Does Tax Competition Reduce Corruption?
| DOI | http://doi.org/10.1111/sjpe.12096 |
| Date | 01 September 2016 |
| Published date | 01 September 2016 |
DOES TAX COMPETITION REDUCE
CORRUPTION?
Yu-Bong Lai*
ABSTRACT
A vast literature asserts that competition among jurisdictions can mitigate cor-
ruption. However, empirical studies do not fully support such a conventional
view. This paper points out a drawback of the conventional view, that is, it
ignores the role played by the owners of immobile factors. Within a standard
tax competition model, we amend the conventional view by incorporating the lob-
bying of the immobile factor owners, and show that fiscal decentralization can
aggravate corruption. We also demonstrate that social welfare can be higher
under decentralization than under centralization in some cases, whereas the
immobile factor owners’ lobbying will prevent such a superiority of decentraliza-
tion from being the outcome. Moreover, we provide an example to show that a
better policy outcome and a higher level of corruption can co-exist, which con-
trasts with the general belief.
II
NTRODUCTION
The effect of fiscal decentralization on economic performance is a hotly-de-
bated issue.
1
The leading role played by centralized states in improving effi-
ciency has been recognized on the one hand, but high levels of rent seeking
arising from monolithic government have also been a concern on the other
(Bardhan and Mookherjee, 2006). Several scholars argue that interjurisdic-
tional competition can serve as a disciplinary device to mitigate corruption
2
(Brennan and Buchanan, 1980; Weingast, 1995; Edwards and Keen, 1996;
Qian and Weingast, 1997; Ahlin, 2001; Arikan, 2004). These papers believe
that competition among different decentralized governments can exercise a
disciplinary force and restrict the monopoly power of a large central govern-
ment. As a result, they argue that decentralization leads politicians to limit
*National Chengchi University
1
For a survey of the related literature, see Bardhan and Mookherjee (2006), Lockwood
(2006), and Shah (2006).
2
Corruption is most commonly defined as the misuse or the abuse of public office for pri-
vate gain (World Bank, 2004). Thus, the monetary payment that is made to influence policy
for private gain can be regarded as a corrupt payment, which can be legal or illegal (Rose-
Ackerman, 2004).
Scottish Journal of Political Economy, DOI: 10.1111/sjpe.12096, Vol. 63, No. 4, September 2016
©2015 Scottish Economic Society.
331
the amount of corrupt earnings that they try to extract, because of the capital
flight.
The empirical evidence, however, has not fully supported such an argument
(we refer to it as the conventional view hereafter). Some studies indicate that
fiscal decentralization is significantly associated with lower corruption
(Estache and Sinha, 1995; Huther and Shah, 1998; Fisman and Gatti, 2002a
b; Arikan, 2004; Fan et al., 2009). Other studies reveal that the effect of fiscal
decentralization on corruption is insignificant, and some even find that cor-
ruption increases with fiscal decentralization (Goldsmith, 1999; Treisman,
2000; Pellegrini and Gerlagh, 2008; Lessmann and Markwardt, 2009).
Although these studies adopt different indicators to measure fiscal decentral-
ization and governance, it appears that the argument that interjurisdictional
competition can mitigate corruption is inconclusive. This present paper aims
at modifying the conventional view, so as to explain why interjurisdictional
competition does not necessarily reduce corruption, and may even aggravate
it in some cases.
We agree with the logic of the conventional view. What we argue is that
the conventional view, which places an emphasis on owners of mobile factors,
ignores the role played by owners of immobile factors. By simultaneously con-
sidering the two types of factor owners, we show that they behave in opposite
ways in bribing policymakers, resulting in ambiguous impacts of fiscal decen-
tralization on corruption.
To elaborate on this idea, we construct a standard tax competition model,
which considers a state that contains a large number of identical jurisdictions.
Competitive firms in each jurisdiction use a mobile factor and an immobile
factor to produce outputs. To be concrete, we let the mobile factor be capital,
and the immobile factor be labor.
3
A tax is imposed on capital according to
the source principle. To highlight the conflict between different groups of resi-
dents, the tax revenues are distributed to residents in a lump-sum manner.
Each jurisdiction contains three types of residents: owners of capital (capital-
ists), workers, and pensioners. Capitalists and workers are assumed to be
organized into distinct groups to affect the capital tax rate by bribing their
policymaker.
As it will become clear later, this paper adopts the common-agency model,
in which lobbies offer the incumbent government contributions that are con-
tingent on the policy decisions taken. Thus, lobbying is viewed as a sale of
public policy for private gains. In this approach, the political contributions
are assumed to be legal, so that they are defined as the lobbying expenses.
However, Schulze and Ursprung (2001) argue that ‘since contributions are
aimed at influencing the government’s policy ⋯, the portrayed interaction
between organized interest groups and government meets the circumstances
of corruption’(p. 68). Even Grossman and Helpman (1994) recognize that
3
Choosing land as the immobile factor does not change the results that follow. Most of
the related papers specify labor as the immobile factor, and we follow this approach. What
we focus on is the relative mobility of the factors under decentralization. The mobility of
labor, at least over a short period of time, is low. This justifies this specification.
332 YU-BONG LAI
Scottish Journal of Political Economy
©2015 Scottish Economic Society
To continue reading
Request your trial