The role of monitoring of corruption in a simple endogenous growth model.
Coppier, Raffaella ; Costantini, Mauro ; Piga, Gustavo 等
I. INTRODUCTION
During the last 30 years, economists from various fields have
contributed to the analysis of corruption. The first paper to receive
widespread attention was published in 1975 (Rose-Ackerman 1975). Since
then a large literature has developed and much attention has been paid
to the relationship between corruption and economic growth. In the
analysis of the consequences of corruption, the literature supports two
opposing positions. On one hand, common wisdom views corruption as an
obstacle (sand) to development and growth; on the other hand, other
researchers, since the pioneering works of Left (1964) and Huntington
(1968), have suggested that corruption may promote efficiency by
allowing firms to bypass government failures of various sorts
(grease corruption). More recent expositions of
efficiency-enhancing corruption can be found in Lui (1985), Acemoglu and
Verdier (1998, 2000), and Barreto (2000). Barreto (2000) presents a
simple neoclassical endogenous growth model where the public
sector's monopolistic position is explicitly considered. Results
indicate that a corruption equilibrium is characterized by lower growth
rates compared to the ideal situation in which public goods are provided
competitively. Barreto (2000) also shows that if the public sector is
subject to significant bureaucratic red-tape, all of the agents within
the economy may prefer the corruption equilibrium, as corruption can
bypass bureaucratic obstacles.
From a theoretical point of view, there are several ways in which
corruption may reduce economic growth. Corruption can act as a tax and
thus lower incentives to invest. Corruption could cause talented people
to engage in rent-seeking rather than productive activities. Corruption
may distort the composition of government expenditure, as corrupt
politicians could be expected to invest in large non-productive projects
from which considerable bribes can be extracted more easily than from
productive activities.
Empirical analysis has also provided evidence of the negative
effects of corruption on economic growth. Mauro (1995) shows that
corruption reduces economic growth, via reduced private investment. (1)
Similar results are obtained by Keefer and Knack (1995) using other
measures of corruption and a different selection of countries. (2)
Recently, the "grease the wheels" hypothesis has been tested
and statistically supported (see Aidt 2009; Campos, Dimova, and Ahmad
2010; Meon and Sekkat 2005, among others).
In this article, we develop a theoretical endogenous growth model
which incorporates corruption. In our model, the social loss brought
about by corruption stems from the fact that firms must bribe a
bureaucrat in order to invest, and consequently devote fewer resources
to the accumulation of capital. Therefore, in our model, corruption has
a negative effect on private investment. Other models share our
framework. Ehrlich and Lui (1999) claim that individuals have an
incentive to compete over the privilege of becoming bureaucrats (the
so-called investment in political capital) since they obtain economic
rents through corruption. This investment in political capital consumes
economic resources which could otherwise be used for production or
investment in human capital. In Del Monte and Papagni (2001), corruption
arises when bureaucrats manage public resources to produce public goods
and services. Corruption reduces the quality of public infrastructure
resulting in a negative effect on economic growth.
The novel feature of this article is the study of the impact of
monitoring of corruption on economic growth. (3) In our theoretical
model, we derive a nonlinear relationship between the level of
monitoring and economic growth, as well as between corruption and
economic growth. (4) At low monitoring levels, the economy experiences
widespread corruption and medium growth rates, whereas no corruption
occurs at intermediate monitoring levels, but low growth rates are
recorded. At high monitoring levels, no corruption takes place and high
growth rates are observed.
The nonlinear relationship between growth and monitoring is finally
investigated empirically over the period 1980-2003 in Italy. To study
this relationship, new measures of monitoring are used. Our empirical
evidence supports the conclusions of the model.
The article is organized as follows: Section II studies the
relationship between the level of monitoring, corruption, and economic
growth. In Section III, empirical implications from the theoretical
model are evaluated. Section IV concludes.
II. THEORETICAL MODEL
Let us consider an infinite horizon economy in continuous time
admitting a representative household with the standard constant relative
risk aversion preferences. We also assume that the representative
household owns a balanced portfolio of all the firms in the economy.
Alternatively, we can think of the economy as consisting of many
households with the same preferences as the representative household in
each household holding a balanced portfolio of all the firms. Following
Acemoglu (2009), the ability to hold a balanced portfolio of projects
with independently disputed returns allows the individual to diversify
the risks and act in a risk-neutral manner. (5) This will imply that the
objective of each firm will be to maximize expected profits (without a
risk premium). Firms manufacture a single homogeneous good y with one
input-capital using one of the two technologies with constant returns to
scale: the modern sector technology and the one of the traditional
sector. Each firm is assumed to have the same quantity of capital k. The
product may be either manufactured for consumption purposes or for
investment purposes. The modern sector technology is y = [a.sub.M]k. The
firms in the modern sector must obtain a license from the government to
access the technology. To obtain such a license, a firm must submit a
project to a bureaucrat and this act involves an implementation cost of
sk. (6) The firm may access the traditional sector without any license
being issued. In this case the output is y = [a.sub.T]k. From this point
onwards, it is assumed that ([a.sub.M] - s) > [a.sub.T] > 0, that
is the modern sector is more profitable than the traditional sector.
While households may invest their capital in the modern sector or in the
traditional one, bureaucrats cannot invest in the production activity,
earning a fixed salary w. (7) Bureaucrats are corruptible, in the sense
that they pursue their own interest, and not necessarily that of the
State. Bureaucrats are open to bribery as they issue the license
required to access the modern sector technology to the firms submitting
a project. The State controls the bureaucrats in such a way that they
have a probability q (monitoring level) of being detected if they
undertake a corrupt transaction.
In this model, the bureaucrat may decide not to ask for a bribe and
to issue the license to those firms who submit a project, or else to ask
for a bribe (b hereafter) in exchange for the license. Since ([a.sub.M]
- s) > [a.sub.T], the firm might find it worthwhile to offer a bribe
to the corrupt bureaucrat with a view to obtaining the necessary license
to access the modern sector. The bureaucrat is assumed to have both
monopolistic power (i.e., after having submitted the project, the firm
cannot turn to any other bureaucrat to obtain the license) and
discretional power over granting the license (i.e., the bureaucrat may
refuse to issue the license without being required to provide any
explanation). If the bureaucrat is detected while performing a corrupt
transaction, he/she incurs a cost (either monetary, moral, or criminal)
equal to mk, where m > 0 (8); the firm, if detected, incurs a cost
(either monetary, moral, or criminal) equal to ck, where c > 0, but
the cost of the bribe paid to the bureaucrat is refunded. (9)
A. Game Description
In the following, we refer to the firm payoff by using the
superscript (F) and to the bureaucrat payoff by using the superscript
(B). These represent the first and the second element of the payoff
vector [[eta].sub.i], i = 1, 2, 3, 4, respectively, as it will become
clear below there are four payoff configurations. Consider the following
three-stage game:
Stage 1. At Stage 1 of the game, the firm decides in which sector
to operate, that is, whether to invest its capital in the modern or in
the traditional sector. Such a decision is tantamount to the decision of
whether to submit the project to the bureaucrat, considering that a
license is needed to invest in the modern sector. Project submission
does not automatically result in the bureaucrat issuing a license, as
he/she may refuse to grant the license unless a bribe b is paid. If the
firm decides not to submit the project (preferring to invest in the
traditional sector instead) the game ends and then the payoff vector is
given by:
(1) [[[eta].bar].sub.1] = ([a.sub.T] k, w).
If the firm decides to submit the project, it asks the bureaucrat
to issue the license. In this case the game continues to Stage 2.
Stage 2. At this stage the bureaucrat, on facing a firm who has
submitted a project incurring a cost sk, may decide to issue the license
without asking for a bribe (b = 0). (10) In this case the game ends and
the payoff vector is given by:
(2) [[[eta].bar].sub.2] = ([a.sub.M] k - sk, w).
Alternatively, if he/she demands the payment of a bribe (b > 0)
from the firm before agreeing to issue the license, the game continues
to Stage 3.
Stage 3. At Stage 3, the payoffs will depend on whether, on one
hand, the agreement between the bureaucrat and the firm is achieved or
not and, on the other hand, whether the bureaucrat and the firm are
reported (with probability q) or not. Should it decide to negotiate the
payment of a bribe with the bureaucrat, the two parties will find the
bribe corresponding to the Nash solution to a bargaining game
([b.sup.NB]). If agreement is not reached, the bureaucrat will refuse to
issue the license; thus the game ends with the bureaucrat receiving
his/her salary and the firm, after having been denied the license, will
be left with no other option but to invest in the traditional sector. In
this case, the game ends and the corresponding payoff vector is given
by:
(3) [[[eta].bar].sub.3] = ([a.sub.T] k - sk, w).
If agreement is reached, the expected payoffs will depend on the
probability q with which the bureaucrat and the firm are monitored. (11)
In this case, the expected payoff vector is given by:
(4) [[[eta].bar].sub.4] = (([a.sub.M] - s)k - qck - [b.sup.NB] (1 -
q), x w - qmk + [b.sup.NB](1 - q)).
It should be noted that [[[eta].bar].sub.2] is preferred to
[[[eta].bar].sub.3] by both agents, and therefore the bureaucrat--will
never ask for a bribe if he/she knows that the agreement will not be
achieved.
B. The Solution to the Game
The model described in the previous section well reflects the
pervasive uncertainty which is typically experienced by firms when
dealing with the Public Administration. The game may be solved by
starting from the last stage using backward induction, determining the
bribe [b.sup.NB] (which is the Nash solution to a bargaining game). The
bribe [b.sup.NB] is the outcome of a negotiation between the bureaucrat
and the firm (see Appendix A for the proof).
PROPOSITION 1. Let q [not equal to] 1. (12) Then there is a unique
non negative bribe ([b.sup.NB]), as the Nash solution to a bargaining
game, given by:
(5) [b.sup.NB] = [alpha][([a.sub.M] - [a.sub.T])k/(l - q) -(q(c -
m)k/(1 - q))],
where [alpha] and (1 - [alpha]) are parameters which can be
interpreted as the bargaining strength measures of the bureaucrat and
the firm, respectively.
Let us assume, without loss of generality, that the firm and the
bureaucrat share the surplus on an equal basis. (13) Thus we have a
standard Nash case, when [alpha] = (1 - [alpha]) = 1/2 and the firm and
the bureaucrat receive equal shares. Hence the bribe is equal to:
(6) [b.sup.NB] = 1/2[(([a.sub.M] - [a.sub.T])k/(1 - q)) -(q(c -
m)k/(1 - q))].
Static Analysis. The game is solved by means of backward induction
starting from the last stage and the solution is formalized by the
following proposition (see Appendix A). (14)
PROPOSITION 2. Define [q.sub.2] = (([a.sub.M] - [a.sub.T])/ (c +
m)) - (2s/(c + m)) and [q.sub.1] = ([a.sub.M] - [a.sub.T])/ (c + m) with
[q.sub.1] > [q.sub.2]. Then, if [q.sub.2] [greater than or equal to]
0, [q.sub.1] [less than or equal to] 1 and (c + m) > 2s:
(C) If q [member of] [0, [q.sub.2]] then the equilibrium payoff
vector is:
(7) [[[eta].sub.4] = ((([a.sub.M] + [a.sub.T])k/2) - sk - (q(c +
m)k/2), w + (([a.sub.M] - [a.sub.T])k/2) - (q(c + m)k/2))
this is the payoff vector connected to equilibrium C (see below);
(B) if [q.sub.2] < q < [q.sub.1] the equilibrium payoff
vector is:
(8) [[eta].sub.1] = ([a.sub.T] k, w)
this is the payoff vector connected to equilibrium B (see below);
(A) If q [member of] [[q.sub.1], 1] the equilibrium payoff vector
is:
(9) [[eta].sub.2] = (([a.sub.M] - s)k, w)
this is the payoff vector connected to equilibrium A (see below).
The previous proposition shows that, depending on the parameter
values, one of the three perfect Nash equilibria is obtained in the
subgames:
* Equilibrium C: corruption and high output. When 0 [less than or
equal to] q [less than or equal to] [q.sub.2], that is if the monitoring
level is low enough, the firm will enter the modern sector and will be
asked to pay a bribe by the bureaucrat. Monitoring intensity is so low
that the difference in gross profits, ([a.sub.M] = [a.sub.T])k, between
the modern and the traditional sector is high enough to outweigh a
(relatively low) expected cost of corruption and the cost of the
project.
* Equilibrium B: no corruption and low output. When [q.sub.2] <
q < [q.sub.1] that is if the monitoring level is intermediate, the
firm will not enter the modern sector and therefore will not ask for a
license. Monitoring intensity is not low enough for the firm to justify
paying for the cost of the project along with the additional expected
cost of paying a bribe. The difference in gross profits between the
modern and the traditional sector does not compensate for the expected
cost of corruption plus the cost of the project. Furthermore, monitoring
intensity is not of a high enough level to deter the bureaucrat from
asking for a bribe where the firm would have paid to pay the cost of the
project.
* Equilibrium A: no corruption and high output. When q [is greater
than or equal to] [q.sub.1], that is if the monitoring level is high
enough, the level of monitoring intensity by the State is so high that
the firm would turn down a request for a bribe even after having paid
the (sunk) cost of submitting a project. Realizing this fact, the
bureaucrat will refrain from asking for a bribe to issue the license.
Thus the firm will enter the modern sector and will not be asked for a
bribe by the bureaucrat.
Notice that in equilibrium B there is no corruption, but low output
compared to equilibrium C, where corruption is at its highest, but
output is higher. Should a State wish to lead the economy toward one of
these three viable equilibria by employing a certain level of
monitoring, it would realize that equilibria A and C imply a greater
output than equilibrium B. Equilibria A and C allow the same output to
be obtained, even though they are considerably different from one
another in terms of level of corruption (which is greatest in C and
nonexistent in A).
From a static perspective, equilibrium A is better than equilibrium
C which implies the same output as equilibrium A but is characterized by
widespread corruption, entailing a higher cost, summarized by parameters
c and m. A is also better than B, while B and C cannot be ranked a
priori.
Dynamic Analysis. Following the work of Del Monte and Papagni
(2007), we expand the game perspective in order to examine the dynamic
consequences of corruption on investment and hence on economic growth.
As we said, the household's satisfaction is derived from
consumption according to a simple constant elasticity utility function:
U = ([C.sup.1 - [sigma]] - 1)/(1 - [sigma])
Each household maximizes utility over an infinite period of time
subject to a budget constraint. This problem is formalized as:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
subject to
[??] = [[eta].sup.F.sub.i] - C,
where C is consumption, [rho] is the discount rate over time,
[[eta].sup.F.sub.i] is the household's payoff.
Since [[eta].sup.F.sub.i] is different in each of the three
equilibria, the problem is solved for each of the three cases. In the
equilibrium with corruption (equilibrium C), the household's payoff
is:
[[eta].sup.F.sub.4] = [([a.sub.M] + [a.sub.T]k/2 - sk - (q(c +
m)k/2)],
thus the constraint is:
[??] = [(([a.sub.M] + [a.sub.T])k/2) - sk - (q(c + m)k/2)] - C.
The Hamiltonian function is:
H = [e.sup.-[rho]t] ([C.sup.1 - [sigma]] - 1/1 - [sigma]) +
[lambda][(([a.sub.M] + [a.sub.T])k/2) - sk - (q(c + m)k/2) - C]
where [lambda] is a costate variable. Optimization provides the
following first-order conditions:
(10) [e.sup.-[rho]t] [C.sup.-[sigma]] - [lambda] = 0
and
(11) [??] = -[lambda][(([a.sub.M] + [a.sub.T])/2)
- s - (q(c + m)/2)].
By differentiating the first condition in Equation (10) with
respect to time and substituting it into the second condition in
Equation (11), the consumption growth rate is obtained:
[y.sup.C.sub.C] = 1/[sigma][(([a.sub.M] + [a.sub.T])/2)
- s - (q(c + m)/2) - [rho]].
In equilibrium A, the household's payoff is
[[eta].sup.F.sub.s] = [a.sub.M]k - sk.
In this case, optimization provides the first-order conditions that
allow the corresponding consumption growth rate to be obtained:
[[gamma].sup.C.sub.A] = 1/[sigma][[a.sub.M] - s - [rho]].
In equilibrium B, the household's payoff is
[[eta].sup.F.sub.1] = [a.sub.T]k
and the corresponding consumption growth rate is
[[gamma].sup.C.sub.B] = 1/[sigma][[a.sub.T] - [rho]].
It should be noted that:
[[gamma].sup.C.sub.A] > [[gamma].sup.C.sub.C] >
[[gamma].sup.C.sub.B]
that is equilibrium A (no corruption, high-level monitoring) has
the highest consumption growth rate; in equilibrium C (pervasive
corruption, low monitoring) the consumption growth rate is intermediate;
and finally in equilibrium B (no corruption, intermediate monitoring
level) the firm invests in the traditional sector, with low profits, low
accumulation of capital, and a low growth rate. Furthermore, it can be
shown that capital and income also have the same growth rate as
consumption. Therefore, of the three equilibria, from a dynamic
viewpoint, equilibrium A is the most conducive to economic growth. This
is shown in Figure 1 in terms of monitoring level and growth rate:
equilibrium A (high-level monitoring without corruption) produces the
highest growth rate since the firms, who are investing in the modern
sector without paying bribes, are able to generate greater accumulation
of capital; in equilibrium C the growth rate is intermediate, since
although the firm manages to invest in the modern sector, it must pay
bribes in order to do so and ends up accumulating less; finally in
equilibrium B the firm invests in the traditional sector, with low
revenues and low accumulation of capital. Thus a nonlinear U-shaped
relationship between the monitoring of corruption and economic growth is
obtained and this shall be tested empirically in the next section.
[FIGURE 1 OMITTED]
III. EMPIRICAL ANALYSIS
The relatively recent Italian nationwide Mani Pulite ("Clean
Hands") scandal, in conjunction with judicial authorities
implementing greater levels of monitoring as a consequence, drastically
affected Italy's economic environment. This Italian experience
lends itself naturally to verifying the impact of the monitoring level
on corruption and growth. (15) This section aims to empirically
investigate the relationship between the level of monitoring of
corruption and economic growth in Italy. The nonlinear character of the
relationship between the monitoring level and the growth rate of income
is formalized by using an empirical specification reflecting a parabolic
relationship between these two variables. The theoretical model is
tested using new monitoring measures. The empirical analysis is first
performed considering all the Italian regions and then three subsamples,
namely the North, Center, and South of Italy. The three subsamples are
considered in order to investigate whether the U-shaped relationship may
simply reflect the structural differences within the subsamples.
A. Data
The empirical analysis is based on annual data from Italian regions
over the period 1980-2003. With the exception of monitoring and human
capital variables, the annual data are drawn from the Prometeia Regional
Accounting data set (courtesy of ISAE). The data relating to monitoring
are selected from ISTAT and the Ministero dell'Economia e delle
Finanze, and data regarding human capital are drawn from the Costantini
and Destefanis (2009) data set. Appendix B provides a detailed
description of the variables and their sources. The descriptive
statistics of the variables are found in Appendix C.
With regard to the monitoring variable, three different measures
are provided with a view to study the effects of monitoring on economic
growth. The first is based on the number of corruption crimes and is
denoted as [M.sub.1]. The second and the third measures use the number
of pertinent judges and police officers as proxies to indicate how much
of the State resources are allocated to fighting corruption-related
crimes. These two measures are denoted as [M.sub.2] and [M.sub.3],
respectively. The first measure ([M.sub.1]) is the ratio between the
number of corruption crimes detected and the estimated total number of
corruption crimes (see Appendix D). This is an expost variable since it
only considers the results of State control activity. Therefore, this
index expresses the effectiveness of the monitoring activity, that is
the monitoring that leads to the corrupt bureaucrat being successfully
charged. The second and third measures are based on the number of judges
assigned to penal law cases ([M.sub.2]) and on the number of police
officers employed in the investigation of corruption crimes ([M.sub.3]),
respectively. Incentives for corruption increase as the probability of
being caught and punished decreases and this probability is positively
dependent on the actions of judges and police officers. These two
proxies are of an ex-ante nature, since they allow us to assess the
level of monitoring implemented by the State.
B. Estimation Methods
The specification of the basic estimated equation corresponds to a
reduced form so as to evaluate the implications of the theoretical
model. We consider the following specification equation:
(12) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
where the disturbance error term, [[epsilon].sub.it], has three
components: the fixed effect (which accounts for any individual-specific
effect), [[mu].sub.i], the unobservable time effect (which accounts for
any time-specific effect and it is individual-invariant),
[[lambda].sub.t], and the idiosyncratic shocks, [v.sub.it] (see Baltagi
2005). [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] is the growth
rate of per capita income at 2000 constant prices, ln [y.sub.it-1] is
the logarithm of the lagged value of the per capita income level,
ln([monitor.sub.it-1]) is the log monitoring level delayed by one
period, (ln([monitor.sub.it-1]))2 is the square of the logarithm of the
monitoring variable lagged by one period, [inv.sub.it] is the share of
investment in Gross Domestic Product (GDP), [conpa.sub.it] is public
consumption over GDP and [h.sub.it] is the stock of human capital. (16)
The index i refers to the cross-section dimension (regions) and the
index t to the time dimension. The share of investment over GDP and the
level of public consumption over GDP are important control variables
(see Barro 1991 and Levine and Renelt 1992). The monitoring variable is
included in the equation with a delay of one period for two reasons.
Firstly, changes in the monitoring level are very likely to require some
time before they influence the agents' decisions. Secondly, any
distortions due to simultaneity, resulting from the possible endogeneity
of the monitoring variable, need to be mitigated, since a higher growth
rate may result in more tax revenue and, therefore, more resources being
allocated to monitoring activity.
Equation (12) is estimated using the system generalized method of
moments (GMM) estimator proposed by Arellano and Bover (1995) and
Blundell and Bond (1998) with a finite sample correction for the
two-step covariance matrix derived by Windmeijer (2005). This estimator
augments the difference GMM estimator developed by Arellano and Bond
(1991) which uses first-differences to remove the unobserved
time-invariant country-specific effects ("fixed effects") and
instruments the right-hand side variables in first-differenced equations
using levels of the series lagged two periods or more. Blundell and Bond
(1998) show that the difference GMM estimator is likely to perform
poorly when the series are persistent. In this case the available
instruments are only weakly correlated with endogenous variables, and
the difference GMM estimator is likely to suffer from serious bias as
well as imprecision. Blundell and Bond (1998) suggest making use of
additional information besides the differences. They consider an
additional assumption which requires a restriction on the initial
conditions. (17) This allows the use of lagged first-difference of the
series as instruments for equations in level. Thus, the system GMM
estimator stacks the equations in first differences and the equations in
levels together and employs both lagged levels and first differences as
internal instruments (see Roodman 2006). (18) The consistency of the GMM
estimators depends on whether lagged values of the explanatory variables
are valid instruments in the growth equation. In the empirical analysis,
we address this issue by considering two specification tests suggested
by Arellano and Bond (1991) and Arellano and Bover (1995). The first is
a Sargan test of over-identifying restrictions, which tests the overall
validity of the instruments by analyzing the sample analog of the moment
conditions used in the estimation process. Failure to reject the null
hypothesis gives support to the model. The second test examines the null
hypothesis that the error term is not serially correlated. As in the
case of the Sargan test, the model specification is supported when the
null of no serial correlation cannot be rejected. The estimation results
are summarized in Section C.
C. Results
The empirical results are reported in Tables 1-4. In the system GMM
estimates, investment (Inv), human capital ([h.sub.it]), and public
consumption (conpa) are treated as endogenous. These endogenous
variables are instrumented with suitable lags of their own differences
for the regression equation (see notes in Table 1). Results of Sargan
tests of over-identifying restrictions for the GMM estimator confirm the
validity of the instruments. For all the regressions, we find that the p
value is larger than 5%. In addition, the results of the Arellano and
Bond (1991) tests for AR(2) autocorrelation in the first-difference
residuals show that null hypothesis of no serial correlation cannot be
rejected and the moment conditions are correctly specified.
The coefficient on the [lny.sub.it] is used to test the convergence
hypothesis. A negative sign denotes conditional convergence of growth
rates. In our estimation, the sign of the parameter [[beta].sub.1] is
negative and also statistically significant in all the cases. The
convergence rate is stronger across the poorest regions (South) than
that across the richest regions (North). Therefore, we can argue that
the forces of convergence are stronger in lower levels of per capita
income and less intensive in higher levels of income. These results are
in line with those in the Italian literature (see, e.g., Cellini and
Scorcu 1995).
The estimated regression coefficients of the square of the
logarithm of the monitoring variables are all positive and statistically
significant (see Tables 1-4). Although these results confirm the
existence of a U-shaped relationship as predicted by the theoretical
model, differences regarding the upward and downward sloping part of the
relationship are found at macro-regional level (North, Center, and
South) and among the monitoring measures (M j, [M.sub.2], and
[M.sub.3]). (19) When considering all the Italian regions as a whole,
the upward sloping part of the relationship is the most relevant for all
the monitoring measures. Furthermore, the positive marginal effects show
that more monitoring implies less corruption and therefore higher
economic growth. In this respect, the most effective instrument of
monitoring is represented by the judicial system which is independent
from the political power (see Della Porta 2001). (20)
When accounting for the different macroarea, contrasting results
are found. As regards the North and the Center of Italy, the upward
sloping part of the relationship is the most relevant. Furthermore,
among the monitoring measures, the most effective ex-ante instrument for
controlling for corruption seems to be the number of pertinent judges
([M.sub.2]), while the police officers measure ([M.sub.3]) is the least
effective.
In the South of Italy, the situation seems to be reversed. The
relevant part of the U-shaped relationship is the downward sloping one.
Also, the negative marginal effects show that a higher level of
monitoring implies a lower economic growth. This result can be
interpreted in light of the theoretical model: the South of Italy mainly
lies in a situation where a rise of monitoring involves paying a higher
bribe, fewer resources are devoted to capital accumulation, and lower
economic growth is expected. In this case the government is unable to
prevent waste, fraud, and mismanagement from occurring in the monitoring
activities and then it is likely to be generally less effective in its
capacity to govern (see, e.g., Putnam 1993). When considering the
different measures of monitoring, a rise in number, the police officers
imply a smaller reduction in the economic growth than that obtained with
an increase in the number of pertinent judges.
As regards the investment/GDP ratio (Inv) variable, the estimated
coefficients are all positive and statistically significant in all the
cases. These results would seem to be in line with the literature
concerning growth models (see, e.g., Levine and Renelt 1992) and similar
findings are also found in other studies of Italian regions (see Auteri
and Costantini 2004). As far as the performance of investment in the
three subsamples is concerned, a larger estimated coefficient is found
for the North of Italy (the estimated coefficient varies from 0.178 to
0.199). This result may be due to the fact that the public
infrastructures are more developed in the North than in the other
macro-areas.
The public consumption variable has positive coefficients in all
cases. When all the Italian regions are considered, the estimated
coefficient values vary from 0.201 to 0.263 and are all statistically
significant. Del Monte and Papagni (2007) found similar results,
although their evidence of a positive impact of public consumption on
economic growth is weaker. As regards the three subsamples, the highest
value of the estimated coefficient of public consumption is found for
the North of Italy. This result reflects the fact that public
expenditures are more efficient in the North than in the other
subsamples. Indeed, Chang and Li (2011) point out that what is important
for the effectiveness of public spending is not the size of the
government, but the quality of the government itself.
With respect to the human capital variable, a positive and
statistically significant effect on economic growth is also found. The
level of education has a crucial impact on growth as it determines the
economy's capacity to carry out technological innovation (see,
e.g., Woo 2009, on education and technology). When the three subsamples
are considered, the strongest impact of human capital on economic growth
is found in the Northern regions (the estimated coefficient varies from
0.050 to 0.053).
IV. CONCLUSIONS
In this article, a new theoretical model of the link between
monitoring of corruption and growth is developed. The model highlights
the nonlinear relationship between the level of State monitoring and
economic growth: when monitoring against corruption is low (high), high
(low) corruption and high income prevail. However, when monitoring is
intermediate, income production in the economy remains low. Similar
results prevail in a dynamic framework. The nonlinear relationship is
investigated using regional data for Italy over the period 1980-2003. An
empirical analysis is then carried out considering all Italian regions
and the three Italian macro-areas (North, Center, and South). The
empirical results confirm the existence of a U-shaped relationship as
predicted by the theoretical model, with differences regarding the
upward and downward sloping part of the relationship found at the
macro-regional level. In particular, we find, as predicted by our model,
that in the Northern and Center part of Italy a direct relation between
monitoring and economic growth prevails, so that making controls more
pervasive depresses corruption and boosts growth. On the contrary, in
Southern Italy more monitoring leads to less economic growth. This is
too coherent with our model where, when monitoring is low to begin
with--and corruption widespread but compatible with investment and
growth--and is raised through greater controls, this induces bureaucrats
to ask for a higher bribe, depressing in turn investment and economic
activity. We believe that a nonlinear pattern of this kind may provide
also interesting insights for the future study of cross-country
differences in the relationship between the fight against corruption and
growth.
ABBREVIATIONS
GDP: Gross Domestic Product
GMM: Generalized Method of Moments
APPENDIX A
THE NASH BARGAINING BRIBE [b.sup.NB]
Let [[PHI].bar].sub.[DELTA]] = [[PHI].sup.(F).sub.[DELTA]],
[[PHI].sup.(B).sub.[DELTA]] be the vector of the differences in the
payoffs between the case of agreement and disagreement regarding the
bribe between the firm and the bureaucrat, In accordance with
generalized Nash bargaining theory, the division between two agents will
solve:
(A1) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
that is
(A2) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
which is the maximum of the product between the elements of
[[PHI].bar].sub.[DELTA]] and where [([a.sub.T]k - sk), w] is the point
of disagreement, that is the payoffs that the firm and the bureaucrat
would obtain respectively if they failed to reach an agreement. The
parameters (1 - [alpha]) and [alpha] can be interpreted as measures of
bargaining strength. It is now easy to check that the bureaucrat gets a
share [alpha] of the surplus [tau], that is the bribe is [b.sup.NB] =
[alpha][tau]. More generally [alpha] reflects the distribution of
bargaining strength between the two agents.
Then the bribe [b.sup.NB] is an asymmetric (or generalized) Nash
bargaining solution and is given by:
(A3) [b.sup.NB] = [alpha] [(([a.sub.M] - [a.sub.T])k(1 - q)) - (q(c
- m)k/(1 - q))]
which is the unique equilibrium bribe in the last subgame, [for
all]q [not equal to] 1.
SOLUTION TO THE STATIC GAME
The static game is solved using backward induction, which enables
the equilibria to be obtained.
Proof of Proposition 2.
Proof Backward induction method.
(3) At Stage 3, the agreement is achieved if, and only if,
[[eta].sup.(F).sub.4] > [[eta].sup.(F).sub.3] [??] ([a.sub.T]k -
sk) < (([a.sub.M] + [a.sub.T])k/2) - sk - (q(c + m)k/2)
that is if the firm negotiates the bribe its payoff is greater than
its payoff if it refuses. That is verified [for all]q < (([a.sub.M] -
[a.sub.T])/(c + m)) = [q.sub.1].
Notice that in order to have an admissible probability set, q must
belong to [0, 1]. It should be noted that [q.sub.1] is greater than one
by assumption.
Furthermore, from now on we assume that [q.sub.1] < 1, that is
the difference in returns between the two sectors must not be greater
than the expected cost of corruption; consequently the presence of the
probability q determines the firm's choice of whether to enter into
the transaction. Then if q < [q.sub.1] the firm negotiates the bribe,
otherwise if q [greater than or equal to] [q.sub.1] it refuses the
bribe. Otherwise, if [q.sub.1] [greater than or equal to] 1 then the
firm will always negotiate the bribe (See Appendix A3 for details).
(2) Ascending the decision-making tree, at Stage 2 the bureaucrat
decides whether or not to ask for a bribe.
* If q [greater than or equal] [q.sub.1] then the bureaucrat knows
that the firm will not accept any bribe. Should he/she decide not to ask
for a bribe, his/her payoff will be w, whereas should he/she decide to
ask for a bribe, he/she knows there is no room for negotiation, and
therefore he/she will refuse to grant the license to the firm, which
will be forced to invest in the traditional sector. In this case the
bureaucrat's payoff will be w. Thus the bureaucrat's payoff is
the same as if he/she decides to ask for a bribe equal to zero. As
noted, in this case of equal payoffs, it may be assumed that the
bureaucrat will prefer to be "honest," and thus not ask for a
bribe.
* If q < [q.sub.1] then the bureaucrat knows that if he/she asks
for a bribe then the firm will start a negotiation and the final bribe
will be [b.sup.NB]. Then, at stage 2 the bureaucrat asks for a bribe if
and only if the bureaucrat's payoff on asking for a bribe is
greater than his/her payoff if he/she does not.
[[eta].sup.(B).sub.4] > [[eta].sup.(B).sub.2] [??]
w + (([a.sub.M] - [a.sub.T])k/2) - (q(c + m)k/2) > w
that holds [for all]q < [q.sub.1]. Thus we can conclude that if
q < [q.sub.1] then the bureaucrat asks for a bribe which the firm
accepts.
(1) At stage 1, the firm has to decide whether to present the
project.
* If q [greater than or equal to] [q.sub.1] then the firm knows
that if it presents a project no bribe will be asked. Should it decide
not to submit the project, its payoff will be equal to [a.sub.T]k,
whereas if it decides to submit its project, its payoff will be equal to
[a.sub.M]k - sk. Therefore, it will present the project if and only if
[[eta].sup.(F).sub.2] > [[eta].sup.(F).sub.1] [??] ([a.sub.M]k -
sk) > [a.sub.T].
The previous inequality is always verified by hypothesis.
* If q < [q.sub.1] then the firm knows that the bureaucrat will
ask for the bribe which it will accept. Should it decide not to submit
the project, its payoff will be [a.sub.T]k, whereas should it decide to
submit the project and to pay the bribe to the bureaucrat, its payoff
will be (([a.sub.M] - [a.sub.T])k/2) - sk - (q(c + m)k/2). Thus the firm
decides to submit the project if and only if
[[eta].sup.F.sub.4] [greater than or equal to] [[eta].sup.F.sub.1]
[??] (([a.sub.M] - [a.sub.T])k/2) - sk - > (q(c + m)k/2) [greater
than or equal to] [a.sub.T]k
which is verified if and only if
q [less than or equal to] ((([a.sub.M] - [a.sub.T]) - 2s)/(c + m))
= [q.sub.2].
Because [q.sub.2] < [q.sub.1] and since we assumed that
[q.sub.1] [less than or equal to] 1, then [q.sub.2] [less than or equal
to] 1. From now on we assume that [q.sub.2] > 0, that is half of the
surplus (as the difference than the returns of the two productivity
sectors) must be greater than the project cost (see Appendix A3 for the
other cases).
EQUILIBRIA UNDER ALL PARAMETER CONDITIONS
If (c + m) [greater than or equal to] 2s we obtain the following
five cases depending on parameter conditions:
If (c + m) < 2s we obtain the following five cases depending on
parameter conditions:
[TABLE A1 OMITTED]
[TABLE A2 OMITTED]
APPENDIX B
TABLE A3
Data and Sources
GDP at market prices 1980-2003: PROMETEIA
2000
Gross Fixed Investment 1980-2003: PROMETEIA
at 2000 prices
Corruption level 1980-2003: ISTAT
"Annuario Statistico e
Giudiziario"
various years
Criminal judges 1980-2003: Ministero
dell'Economia e delle Finanze
"Dipendenti delle
Amministrazioni Statali"
various issues
Police forces 1980-2003: Ministero del
Tesoro
"Dipendenti delle
Amministrazioni Statali"
various issues
Population 1980-2003: PROMETEIA
Public infrastructures 1980-2003: PROMETEIA
spending
at 2000 prices
Public consumption 1980-2003: PROMETEIA
at 2000 prices
Human capital 1980-2003: Costantini and
Destefanis (2009)
Notes: The legal statistics of ISTAT are one of the main
sources for region-based corruption analysis. Corruption
crimes fall into two classes of crimes considered by ISTAT.
The first class includes crimes by public officials considered
by the criminal code (arts. "314" and "322") and referred
to as embezzlement of public funds or misappropriation
(art. "324"); the second class concerns private interests
in official deeds. The data considered in this study refer
to the total number of crimes classified by ISTAT with
classification numbers from "286" to "294," namely: "286"
Embezzlement of public funds; "287" Embezzlement by
drawing profit from another's error; "288" Misappropriation
to the damage of private individuals; "289" Extortion; "290"
Corruption for official deeds; "291" Corruption for deeds
contrary to official duties; "292" Corruption of a party
in charge of a public service; "293" Corruptor's liability;
"294" Incitement to corruption; Police forces data include:
Arma dei Carabinieri (paramilitary police) and Polizia di
Stato (state police) (see "Conto Annuale," "Dipendenti delle
amministrazioni statali", codice "9," Ministero del Tesoro);
Judges data include several categories (see codice "12").
APPENDIX C
TABLE A4
Descriptive Statistics
No. of
Observa
Variables tions M Min Max SD
GDP 479 4.4 -9.6 5.6 0.559
(growth rate)
M, 478 2.230 0.967 3.204 0.336
M2 460 5.446 0.693 7.577 1.155
M3 460 8.448 5.575 10.822 1.010
Investment/GDP 480 0.228 0.142 0.448 0.056
Public 480 0.248 0.129 0.396 0.067
consump
tion/GDP
Human capital 480 7.224 4.498 9.144 1.060
Notes: The growth rate of real per capita GDP is
expressed in percentages. With reference to public
consumption and investment, the unit of measurement used is
millions of Euro at 2000 constant prices. [M.sub.1],
[M.sub.2], and [M.sub.3] indicate the three measures of
monitoring.
APPENDIX D
We consider the ratio between the number of detected corruption
crimes and the estimated total number of corruption crimes. The number
of reported corruption crimes is both a function of the corruption level
and a function of the level of prevention in place to reduce the
phenomenon. The probability of being detected, q, may be estimated by
the ratio between detected corruption crimes, [C.sub.o], and the
estimated total number of corruption crimes, [C.sub.e]:
(A4) q = [C.sub.o]/[C.sub.e]
Most econometric studies find that corruption is a function of
several variables (the legal system, government intervention,
probability of being detected, etc.). Therefore, we can define the
estimated total number of corruption crimes:
(A5) [C.sub.e] = A * IP
where IP is public infrastructure spending, and the constant A
represents all the other variables which affect corruption. The
rationale for focusing on public infrastructure spending is that
activities surrounding public works construction are the classic locus
of illegal monetary activities between public officials, both elected
and appointed, and businesses. Although corruption occurs in settings
other than public works contracting, the process of public works
contracting is, because of inherent informational asymmetries,
especially vulnerable, as substantial empirical and theoretical
literatures suggest (see McMillan 1991; Porter and Zona 1993).
We assume that q is a nonlinear function of the monitoring level:
(A6) q = [Monitoring.sup.[alpha]]
By substituting Equations (A5) and (A6) in Equation (A4), we have:
(A7) [Monitoring.sup.[alpha]] = [C.sub.o]/(A * IP) [??]
(A8) [C.sub.o]/IP = A * [Monitoring.sup.[alpha]]
Taking logs, Equation (A8) is written as follows:
(A9) log[C.sub.o]/IP = logA + [alpha]logMonitoring
Then we use log[C.sub.o]/IP as a proxy for the dynamic of
logMonitoring:
logmonitor = logdetected corruption crimes/ x public infrastructure
spending.
REFERENCES
Acemoglu, D. Introduction to Modern Economic Growth. Princeton, NJ:
Princeton University Press, 2009.
Acemoglu, D., and T. Verdier. "Property Rights, Corruption and
the Allocation of Talent: A General Equilibrium Approach." Economic
Journal, 108, 1998, 1381-403.
--. "The Choice between Market Failures and Corruption."
American Economic Review, 90(1), 2000, 194-211.
Aidt, T. "Corruption, Institutions, and Economic
Development." Oxford Review of Economic Policy, 25(2), 2009,
271-91.
Arellano, M., and S. Bond. "Some Test of Specification for
Panel Data: Monte Carlo Evidence and Application to Employment
Equations." Review of Economic Studies, 58, 1991, 277-97.
Arellano, M., and O. Bover. "Another Look at the Instrumental
Variables Estimation of Error Components Models." Journal of
Econometrics, 68, 1995, 29-51.
Auteri, M., and M. Costantini. "Fiscal Policy and Economic
Growth: The Case of Italian Regions." Review of Regional Studies,
34, 2004, 72-93.
Baltagi, B. Econometric Analysis of Panel Data. 3rd ed. Chichester,
UK: Wiley, 2005.
Barreto, R. "Endogenous Corruption in a Neoclassical Growth
Model." European Economic Review, 44, 2000, 35-60.
Barro, R. J. "Economic Growth in a Cross Section of
Countries." Quarterly Journal of Economics, 106, 199t, 407-44.
Blackburn, K., N. Bose, and M. E. Haque. "The Incidence and
Persistence of Corruption in Economic Development." Journal of
Economic Dynamics and Control, 30, 2006, 2447-67.
--. "Endogenous Corruption in Economic Development."
Journal of Economic Studies, 37, 2010, 4-25.
Blundell, R., and S. Bond. "Initial Conditions and Moment
Restrictions in Dynamic Panel Data Models." Journal of
Econometrics, 87, 1998, 115-43.
Campos, N., R. Dimova, and S. Ahmad. "Whither Corruption? A
Quantitative Survey of the Literature on Corruption and Growth."
CEPR Discussion Paper No. 8140, 2010.
Cellini, R., and A. E. Scorcu. "How Many Italies? What Data
Show about Growth and Convergence across Italian Regions, 1970-91."
Working Paper 251, Universit a di Bologna, 1995.
Cerqueti, R., R. Coppier, and G. Piga. "Corruption, Growth and
Ethnolinguistic Fractionalization: A Theoretical Game Model."
Journal of Economics, 106(2), 2012, 153-81.
Chang, T. S., and Y. Li. "A Macroeconomic Model on the Quality
of Government and Economic Performance." Economic Inquiry, 49,
2011, 174-84.
Costantini, M., and S. Destefanis. "Cointegration Analysis for
Cross-sectionally Dependent Panels. The Case of Regional Production
Functions." Economic Modelling, 26, 2009, 320-27.
Del Monte, A., and E. Papagni. "Public Expenditure, Corruption
and Economic Growth." European Journal of Political Economy, 17(1),
2001, 1 - 16.
Della Porta, D. "A Judges' Revolution? Political
Corruption and the Judiciary in Italy." European Journal of
Political Research, 39, 2001, 1-21.
Della Porta, D., A. Vannucci. Un Paese Anormale. Come la Classe
Politica ha Perso l'Occasione di Mani Pulite, Edizioni Laterza,
1999.
--. "The Determinants of Corruption in Italy: Regional Panel
Data Analysis." European Journal of Political Economy, 23(2), 2007,
379-96.
Ehrlich, I.. and F. T. Lui. "Bureaucratic Corruption and
Endogenous Economic Growth." Journal of Political Economy, 107(6),
1999, 270-93.
Fan, S. C. "Kleptocracy and Corruption." Journal of
Comparative Economics, 34, 2006, 57-74.
Friehe, T. "Correlated Payoffs in the Inspection Game: Some
Theory and an Application to Corruption." Public Choice, 137, 2008,
127-43.
Harstad, B., and J. Svensson. "Bribes, Lobbying, and
Development." American Political Science Review, 105(1), 2011,
46-63.
Hijzen, A., T. Inui, and Y. Todo. "Does Offshoring Pay?
Firm-Level Evidence from Japan." Economic Inquiry, 48, 2010,
880-95.
Huntington, S. P. Political Order in Changing Societies, New Haven,
CT: Yale University Press, 1968.
ISTAT. "Annuario Statistico e Giudiziario." Various
years.
--. "Conti Economici Nazionali." Various years*
--. "Conti Economici Regionali." Various years.
--. "Opere Pubbliche." Various years*
Keefer, P., and S. Knack. "Institutions and Economic
Performance: Cross-Country Test Using Alternative Institutional
Measures." Economics and Politics, 1995, 207-27.
Left, N. H. "Economic Development through Bureaucratic
Corruption." American Behavioral Scientist, 1964, 8-14.
Levine, R., and D. Renelt. "A Sensitivity Analysis of
Cross-Country Growth Regression." American Economic Review, 82,
1992, 942-63.
Lui, F. "An Equilibrium Queuing Model of Bribery."
Journal of Political Economy, 1985, 760-81.
Mauro, P. "Corruption and Growth." The Quarterly Journal
of Economics, 110(3), 1995, 681-712.
--. "The Persistence of Corruption and Slow Economic
Growth." IMF Staff Papers, 51 (1), 2004, 1 - 18.
McMillan, J. "Dango: Japan's Price-Fixing
Conspiracies." Economics and Politics, 3, 1991, 201-18.
Meon, P., and K. Sekkat. "Does Corruption Grease or Sand the
Wheels of Growth?" Public Choice, 122, 2005, 69-97.
Mishra, A. "Persistence of Corruption: Some Theoretical
Perspective." World Development, 34, 2006, 349-58.
Mocan, N. "What Determines Corruption? International Evidence
from Microdata." Economic Inquiry, 46, 2008, 493-510.
Murphy, K., A. Shleifer, and R. Vishny. "Why Is Rent-Seeking
So Costly to Growth?" AEA Papers and Proceedings, 1993, 409-14.
Porter, R. H., and J. D. Zona. "Detection of Bid Rigging in
Procurement Auctions." Journal of Political Economy, 101, 1993,
518-38.
Putnam, R. D. Making Democracy Work: Civie Traditions in Modern
Italy. Princeton, NJ: Princeton University Press, 1993.
Reed, R. W. "The Determinants of U.S. State Economic Growth: A
Less Extreme Bounds Analysis." Economic Inquiry,, 47, 2009,
685-700.
Roodman, D. "How to Do xtabond2: An Introduction to
'Difference" and "System' GMM in Stata."
Working Paper No. 103, Center for Global Development, 2006.
Rose-Ackerman, S. "The Economics of Corruption." Journal
of Public Economics, 4, 1975, 187-203.
--. Corruption and Govermnent. Cambridge, UK: Cambridge University
Press, 1999.
Shaw, P., M. Katsaiti, and M. Jurgilas. "Corruption and Growth
under Weak Identification." Economic Inquiry, 49, 2011,264-75.
Svensson, J. "Eight Questions about Corruption." Journal
of Economic Perspectives, 19(3), 2005, 19-42.
Windmeijer, F. "A Finite Sample Correction for the Variance of
Linear Two-step GMM Estimators." Journal of Econometrics, 126,
2005, 25-51.
Woo, J. "Productivity Growth and Technological Diffusion
through Foreign Direct Investment." Economic Inquiry, 47, 2009,
226-48.
(1.) For a recent review of Mauro's influential work see Shaw,
Katsaiti, and Jurgilas (2011).
(2.) Since corruption data are available, almost exclusively, at
the aggregate country level, most of the empirical literature has
focused on cross-country analysis. An exception is, for example, Mocan
(2008) who investigates the determinants of being asked for a bribe at
the individual level.
(3.) The implicit assumption in our work. as in most of the
literature, is that a higher monitoring level will reduce corruption.
However, as stressed by Svensson (2005), the institutions are weak in
many poor countries and, therefore, providing more resources for
monitoring activities might not be the right strategy in order to reduce
corruption. In fact, little evidence exists--Hong Kong and Singapore are
the most cited exceptions--that allocating more resources to monitoring
institutions will reduce corruption.
(4.) See also Cerqueti, Coppier, and Piga (2012) for a nonlinear
relationship between corruption and economic growth.
(5.) "When individuals are risk averse, this may imply that
there should be a risk premium associated with such stochastic streams
of income. This is not necessarily the case, however, when the following
three conditions are satisfied:
(1) there are many firms involved in research;
(2) the realization of the uncertainty across firms is independent;
(3) consumers and firms have access to a 'stock market.'
where each consumer can hold a balanced portfolio of various research
firms.
When this is the case, even though each firm's revenue is
risky, the balanced portfolio held by the representative household will
have deterministic returns," pp. 562-563.
(6.) The cost of the project submission to the bureaucrat is a
function of the investment. The underlying assumption is that, as the
size of the investment grows, the cost for the firm's bureaucratic
practices also grows.
(7.) It is assumed that no arbitrage is possible between the public
and the private sector allowing bureaucrats to become households, even
if their salary w is lower than the firm's net return. This may be
assumed since although individuals in the population (bureaucrats) have
a job, they have no access to capital markets, and therefore may not
become households.
(8.) Like Harstad and Svensson (2011), we assume that the penalty
increases in k because the penalty for such a serious crime is larger.
(9.) See Rose-Ackerman (1999) for details regarding the assumption
of a non-constant punishment function.
(10.) If agents are indifferent about whether to ask for a bribe or
not, they will prefer to be honest.
(11.) It is possible to consider q to depend on the number of
corrupt individuals or on other economic variables as, for example,
fiscal revenues. We deliberately avoid doing this in order to better
highlight how the monitoring level can influence the economic growth
rate (this approach has been also considered by Blackburn, Bose, and
Haque 2006, 2010; Del Monte and Papagni 2001; Fan 2006; Friehe 2008;
Mishra 2006). Furthermore, there is no clear-cut evidence regarding how
the level of corruption may endogenously affect the monitoring activity.
In fact, some literature considers that the greater the corruption rate,
the more likely the detection is (see, e.g., Barreto 2000); while other
authors assume that the more corrupt people there are, the less might be
the probability of being caught (see Mauro 2004; Murphy, Shleifer, and
Vishny 1993).
(12.) If q = 1 this stage of the game is never reached.
(13.) Leaving the bargaining strength of the bureaucrat and of the
firm generic, that is equal to a and 1 - [alpha], respectively, does not
change qualitatively the results of the static game and, therefore of
the dynamic one. The proof is available upon request.
(14.) We here focus on the case where parameters allow the greatest
number of equilibria depending on the level of monitoring by the State.
In Appendix A we show the results under all parameter conditions.
(15.) Mani pulite (Italian for "Clean Hands") was a
nationwide Italian judicial investigation into political corruption,
held in the 1990s. As Della Porta and Vannucci (1999) said "In
Italy the history of corruption does not begin (let alone end) on
February 17, 1992 (the official date that Clean Hands began). What
starts at that point are the extraordinary events of public exposure of
corruption, a scandal affecting the highest levels of the political and
economic system, causing the most serious political crisis of the
Italian Republic". The corruption system that is uncovered by these
investigations is usually referred to as Tangentopoli, or
"bribesville."
(16.) For a recent analysis on the determinants of growth. see Reed
(2009).
(17.) For further details see Blundell and Bond (1998).
(18.) For a discussion on endogeneity and system GMM estimator see,
e.g., Hijzen, Inui, and Todo (2010).
(19.) We would like to thank a referee for having raised this
point.
(20.) In particular the author finds evidence that for political
corruption, "the efficacy of the magistracy is, to a large extent,
determined by its degree of independence from political authority.... In
Italy, there is an unusually high level of (at least formal)
independence of the judiciary from political power. Mechanisms which can
allow for a certain degree of control by politicians over judges are not
available in Italy; the Constitution ensures that the magistracy could
not become the 'long arm of the government' as it had been
during the fascist regime."
RAFFAELLA COPPIER, MAURO COSTANTINI, and GUSTAVO PIGA *
* The authors would like to thank Francesco Busato, John Bennett,
Luca De Benedictis, Gianni De Fraia, Pietro Reichlin, Maria Cristina
Rossi. and seminar participants at the Department of Economic Sciences
at the University of Rome "La Sapienza" and at the Department
of Economic and Financial Institutions at the University of Macerata for
their very helpful suggestions. A special thanks to Francesco Nucci. The
usual disclaimer applies.
Coppier: Department of Economic and Financial Institutions,
University of Macerata, Macerata 62100, Italy. Phone +39 07332583245,
Fax +39 07332583206, E-mail raffaellacoppier@unimc.it
Costantini: Department of Economics and Finance, Brunel University,
Uxbridge UB8 3PH, UK. Phone +44 (0)1895 267958, Fax +44 (0)1895 269770,
E-mail Mauro.Costantini@brunel.ac.uk
Piga: Department of Business, Government, Philosophy Studies,
University of Rome-Tor Vergata, Rome 133, Italy. Phone -t-39 0672595701,
Fax +39 062020500, E-mail gustavo.piga@uniroma2.it
doi: 10.1111/ecin.12007
TABLE 1
Growth and Monitoring (All Italian Regions,
1980-2003)
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCCI]
Variables [M.sub.1] [M.sub.2]
ln([y.sub.i,t-1]) -0.082 -0.088
(2.83) *** (2.84) ***
ln([monitor.sub.i,t-1])) -0.079 -0.099
(2.72) *** (3.09) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.013 0.012
(3.17) *** (2.40) **
Inv 0.161 0.174
(3.04) *** (3.05) ***
conpa 0.263 0.223
(2.58) ** (2.42) **
[h.sub.it] 0.050 0.058
(3.13) *** (2.63) **
constant 0.321 0.334
(1.34) (3.04) ***
Sargan test: [chi square] 16.92 16.79
p values 921 0.928
z-statistics 0.245 0.235
p values 0.749 0.768
No. of observations 475 457
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCCI]
Variables [M.sub.3]
ln([y.sub.i,t-1]) -0.096
(2.91) ***
ln([monitor.sub.i,t-1])) -0.123
(2.93) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.009
(3.00) **
Inv 0.168
(2.95) ***
conpa 0.201
(2.37) **
[h.sub.it] 0.059
(2.81) ***
constant 0.310
(1.35)
Sargan test: [chi square] 16.35
p values 0.937
z-statistics 0.456
p values 0.659
No. of observations 457
Noter: Instruments used for Equation (12) are [DELTA]ln ([y.sub.i,t-
2]), [DELTA]ln([monitoring.sub.i,t-2]),
[DELTA][(ln([monitoring.sub.i,t-2])).sup.2], [DELTA]In[v.sub.i,t-1],
[DELTA][conpa.sub.i,t-1] and [DELTA][h.sub.it-1]. t-Statistics are in
parentheses. z-statistics indicate Arellano-Bond test for second-
order autocorrelation in the first-difference residuals. ***
Significant at 1%; ** significant at 5%; * significant at 10% levels,
respectively.
TABLE 2
Growth and Monitoring (Northern Italian
Regions, 1980-2003)
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII]
Variables [M.sub.1] [M.sub.2]
ln([y.sub.i,t-1]) -0.068 -0.073
(2.96) *** (2.81) ***
ln([monitor.sub.i,t-1]) -0.078 -0.095
(3.01) *** (2.97) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.020 0.016
(2.86) *** (2.67) **
Inv 0.183 0.199
(3.21) *** (2.93) ***
conpa 0.276 0.237
(2.56) ** (2.52) **
[h.sub.it] 0.053 0.052
(3.12) *** (2.26) **
constant 0.341 0.352
(1.48) (3.01) ***
Sargan test: [chi square] 16.17 16.41
p values .948 .924
z-statistics 0.225 0.400
p values .779 .700
No. of observations 192 187
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII]
Variables [M.sub.3]
ln([y.sub.i,t-1]) -0.085
(2.93) ***
ln([monitor.sub.i,t-1]) -0.0122
(2.84) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.008
(2.67) **
Inv 0.178
(2.58) **
conpa 0.232
(2.64) **
[h.sub.it] 0.050
(2.38) **
constant 0.315
(1.31)
Sargan test: [chi square] 16.51
p values .927
z-statistics 0.435
p values .690
No. of observations 187
Notes: Instruments used for Equation (12) are
[DELTA]ln([y.sub.i,t-2]), [DELTA]ln([monitoring.sub.i,t-2]),
[DELTA][(ln([monitoring.sub.i,t-2])).sup.2], [DELTA]ln[v.sub.it-1],
[DELTA][conpa.sub.it-1] and [DELTA][h.sub.it-1]. t-Statistics are in
parentheses. z-statistics indicate Arellano-Bond test for second-
order amocorrelation in the first-difference residuals.
***Significant at 1%; ** significant at 5%; * significant at
10% levels, respectively.
TABLE 3
Growth and Monitoring (Central Italian
Regions, 1980-2003)
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII]
Variables [M.sub.1] [M.sub.2]
ln([y.sub.i,t-1]) -0.073 -0.079
(2.81) *** (2.93) ***
ln([monitor.sub.i,t-1]) -0.071 -0.102
(2.84) *** (2.91) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.017 0.011
(2.83) *** (2.75) ***
Inv 0.109 0.139
(3.03) *** (3.16) ***
conpa 0.214 0.209
(2.09) * (2.01) *
[h.sub.it] 0.029 0.031
(2.42) ** (2.39) **
constant 0.402 0.389
(1.61) (1.46)
Sargan test: [chi square] 15.90 16.02
p values .964 .951
z-statistics 0.228 0.329
p values .772 .742
No. of observations 94 87
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII]
Variables [M.sub.3]
ln([y.sub.i,t-1]) -0.094
(2.76) ***
ln([monitor.sub.i,t-1]) -0.116
(3.29) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.007
(2.33) **
Inv 0.136
(2.52) **
conpa 0.201
(2.05) *
[h.sub.it] 0.036
(2.77) **
constant 0.401
(1.42)
Sargan test: [chi square] 15.02
p values .979
z-statistics 0.328
p values .744
No. of observations 87
Notes: Instruments used for Equation (12) are
[DELTA]ln([y.sub.i,t-2]), [DELTA]ln([monitoring.sub.i,t-2]),
[DELTA][(ln([monitoring.sub.i,t-2])).sup.2], [DELTA]ln[v.sub.it-1],
[DELTA][conpa.sub.it-1] and [DELTA][h.sub.it-1]. t-Statistics are in
parentheses. z-statistics indicate Arellano-Bond test for second-
order amocorrelation in the first-difference residuals.
*** Significant at 1%; ** significant at 5%; * significant at
10% levels, respectively.
TABLE 4
Growth and Monitoring (Southern Italian
Regions, 1980-2003)
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII]
Variables [M.sub.1] [M.sub.2]
ln([y.sub.i,t-1]) -0.103 -0.101
(3.21) *** (3.26) ***
ln([monitor.sub.i,t-1]) -0.070 -0.101
(2.92) *** (2.89) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.013 0.008
(3.25) *** (2.67) ***
Inv 0.100 0.101
(2.13) ** (2.24) **
conpa 0.199 0.198
(1.48) (1.83) *
[h.sub.it] 0.011 0.021
(1.83) * (1.75) *
constant 0.378 0.362
(1.61) (2.65) **
Sargan test: [chi square] 16.91 17.10
p values .919 .898
z-statistics 0.268 0.202
p values .740 .792
No. of observations 189 183
Dependent Variable:
[MATHEMATICAL EXPRESSION
NOT REPRODUCIBLE IN ASCII]
Variables [M.sub.3]
ln([y.sub.i,t-1]) -0.108
(3.18) ***
ln([monitor.sub.i,t-1]) -0.121
(3.18) ***
[(ln([monitor.sub.i,t-1])).sup.2] 0.007
(2.33) ***
Inv 0.102
(2.17) **
conpa 0.178
(1.82) *
[h.sub.it] 0.012
(1.71) *
constant 0.328
(1.54)
Sargan test: [chi square] 16.87
p values .919
z-statistics 0.209
p values .790
No. of observations 183
Notes: Instruments used for Equation (12) are
[DELTA]ln([y.sub.i,t-2]), [DELTA]ln([monitoring.sub.i,t-2]),
[DELTA][(ln([monitoring.sub.i,t-2])).sup.2], [DELTA]ln[v.sub.it-1],
[DELTA][conpa.sub.it-1] and [DELTA][h.sub.it-1]. t-Statistics are in
parentheses. z-statistics indicate Arellano-Bond test for second-
order amocorrelation in the first-difference residuals.
*** Significant at 1%; ** significant at 5%; * significant at
10% levels, respectively.
