the rate .2 A fixed fraction is saved out of the
post-tax income in period t,3 which is consumed at the next period when agents
are old without bequest. Following the conventional models, initial capital stock
K1 is owned by the L0 persons who are old in period 1.

2.2 Public good


We now introduce Gt . The public projects are undertaken by different bureaucrats. Some policies are favorable for higher economic growth, and others are
not. The level of public good Gt is determined by the local and nation-wide distribution of the public inputs through inter- and intra-jurisdictional spillovers,
which we call complementarity. A crucial question for aggregation is whether
a good project in the nation pulls the national product up, or a bad one drags
the nation down. Our concern is captured by the following formulation as an
extension of Nishimura (2006).
Each jurisdiction provides a public service, such as infrastructure, education,
or regulation policies. There is a continuum of projects in each jurisdiction j.
Let gt (j, m), m [0, 1] M be project m in jurisdiction j at period t. The public
good Gt is a function of these public projects.
Tasks are divided between the central bureaucrat and the local bureaucrat. There are two types of bureaucrats who can provide public services
either with high (h) or low (l) quality. A proportion of bureaucrats has high
2 Capital income tax can be incorporated in the model, but such complication is not essential to

the analysis.
3 A constant fraction of savings is similar to the conventional Solow-Swan growth model, but

it can also be derived from a micro-founded model where households have a logarithmic utility
function.

344

N. Akai et al.

ability. When a project m is assigned to the central government, one bureaucrat, who can be type h with probability , is in charge of public project gt (j, m)
for all j. When a project m is assigned to a local government, one bureaucrat in a local government in each jurisdiction, who can be of type h with
probability , is in charge of one public service, gt (j, m). The quality of each
public service in period t is a random variable, depending on the bureaucrats
ability.
At each period, the local bureaucrats in each jurisdiction and the central
bureaucrats are randomly and independently chosen. At the end of period t,
the bureaucrat who is assigned to jurisdiction js project m decides on the
project, which turns out to be good or bad in the beginning of the next period
(period t + 1) according to the ability of the bureaucrat in charge. Each project
is financed by the tax revenue. Let Tt be the tax revenue per jurisdiction at
period t, and let Tt (j, m) be the amount of public funds to finance jurisdiction
1
js project m at period t. The fiscal budget is balanced: 0 Tt (j, m)dm = Tt
for all j. Tasks are symmetric, so that an equal proportion of tax revenue is
allocated to each project: Tt (j, m) = Tt for all j and m. The bureaucrat in charge
of jurisdiction js project m uses the public fund Tt (j, m) to provide the public
service of the next period, gt+1 (j, m). The bureaucrats employ the following
linear production functions.

gt+1 (j, m) = Tt (j, m) if the bureaucrat has a high ability, and


gt+1 (j, m) = Tt (j, m) if the bureaucrat has a low ability, with > > 0. (2)
The parameters and represent the efficiency of the public service. We assume
that the values of (the proportion of high ability bureaucrats), and [the
parameters for the production function (2)] are the same for central and local
governments.4

4 Our formulation is an application of Sahs (1991) fallibility. Bureaucrats cannot perfectly choose

policies for a desirable outcome. As a result, the outcome of policies becomes stochastic. A society is
called more centralized or less centralized depending on whether there are few or many decisionmakers (which is slightly different from our setup). The aggregate performance of all fallible decisions is different, depending on the degree of decentralization. One may argue that the central
bureaucrats perform better than the local bureaucrats. On the other hand, there is a counterargument that lower-tier government can respond better to local needs than higher-tier governments.
For example, Crewson (1995) used the score of the Armed Forces Qualifying Test (AFQT) as an indicator of quality of the workers. He showed that the U.S. federal employees hired during the 1980s
have scores of the AFQT that are significantly higher than those of private sector employees, whereas there is no significant difference in the state sector. On the other hand, Bardhan (2002) reports
several empirical findings in developing countries that public services by municipalities respond
better to local needs than those by the central government. The focus of our analysis is to examine the economic effects of fiscal decentralization (represented by higher allocation of resources
to the lower-tier government) without assuming a priori which bureaucrats perform better on
average.

Complementarity, fiscal decentralization and economic growth

345

We then consider the following aggregation process that embodies our


motivation of the paper:

1

gt (j) gt (j, m) dm

Gt =

0
J

1
j=1

gt (j)

(3)

(4)

(3) represents the public input in jurisdiction j as the composition of


public projects within the jurisdiction. The value of corresponding to
Bnabou (1996) local complementarity, represents the complementarity between the projects within a jurisdiction (among the different projects, or among
the same project accomplished by different bureaucrats). For example, the overall educational level in a jurisdiction depends on the quality of the education
provided by national and local schools. A higher indicates higher complementarity.5 Intuitively, the unsuccessful local project drags the aggregate performance down. Consider, for example, infrastructure in a country where the
regional production process is highly complementary. If the quality of one road
in a jurisdiction is poor, it reduces the efficiency of production in the jurisdiction.6 On the other hand, substitutability becomes higher when public projects
have an experimental nature, where the level of technology is defined by the
advances made by the most successful jurisdiction (Cornes and Sandler 1996,
pp. 5455).7
(4) represents the national level of the public good as a function of jurisdictional public inputs. The value of represents the degree of global complementarity
between public services in the terminology of Bnabou (1996).
The production functions (3) and (4) resemble the social composition function by Hirshleifer (1983) and its generalization by Cornes (1993) Cornes and
Sandler (1996) which discuss complementarity and substitutability in the public good provision. The same formulation is found in Matsuyama (1995) and
Bnabou (1996) which explain the mechanism behind economic development.
It turns out that the value of relative to is central in determining the pattern
of expected growth in relation to fiscal decentralization.
5 The limiting case of high complementarity, g (j) = min g (j, m) (where = ), is referred to as
t
m t
the weakest-link by Hirshleifer (1983).
6 In the former Soviet Union and the East European countries, industries are located according

to a comparative advantage of each region, and the industrial linkage has a highly complementary
structure (see also Prudhomme 1995; Montinola et al. 1996).
7 The limiting case of high substitutability, g (j) = max g (j, m) (where = ), is referred to as
t
m t

the best-shot by Hirshleifer (1983). See also Cremer et al. (1996). See Montinola et al. (1996) for
the case of China.

346

N. Akai et al.

In each jurisdiction, the local government undertakes the projects in the


interval [0, ], and the central government undertakes those in [ ,1]. The value
of (hence the proportion undertaken by the local government) is assumed to
be the same in every jurisdiction. Since the projects are symmetric, the value of
also indicates the spending share of the local government.8
2.3 Economic growth
Due to the uncertainty in the productivity of the bureaucrats, the growth of
the national output is stochastic. Let E denote an expectation operator.
We can

Yt+1
derive the following result regarding expected GDP growth, E ln Yt :
 Y

t+1

E ln Y

t
> 0,
Proposition Suppose that < min {, 0}. Then

=0
 Y

 Y
t+1

t+1
E ln Y
2 E ln Y

t
t
< 0 and
< 0 for all [0, 1]. That is, the relationship

=1
between fiscal decentralization and expected economic growth is hump-shaped:
a unique interior solution maximizes expected growth.
The proof of the proposition is given in Appendix A.
The point here is the interaction of two effects, which are risk-sharing between jurisdictions (represented by ) and risk-sharing between central and
local governments (represented by ). When (the value of local complementarity) is negative, risk-sharing between the central government and the local
government within a jurisdiction is beneficial for expected growth, since good
projects within a jurisdiction dominate in a local aggregation process (gt (j) in
(3)). On the other hand, this vertical risk-sharing brings another risk, namely,
a correlated risk across jurisdictions undertaken by the central government. In
this model, the performance of the central governments public project is more
volatile than that of the local governments as a whole, since the success/failure
of the performance of the public project depends on the ability of the central
bureaucrat. A good project promotes national production as a whole, whereas
a bad project drags a country in the opposite direction. On the other hand, if
a project is undertaken at the local level, there is a kind of risk diversification
of the performance across different bureaucrats. This risk diversification by
decentralization is conducive to economic growth when (the value of global
complementarity) is negative, and vice versa when > 0 (Nishimura 2006).
When < , the first effect dominates the second, so that an interior optimum
maximizes expected growth, and the relationship between fiscal decentralization and economic growth is concave.
The case of < min{, 0} is only a representative case which is consistent
with our empirical analysis in the next section. Various patterns can emerge,
8 Recall that T (j, m) = T for all j and m. Then the amount
t
t

Tt is spent by the local bureaucrat.

m[0,] Tt (j, m)dm = Tt

m[0,] dm =

Complementarity, fiscal decentralization and economic growth

347

Table 1 Relationship between fiscal decentralization and expected economic growth


\

< min{, 0}

=0

> max{, 0}

<0
>0

Hump-shaped, interior optimum


Hump-shaped, interior optimum

Positive and monotonic


Negative and monotonic

U-shaped, = 1 is optimal
U-shaped, = 0 is optimal

depending on the values of and . For example, when < 0 and is close
to 0, the second effect mentioned above dominates, and there is a positive
monotonic relationship between fiscal decentralization and economic growth.
With being close to 0 and > 0 (global complementarity), there is a negative
monotonic relationship.9 On the other hand, when > 0 and sufficiently high,
the relationship between fiscal decentralization and economic growth tends to
be U-shaped, since the first effect mentioned above is reversed.10 In order to
examine which possibility applies for a particular country, we need to examine
the values of and , i.e., how much higher the degree of local substitutability
is compared with global substitutability (Table 1).
Theoretical models of Davoodi and Zou (1998) and Xie, Zou and Davoodi
(1999) demonstrate that there exists a theoretically optimal degree of fiscal
decentralization to maximize the economic growth rate. Their theoretical model
presumes the productivity of each governments spending, which is proportional
to the optimal spending share. However, they did not estimate the production
function. On the other hand, our model illuminates the relationship between
fiscal decentralization and economic growth from alternative complementarity
structures.

3 Empirical analysis with US cross-section data


In this section, in order to empirically verify our theoretical results, we examine
the relationship between fiscal decentralization and economic growth by using
state-level cross-section data for the United States. As outlined in footnote
1, Akai and Sakata (2002) argue that the data used have to be free from the

9 A sufficient condition for positive monotonic relationship

E ln

Yt+1
Yt

and < 0. A sufficient condition for negative monotonic relationship


> 0 for all

E ln Yt+1
Yt

is = 0


< 0 for all

is = 0 and > 0. More generally, for all < 0, there is a critical value of , denoted by (),
where the expected growth is an increasing function of for all [ (), 0]. The () satisfies
< ()  0 for all < 0. There is also a value () such that ()  0 for all , where the
expected growth is a decreasing function of for all [ (), 0].
Y

10 The relationship is U-shaped iff

t+1
2 E ln Y
t

> 0 for all . Sufficient conditions for


2
U-shapedness are > max{, 0}, which is complementary to the case of the proposition.

348

N. Akai et al.

early stage of economic development and do not embody cultural and historical
differences. Taking account of these points, we use a panel data set of the 50
states of United States over the period of 19921997.11 We have collected the
data from the USA COUNTIES 1998 (CD-ROM), the Statistical Abstract of
the United States, and State and Metropolitan Area Data Book 199798 (CDROM).
3.1 Estimation of the production function
We first estimate the production function (1). Due to the limitations of the
data, we use the values of 1992 to estimate the production function. We add a
subscript i to indicate state i, and in this subsection, we omit the subscript t. The
logarithmic expansion of (1) is:
LGSPi = B + (1 ) LKi + (LGi + LLi ) ,

(5)

where LGSPi is the logarithm of the per capita Gross State Product of state i,
LKi is the logarithm of the per capita private investment of state i, LLi is the
number of workers per capita in state i. Each is the value in the year 1992. The
value of LGi corresponds to the logarithmic form of the regional production
function (4).
We construct LGi in the following way. Let i be the local expenditure
share in the total budget of the governments (the ratio of local government
expenditure to combined state and local government expenditure). Let li (j) be
the contribution by the local bureaucrat and ci (j) be that of a central bureaucrat
in jurisdiction j, respectively. Then (3) implies:
1

gi (j) = (li (j) i + ci (j) (1 i )) .12

(6)

In our empirical specification, li (j) represents the fiscal expenditure of the


jurisdictional government j in state i per capita, and ci (j) represents the fiscal
expenditure of the state government per capita in state i divided by the number
of jurisdictions (counties) in state i. Each of them is defined by public expenditure minus payroll to the public workers divided by i and 1 i respectively to
represent the expenditure per project. This value of net expenditure captures
the efficiency of the public service, i.e., the values of and in (2), since a high
11 Washington DC is not included.

12 The local bureaucrat transforms the public projects m [0, ] to
dm =
m[0,] gt+1 (j, m)



{Tt (j, m)} dm = (Tt )


with probability , and
m[0,] dm = (Tt)

 m[0,]
dm =
dm = (T )
with prot
m[0,] {Tt (j, m)}
m[0,] dm = (Tt )
m[0,m] gt+1 (j, m)

bability 1 . The same rearrangement applies for the projects by the central government. Substitution of the rearranged terms into (3) derives (6).

Complementarity, fiscal decentralization and economic growth

349

value of the net expenditure implies efficiency of the jurisdictional intermediate


public good.
We then consider the following equation:

LGi = ln

Ji

1
j=1

pi (j)gi (j)

(7)

where Ji is the number of the jurisdictions in state i. Since the number of


counties/provinces in each state differs across states, we have to make some
adjustment, where pi (j) is a weighting parameter. If we set the number of J as
the lowest common multiple of the number of the jurisdictions in the fifty states,
then, setting pi (j) = J/Ji , the equation becomes equivalent to the logarithmic
form of (4). However, such a number is too large for a usual econometric
software to accommodate. To circumvent this problem, we set J to 1000, an
arbitrary high number, and use the following method. First, arrange the numbers
such that j1  j2 iff|gi (j1 ) j gi (j)/Ji |  |gi (j2 ) j gi (j)/Ji |, that is, to assign a
lower number to a jurisdiction which has a value of gi (j) closer to the average
of gi (j)s in state i. Let xi J [J/Ji ]Ji , where [J/Ji ] is the Gauss number of
J/Ji . The weight pi (j) satisfies:
(a) pi (j) = [J/Ji ] + 1 if 1  j  xi ,
(b) pi (j) = [J/Ji ] otherwise.
That is, the weighting system, which satisfies j pi (j) = J, is to lower the impact
of the jurisdictions further from the average of the jurisdictional expenditure
in the state. This is called Case (i). To test the robustness, we also consider
the polar case where condition (a) above is replaced with pi (j) = [J/Ji ] + 1 if
Ji xi + 1  j  Ji , that is, if gi (j) is further from the average of gi (j)s in state i.
This is called Case (ii).
(6) and (7) are substituted into (5) to estimate the production function.
We use the nonlinear least squares method. Our estimation result is shown in
Table 2.
The null hypothesis of = 0 is rejected with 1% significance level. In addition, we test the null hypothesis of = 0 by Wald test. The 2 (1) statistics
are 7.937 (P value = 0.005) and 7.866 (P value = 0.005), respectively, so that
the value of is significantly negative. Therefore, the parameter values of

Table 2 Estimation of the production function


Case (i)

Case (ii)

Parameter

Estimate

t-statistic

P-value

Parameter

Estimate

t-statistic

P-value

1.19521
0.39793
0.548132

4.54942
3.45696
9.41181

[0.000]
[0.001]
[0.000]

1.19792
0.404898
0.548103

4.5428
3.67224
9.41356

[0.000]
[0.000]
[0.000]

350

N. Akai et al.

and satisfy the assumptions of the proposition.13 In the next section, we


empirically verify our theoretical results.
3.2 Growth regression model
To test our theoretical results, a stylized growth regression model is augmented
with variables representing (the degree of fiscal decentralization). We use
a panel data set of the 50 states of United States over the period of 1992
1997. The annual growth rate of Gross State Product (GSP) of state i at time t,
GSPit , is regressed upon, along with other control variables, a measure of
fiscal decentralization, it . We assume a quadratic functional form GSPit =
of the production
a + b1 it + b2 (it )2 . From the proposition and our estimation

Yt+1


E ln Y
t
|=0 > 0 , and
function, the estimate of b1 should be positive since

 Y


t+1
2 E ln Y
t
the sign of b2 should be negative since
<
0
for
all

[0,
1]
, with
2




Yt+1

E ln Y
t

0 < b1 /(2b2 ) < 1 since


<
0
.

=1

The indicator of fiscal decentralization is defined in accordance with the


fiscal authority that the central/local government has.14 Following the literature, we adopt two indicators of fiscal decentralization: (1) the local expenditure share in the total budget of the governments (the ratio of local
government expenditure to combined state and local government expenditure);
and (2) the local revenue share in the total budget of the governments (the ratio
of local government revenue to combined state and local government revenue).
Both can correspond to the value of in the theoretical model, but they differ in reality because of the intergovernmental grants. The first indicator can
represent the local governments authority when they have an influence over
their expenditure, that is, the amount of taxes to be collected and the type of
expenditure to be made, implicitly assuming that all intergovernmental grants
are non-matching or lump-sum grants. In calculating the expenditure share,
the measure includes all grants from other governments. The second indicator
represents the local governments authority when the government collecting
revenue also has authority over its own revenue, that is, the tax amount to be
collected and the type of expenditure to be made, assuming that all intergovernmental grants are conditional or matching grants. In calculating the revenue
share, we use the own revenue that does not include any grants from other
13 There are two additional remarks. First, the value of B is not reported here, since it is not central

to our study. The values are not significant in both cases. Since this is not a dynamic growth analysis,
the implication for the Solow residual is not entirely clear. Second, estimations are done using a
more generalized model of LGSPi = B+1 LKi + 2 LGi + 3 LLi . The conclusion of < min{, 0}
is robust.
14 In Hoxby (1999), for example, the centralized system is defined as a system where the state
collects and allocates revenue among districts (p. 21).

Complementarity, fiscal decentralization and economic growth

351

governments. Both indicators are commonly used in research on fiscal decentralization, for example, Xie, Zou and Davoodi (1999) and Wallis and Oates
(1988).15,16
The other control variables are chosen on the basis of previous empirical studies of economic growth (for example, Barro and Sala-i-Martin 1995;
Davoodi and Zou 1998). These variables are annual growth rate of the population (POP), the logarithm of the initial GSP per capita (GSP), the initial
education level (EDUC), the initial Liberal vs. Conservative tendencies measured by the initial share of the seats in the state legislature held by Democrats
(LIB vs. CON), the initial Gini coefficient (GINI), the initial percentage of patents (PATENTS), and the initial percentage of export in GSP (OPENNESS).
Detailed descriptions of the data are given in Appendix B. We use explanatory
variables measured at each initial fiscal year of GSPit to deal with potential
endogeneity problems, with the exception of POP.
Our growth equation is the following:
GSPit = a + b1 it + b2 (it )2 + b3 POPit + b4 GSPit + b5 EDUCit
+ b6 LIBvsCONit + b7 GINIit + b8 PATENTSit
+ b9 OPENNESSit + it

(8)

To test robustness, we examine the versions that include time and regional
dummies, and exclude some explanatory variables. We also examine the case
where the dependent variable is the average growth rate of the sample period.
The maximum likelihood method is used for estimation.
3.3 Results
The results are shown in Tables 3 and 4. Tables 3 and 4 present the results
when we adopt revenue and expenditure respectively as a fiscal decentralization
indicator.
The first column of estimated results is for a model that includes all control
variables and a constant term. Regardless of the definition of fiscal decentralization, the estimated value of b2 is significantly negative and that of b1 is
significantly positive in both tables, with 0 < b1 /(2b2 ) < 1. The results are
consistent with the proposition. We will examine the value of b1 /(2b2 ), the
estimated value of the optimal degree of fiscal decentralization, in the next
section.
The second column shows the regression result with only a fixed time effect.
The introduction of a fixed time effect does not affect the results that b2 is
15 Xie, Zou and Davoodi (1999) use the expenditure share as an indicator of fiscal decentralization, while Wallis and Oates (1988) use expenditure and revenue shares as indicators of fiscal
decentralization. See also Akai and Sakata (2002, pp. 9798), for a recent empirical analysis.
16 The result of the production function estimation in Sect. 3.1 is invariant when we use the revenue

indicator for i under several alternative specifications.

No
Yes

Yes
No

Yes
No

0.03
[2.36]
0.14
[3.06]***
0.15
[2.63]***
0.34
[1.60]
0.0001
[0.46]
0.16
[2.29]**

994.91
0.47

Yes
Yes

0.005
[0.36]
0.10
[2.29]**
0.11
[1.97]**
0.28
[1.54]
0.0001
[0.95]
0.08
[1.16]
0.01
[1.37]
0.03
[1.00]
0.02
[0.75]
0.003
[0.02]
1,052.82
0.45

Yes
Yes

0.01
[1.15]
0.12
[3.05]***
0.13
[2.52]**
0.30
[1.77]*
0.0002
[1.06]
0.08
[1.37]

1,051.21
0.46

No
No

0.01
[0.55]
0.24
[2.10]**
0.25
[1.55]
0.11
[0.27]
0.001
[3.05]***
0.04
[0.40]

171.11
0.48

No
Yes

0.03
[1.64]
0.12
[2.30]**
0.14
[2.09]**
0.35
[1.44]
0.0000
[0.25]
0.15
[1.99]**
0.003
[0.62]
0.01
[0.16]
0.01
[0.50]
0.01
[0.42]
995.37
0.43

No
No

0.01
[1.54]
0.14
[3.84]***
0.15
[3.11]***
0.10
[0.89]
0.0002
[2.05]**
0.06
[1.67]*

1,044.26
0.47

0.001
[0.11]
0.12
[2.94]***
0.14
[2.44]**
0.06
[0.48]
0.0003
[2.21]**
0.02
[0.39]
0.01
[1.50]
0.02
[0.83]
0.002
[0.11]
0.004
[0.32]
1,045.76
0.43

(7)

0.02
[1.61]
0.15
[3.09]***
0.17
[2.68]***
0.08
[0.47]
0.0001
[0.77]
0.08
[1.35]
0.002
[0.43]
0.01
[0.36]
0.01
[0.41]
0.01
[0.62]
988.41
0.44

(6)

Figures in parentheses are the absolute values of t-statistics. Asterisks indicate variables whose coefficients are significant at the 10%(*), 5%(**) and 1%(***)
levels, respectively. The sample size is 300 or 50. Due to limits on space, we do not report the results for the estimated coefficients of the individual dummy
variables in the table. We present fifty states divided into eight areas as follows. New England = Connecticut, Maine, Massachusetts, New Hampshire, Rhode
Island, Vermont; Mideast = Delaware, Maryland, New Jersey, New York, Pennsylvania; Great Lakes = Illinois, Indiana, Michigan, Ohio; Plains = Iowa, Kansas,
Minnesota, Missouri, Nebraska, North Dakota, South Dakota, Wisconsin; Southeast = Alabama, Arkansas, Florida, Georgia, Mississippi, Kentucky, Louisiana,
North Carolina, Oklahoma, South Carolina, Tennessee, Virginia; Southwest = Arizona, New Mexico, Texas; Rocky Mountian = Colorado, Idaho, Montana,
Utah, Wyoming; Far West = Alaska, California, Hawaii, Nevada, Oregon, Washington

Log likelihood
Optimal value of fiscal decentralization
Variables included in growth equation:
Regional fixed effects
Time fixed effects

Openness

Patents

GINI

LIB versus CON

EDUC

GSP

POP

Decentralization2

Decentralization

Constant

(5)

(8)

(4)

(2)

(1)

(3)

Dep. ver: average growth

Dep. ver: annual growth

Table 3 Regression results for the revenue indicator

352
N. Akai et al.

No
Yes

Yes
No

Yes
No

0.05
[2.72]***
0.20
[2.65]***
0.20
[2.24]**
0.21
[0.98]
0.0000
[0.05]
0.13
[1.77]*

997.97
0.50

Yes
Yes

0.02
[0.87]
0.15
[1.94]*
0.15
[1.64]*
0.20
[1.07]
0.0002
[1.31]
0.05
[0.77]
0.004
[0.86]
0.03
[1.19]
0.02
[0.69]
0.004
[0.30]
1,054.06
0.50

Yes
Yes

0.03
[1.48]
0.17
[2.40]**
0.17
[2.03]**
0.19
[1.11]
0.0002
[1.81]*
0.05
[0.74]

1,052.90
0.50

No
No

0.016
[0.59]
0.36
[2.83]***
0.34
[2.36]**
0.10
[0.28]
0.001
[3.25]***
0.12
[1.50]

174.50
0.53

No
Yes

0.04
[2.16]**
0.19
[2.26]**
0.19
[1.92]*
0.22
[0.92]
0.0000
[0.01]
0.13
[1.71]*
0.000
[0.09]
0.01
[0.31]
0.005
[0.18]
0.003
[0.21]
998.06
0.50

No
No

0.03
[1.99]**
0.22
[3.11]***
0.23
[2.77]***
0.02
[0.20]
0.0003
[2.57]***
0.03
[0.91]

1,046.36
0.48

0.02
[1.26]
0.21
[2.90]***
0.22
[2.58]***
0.02
[0.15]
0.0002
[2.15]**
0.01
[0.22]
0.003
[0.94]
0.03
[1.53]
0.01
[0.27]
0.01
[0.58]
1,047.79
0.48

(7)

0.05
[2.61]***
0.26
[3.33]***
0.27
[2.96]***
0.03
[0.17]
0.0001
[0.57]
0.07
[1.27]
0.001
[0.25]
0.01
[0.22]
0.01
[0.40]
0.004
[0.27]
992.63
0.48

(6)

Figures in parentheses are the absolute values of t-statistics. Asterisks indicate variables whose coefficients are significant at the 10%(*), 5%(**) and 1%(***)
levels, respectively. The sample size is 300 or 50. Due to limits on space, we do not report the results for the estimated coefficients of the individual dummy
variables in the table. We present fifty states divided into eight areas as follows. New England = Connecticut, Maine, Massachusetts, New Hampshire, Rhode
Island, Vermont; Mideast = Delaware, Maryland, New Jersey, New York, Pennsylvania; Great Lakes = Illinois, Indiana, Michigan, Ohio; Plains = Iowa, Kansas,
Minnesota, Missouri, Nebraska, North Dakota, South Dakota, Wisconsin; Southeast = Alabama, Arkansas, Florida, Georgia, Mississippi, Kentucky, Louisiana,
North Carolina, Oklahoma, South Carolina, Tennessee, Virginia; Southwest = Arizona, New Mexico, Texas; Rocky Mountian = Colorado, Idaho, Montana,
Utah, Wyoming; Far West = Alaska, California, Hawaii, Nevada, Oregon, Washington

Log Likelihood
Optimal value of fiscal decentralization
Variables included in growth equation:
Regional fixed effects
Time fixed effects

Openness

Patents

GINI

LIB versus CON

EDUC

GSP

POP

Decentralization2

Decentralization

Constant

(5)

(8)

(4)

(2)

(1)

(3)

Dep. ver: average growth

Dep. ver: annual growth

Table 4 Regression results for the expenditure indicator

Complementarity, fiscal decentralization and economic growth


353

354

N. Akai et al.

significantly negative and b1 is significantly positive. To examine the robustness


of the results in the second column, we exclude four control variables, LIB
versus CON, GINI, PATENTS and OPENNESS in the third column. Even in
this case, similar results hold. We next introduce a regional fixed effect.17 We
divide the fifty states into eight different geographical areas (New England,
Mideast, Great Lakes, Plains, Southeast, Southwest, Rocky Mountain, and Far
West; see the description in the table for a detailed explanation). The fourth
and fifth columns present the results with only a fixed regional effect, and the
sixth and seventh columns show the results with both fixed regional and time
effects. The coefficients of b1 and b2 show the same statistical properties as in
the previous cases.
The eighth column shows the result where the dependent variable is the
average growth rate of the sample period, and all explanatory variables are
those of 1992. Except for the fact that the significance of b2 in Table 3 is weaker,
our result is robust.
Next, turning to other characteristics of coefficient estimates, some conclusions emerge from the results in Tables 3 and 4. First, we have controlled
the quality of regional human capital by using two variables (PATENTS and
EDUC). EDUC has the expected positive effect, which is statistically significant
in several regressions, while the effect of PATENTS is not. Thus we evaluate
the results that a higher level of education enhances economic growth in the
region. Second, the estimated coefficient of GSP is negative and significant in
regression models (2), (3), and (8) in Table 3, and (2), (3), (7) and (8) in Table 4,
which is consistent with a conventional convergence property.
In this section, we examine our theoretical results with the use of the conventional growth regression. Consistent with our theoretical result applied to the
estimated parameters of and which are statistically significant, we show
convincing evidence of a hump-shaped relationship between fiscal decentralization and economic growth. This result is robust when we use the revenue and
expenditure indicator as a variable of fiscal decentralization.
3.4 Optimal degree of decentralization
In our model, the growth of the GSPit is maximized when it = b1 /(2b2 ).
The values of b1 /(2b2 ) are 4348 % in the case of the revenue indicator, and
4853% in the case of the expenditure indicator. An intriguing issue is whether
they are higher or lower than the actual data. The average of the data is 0.38
in the case of the revenue indicator (that is, on average, thirty eight percent of
the total government revenue goes to the local governments) and 0.44 in the
case of the expenditure indicator. The hypothesis that the optimal degree of
17 It should be noted that a state fixed effect is highly correlated to fiscal decentralization measure
in each state because there is little variation of fiscal decentralization measure across time. This
creates a multicolinearity problem, which causes distortion in estimating the real effect of fiscal
decentralization. In order to avoid this problem, we adopt regional dummies, instead of state
dummies.

Complementarity, fiscal decentralization and economic growth

355

Table 5 Optimality of fiscal decentralization


Model

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Revenue indicator

3.451
[0.06]*
2.424
[0.12]

3.178
[0.07]*
2.051
[0.15]

4.130
[0.04]**
2.406
[0.12]

1.961
[0.16]
1.468
[0.23]

2.941
[0.09]*
1.970
[0.16]

2.091
[0.15]
1.156
[0.28]

3.089
[0.08]*
1.694
[0.19]

1.006
[0.32]
1.711
[0.19]

Expenditure indicator

Null hypothesis is defined as: optimal value = average value. Figures in parentheses are p-values

fiscal decentralization is significantly higher than the data average is tested by


the statistical significance of the null hypothesis of b1 /(2b2 ) = the average
of the data, i.e., the existing spending shares (or the revenue shares) for local
governments have been consistent with growth maximization. We use the Wald
statistics of the null hypothesis.18 Results are shown in Table 5. In the case
of the revenue indicator, the null hypothesis is rejected in case (3) with 5 %
significance level, and cases (1), (2), (5), and (7) with 10% significance level. In
the case of the expenditure indicator, the null hypothesis cannot be rejected.
Our results show that further revenue decentralization is conducive to economic growth, but the expenditure share is consistent with maximal economic
growth. Since revenue is less decentralized than expenditure on the average of
the data, our results also suggest mitigating the divergence of expenditure and
revenue assignments for higher economic growth.
4 Conclusion
The effect of fiscal decentralization on economic growth has been a major
focus of debate and discussion in the context of recent public reforms. It
has been suggested that some countries succeed with fiscal decentralization
and others do not. The empirical works regarding whether there is positive or
negative relationship between economic growth and fiscal decentralization are
still controversial. In this paper, we approach this issue from a different angle
than the previous literature. Taking account of complementarity in global and
local levels, we found that, in general, the relationship between fiscal decentralization and economic growth is not linear, so that the economic growth
implication of fiscal decentralization depends on the structure of complementarity. Using data for the United States, we first show that our specification of
the production function fits well. From our growth regression model, we observed a hump-shaped relationship between fiscal decentralization and economic
growth, which is consistent with our theoretical result applied to the estimated
parameters. We also examined whether the optimal level of fiscal decentralization derived by the data is higher or lower than that of the actual data. In
18 The Wald statistics = (b /(2b )-the average of the data)2 /V(b

2
1
1 /(2b2 )), where V(b1 /(2b2 ))

is a root-n consistent estimator of the variance obtained by the delta method (e.g., Greene (1993),
p. 297).

356

N. Akai et al.

some cases, the optimal degree of fiscal decentralization conducive to economic


growth is higher than the average of the data, and hence further decentralization
is recommended for economic growth.
As far as we know, no theoretical work discusses the hump-shaped relationship between them with a formal model that identifies the reason. Thiessens
(2003) empirical work shows the hump-shaped relationship between fiscal
decentralization and economic growth in the group of high-income OECD
countries, but he does not identify the reason for the humped shape. Given
this background, this paper is the first work which discusses the hump-shaped
relationship between fiscal decentralization and economic growth theoretically,
with consistent empirical analyses. We first provided a theoretical model that
explains a relationship between fiscal decentralization and economic growth,
extending a conventional macroeconomic model with central and local governmental sectors. Second, we estimated the production function consistent with
theoretical hypothesis. And finally, we examined whether the current level of
fiscal decentralization is optimal or not. Given that existing theories cannot
explain the hump-shapedness, and that the comprehensive analysis regarding
this issue has not existed yet, our paper has provided persuasive explanations
theoretically and empirically. The macroeconomic implications of fiscal decentralization depend on the complementarity structures (Table 1). Further
analyses with different data sets may complement our analyses.
Appendix A
In order to prove the proposition, we first derive the market clearing conditions.
There are two markets in this economylabor and capital. As in the conventional model, the goods market is complementary to the capital market. First, the
labor market equilibrium condition is met where the competitive firms labor
demand equals (exogenous) supply of labor. This is equivalent to the following
relationship between the wage rate wt , the national output Yt and the labor
supply Lt :19
1

wt = AKt

Gt Lt

Yt
Lt

(A.1)

We now show the condition for capital (goods) market equilibrium, following
the conventional models (e.g., Barro and Sala-i-Martin (1995, Eqs. (3.A.10)
(3.A.12))):
Lemma The condition for the capital market equilibrium implies:
Kt+1 = a(1 )Yt ,

(A.2)

where a is a fraction of the net income that is saved.


19 The expression of competitive firms capital demand, which determines the interest rate r , is
t

omitted, since this condition is not used in the following derivation.

Complementarity, fiscal decentralization and economic growth

357

Proof of the Lemma The condition for the capital market equilibrium (or the
goods market equilibrium) is where aggregate net investment equals total net
income minus total consumption. Tax revenue is used for public projects, so
that there is no net savings from the government sector. Formally, let Ct and
Dt denote the consumption of the young and old, respectively, and let rt be the
interest rate. Then:
Kt+1 Kt = (1 )wt Lt + rt Kt Ct Lt Dt Lt1 .
By an assumption of constant fraction of savings, (1 )wt Ct = a(1 )wt
and Dt = (1 + rt )a(1 )wt1 . The above equation is rearranged to:
Kt+1 = a(1 )wt Lt + (1 + rt )(Kt a(1 )wt1 Lt1 ).
At the beginning, initial capital stock K1 is owned by L0 , so that D1 L0 =
(1 + r1 )K1 [Barro and Sala-i-Martin (1995, p. 120, footnote 28)]. Since D1 L0 =
(1+r1 )a(1 )w0 L0 , the difference equation above implies Kt+1 = a(1 )wt Lt .
Combining with (A.1), we derive (A.2).


Proof of the Proposition Let n (0  n  J) be the number of successful jurisdictions (those who draw ). This is a random
variable which follows a binomial
J
distribution with probability y Pr(n|J)
n (1 )Jn .
n
Let Tt be the tax revenue per jurisdiction at period t. Recall that tasks
and jurisdictions are symmetric, so that an equal proportion of tax revenue
is allocated to each project: Tt (j, m) = Tt for all j and
 m. The local bureaucrat transforms the public projects m [0, ] to m[0,] gt+1 (j, m) dm =


{Tt (j, m)} dm = (Tt ) m[0,] dm = (Tt ) with probability
m[0,]

, and m[0,m] gt+1 (j, m) dm = m[0,] {Tt (j, m)} dm = (Tt ) m[0,]
dm = (Tt ) with probability 1 . The same rearrangement applies for
the projects undertaken by the central government. Recall that labor income
taxation takes place [so that the tax revenue = wt Lt = Yt by (A.1)].
Recall also that the jurisdictions are symmetric, so that Tt = Yt /J.
Substituting these expressions into (3) and (4), we now describe Gt+1 by the
ability of the central bureaucrat ( or ) and the number of successful jurisdictions (n), as follows. When the central bureaucrat has a high ability (which
yields gt+1 (j, m) = Tt (j, m) for m [ , 1]) and the number of successful local
bureaucrats is n(0  n  J),

Gt+1 =

n
J n
( + (1 )) +
( + (1 ))
J
J

Yt
.
J
(A.3a)

358

N. Akai et al.

When the central bureaucrat has a low ability (which yields gt+1 (j, m) = Tt (j, m)
for m [ , 1]) and the number of successful local bureaucrats is n (0  n  J),

Gt+1 =

n
J n
( + (1 )) +
( + (1 ))
J
J

Yt
.
J
(A.3b)

Notice that any realization of n(0  n  J) is a-priori possible.


Substituting (A.2), into (1),
Yt+1 = A(a(1 )Yt )1 (Gt+1 Lt+1 ) .
Substituting (A.3a) and (A.3b) into the above equation, for c = , and 0 
n  J, we obtain:
Yt+1
c
= k + ln t+1
(n),
Yt


 
c (n)(c = , , 0  n  J) is
, and t+1
L
where k = ln Aa(1 )1
J t+1
ln

a random variable such that:


1

n J n

+
( + (1 ))
=
ln
J
J
with probability Pr(, n) Pr(n|J),
1

n
J n
c

(n) = t+1
(n) ln
t+1
( + (1 )) +
J
J
with probability Pr(, n) (1 ) Pr(n|J),

with > 1.

c
t+1
(n)

t+1 (n)

Taking expectation of the above expression (i.e., multiplying both sides by


Pr(c, n) and summing up with respect to (c, n)),



Yt+1

ln
Pr(c, n) = k +
ln t+1 (n) Pr(n|J)
c,n
n
Yt


ln t+1
(n)(1 ) Pr(n|J).
+

Or, rewriting the expression:







Yt+1

n J n
E ln
= k+
En ln
+
( + (1 ))
Yt

J
J



Jn
En ln
( + (1 )) +
+ (1 )

J
J

(A.4)

Complementarity, fiscal decentralization and economic growth

359

where En is theexpectation operator with respect to the binomial distribution


J
c (n)]  ln c (n) Pr(n|J) for
Pr(n|J) =
n (1 )Jn , i.e., En [ln t+1
n
t+1
n
c = , . Taking the derivative with respect to :

E ln

Yt+1
Yt



1)( + (1 )) 1
Jn

J (
=
En

n
(1 ))

+ Jn
J
J ( +



nJ (1 )( + (1 )) 1

+ (1 )
. (A.5)
En

( + (1 )) + Jn
J

Evaluated at = 0,

E ln

Yt+1
Yt

=0



1)+
Jn

J (
En
=

 n

En (1 )
+ (1 )

J


J n
n

En
( 1) (1 ) (1 )
=

J
J

(1 )( 1)(1 ) > 0.
=

n
Notice that En
= , > 1 and < 0.
J

Also, evaluated at =1,



E ln

Yt+1
Yt

=1



1) + (1 ) n (1 )+
Jn

J (
J
= En
.
n

+ Jn
J
J

We first prove that this value is negative when is negative. Let


1

1) + (1 ) n (1 )+ and B(n) = n + Jn
(
.
A(n) = Jn
J
J
J
J
 Y

t+1

E ln Y
t

= {Covn (A(n), B(n)) + En [A(n)]En [B(n)]}, where Cov


Then

=1

denotes covariance. By supposition, both A(n) and B(n) are decreasing


functions of n,20 so that the covariance is positive. Also,
E [A(n)] = (1 )
 Y n

t+1

E ln Y
t

( 1)(1 + ) > 0 and En [B(n)] > 0, so that


< 0.

=1

20 A(n) = 1 ( 1)( + (1 )+ )
J
2
 n
1  1 < 0 (since > 1).
nJ + Jn
J
J

< 0 (since

> 1) and B(n)


n

360

N. Akai et al.

Also,

2 E ln

Yt+1
Yt
2

En 


(1 )

2

1)( + (1 )) 1


2
(1 ))
+ Jn
J ( +

Jn
J (

En



+
1 En

+ (1 )) 1

2

n
+ Jn
(

+
(1

))
J
J

n
)(
J (1

Jn
1)2 (
J (
n
+ Jn
J
J (



+ (1 )
1 En

2

+ (1 )) 2

+ (1 ))

n
)2 ( + (1 )) 2
J (1

n
+ (1 )) + Jn
J (
J


.

The supposition of < min{, 0} implies that the expression is negative for
all .




Y
Y
> E ln Yt+1
(Nishimura 2006).21
Finally, when > 0, E ln Yt+1

t
t
=0
 Y

 Y =1

t+1

t+1
E ln Y
2 E ln Y
t
t

To be consistent with
> 0 and
< 0 for all [0, 1], it
2

=0
 Y

t+1

E ln Y
t

< 0.
has to be

=1

Inspection of (A.5) shows that, when



 Y

resp.

E ln

t+1
Yt

0

0,


E ln

Yt+1
Yt

if < 0 (resp. > 0). Such relationship continues to



2 E ln

Yt+1
Yt
2

hold if is sufficiently close to 0. Also,


> 0 for all [0, 1] when
> max{, 0} (complementary to the case
of
the
proposition).
Combined with



Yt+1

Yt+1

> 0, we de E ln Yt

iff <
Nishimuras (2006) result that E ln Yt

=0

=1

rive the relationship between fiscal decentralization and economic growth in


Table 1.

21 = 0 corresponds to the centralized regime, and = 1 corresponds to the decentralized regime

of Nishimura (2006). Notice that the sign of in (4) is opposite to Nishimuras (2006) Eq. (5).

Complementarity, fiscal decentralization and economic growth

361

Appendix B Variable definitions, means, and standard deviations

Variable

Mean

Standard deviation Definition

GSP

0.0134

0.0094

POP

0.0047

0.0040

GSP
EDUC

24.1878 4.3925
0.0950 0.0137

LIB versus CON

0.5419

0.1582

GINI

0.0957

0.0233

PATENTS

0.0203

0.0290

OPENNESS

0.0588

0.0367

Indicators of fiscal
decentralization
Expenditure indicator 0.4402

0.0730

Revenue indicator

0.0824

0.3820

Average annual growth rate of real GSP


per capita over the 19921998 period
The lag of average annul growth rate of
state population
Per capita real GSP
Percentage of high school graduates in
total population aged 1824 years
The share of seats in state legislature held
by Democrats
Income difference between counties
within a state
The states share of total US patents
issued
Ratio of statess exports to other countries
and other states to nominal GSP

Ratio of local government expenditure to


state and local government expenditure
Ratio of local government revenue to
state and local government revenue

All variables except for GSP and POP relate to 1992, 93, 94, 95, 96, 97 each period
Sources: USA COUNTIES 1998 (CD-ROM), Statistical Abstract of United States, and States and
Metropolitan Area Data Book 199798 (CD-ROM)

References
Akai N, Sakata M (2002) Fiscal decentralization contributes to Economic growth: evidence from
state-level cross-section data for the United States. J Urban Econ 52:93108
Bardhan P (2002) Decentralization of Governance and Development. J Econ Perspect 16:185205
Barro RJ (1990) Government spending in a simple model of endogenous Growth. J Polit Econ
98:S103-S125
Barro RJ, Sala-i-Martin X (1995) Economic Growth. McGraw-Hill, New York
Bnabou R (1996) Heterogeneity, stratification, and growth: macroeconomic implications of community structure and school finance. Am Econ Rev 86:584609
Conybeare J, Murdoch JC, Sandler T (1994) Alternative collective-goods models of military
alliances: theory and evidence. Econ Inquiry 32:525542
Cornes R (1993) Dyke maintenance and other stories: some neglected types of public goods.
Q J Econ 108:259271
Cornes R, Sandler T (1996) The theory of externalities, public goods. and club goods, 2nd edn.
Cambridge University Press, New York
Cremer J, Estache A, Seabright P (1996) Decentralizing public services: what can we learn from
the theory of the firm? Revue dEconomie Politique 106:3760
Crewson PE (1995) A comparative analysis of public and private sector entrant Quality. Am J Polit
Sci 39:628639
Davoodi H, Zou HF (1998) Fiscal decentralization and economic growth: a cross-country study.
J Urban Econ 43:244257
Ebel RD, Yilmaz S (2002) On the measurement and impact of fiscal decentralization. World bank
policy research working paper no. 2809
Greene WH (1993) Econometric Analysis. Macmillan, New York

362

N. Akai et al.

Hirshleifer J (1983) From weakest-link to best-shot: the voluntary provision of public goods. Publ
Choice 41:371386
Hoxby M (1999) The productivity of schools and other local public goods producers. J Publ Econ
74:130
Lin JY, Liu Z (2000) Fiscal decentralization and economic growth in China. Econ Cult Change
49:121
Matsuyama K (1995) Complementarity and cumulative process in models of monopolistic competition. J Econ Lit 33:701729
Montinola G, Qian Y, Weingast BR (1996) Federalism, Chinese style: the political basis for economic
success. World Polit 48:5081
Nishimura Y (2006) Human fallibility, complementarity and fiscal decentralization. J Publ Econ
Theory 8:487501
Prudhomme R (1995) The dangers of decentralization. World Bank Res Observ 10:201220
Sah RK (1991) Fallibility in human organizations and political systems. J Econ Perspect 5:6788
Sato M, Yamashige S (2005) Decentralization and economic development: an evolutionary
approach. J Publ Econ Theory 7:497520
Thiessen U (2003) Fiscal decentralization and economic growth in high-income OECD countries.
Fiscal Stud 24:237274
Wallis JJ, Oates WE (1988) Decentralization in the public sector: local government. In: Harvey S,
Rosen H (eds) Fiscal federalism: quantitative studies. The University of Chicago Press, Chicago
and London
Xie D, Zou HF, Davoodi H (1999) Fiscal decentralization and economic growth in the United
States. J Urban Econ 45:228239
Zhang T, Zou HF (1998) Fiscal decentralization, public spending, and economic growth in China.
J Publ Econ 67:221240

Potrebbero piacerti anche