Job Rotation and Tax Administration in the Developing
Countries
December 20, 2003
Job Rotation, Corruption, and
Tax Administration in the Developing Countries (Do Ngoc
HuynhVisiting Scholar,
Policy Research Institute, Ministry of Finance,
JapanAbstract
This paper is purposed to make a theoretical model to illustrate
the role of job rotation in reducing tax evasion and corruption and
consider an appropriate framework to apply the job rotation system
in tax administration in the developing countries. For the
theoretical model, we consider the role of rob rotation in a simple
framework based on the principal-agent and cooperation relations in
the theory of tax administration, and find that there would be an
optimal time for tax inspectors to work with the same taxpayers
that maximizes the expected government revenue. This would play a
role of the cornerstone for applying the job rotation system in tax
administration. Then, based on the Japanese experience, we discuss
the issue of how to apply the job rotation system in tax
administration with emphases on difficulties, feasibility, and
necessary conditions for the implementation process under the
context of socio-economic conditions in the developing
countries.
Keywords: Rotation; Specification; Tax Evasion; Corruption; Tax
AdministrationJEL Classification: J53, L22, C70, K42, D73
I. Introduction
The issue of tax evasion and corruption in tax administration
has always been one of the most disputed topics for policy-makers
and economists in the developing world nowadays. In the developed
countries, where the tax system is quite functioned in the
condition of high economic development, corruptions in general and
corruption in tax administration in particularly are suppressed at
relatively low levels. Being different from the situation of
corruptions in these countries, it is widely accepted that the
phenomenon of corruptions has been appeared everywhere in the
developing countries, especially in the process of tax compliance.
There are, generally, many reasons for the popularity of tax
evasion and corruption in these countries. However, they could be
divided in three main groups consisting of the tax policy, the
organization and management of tax administration, and the checking
system (Das- Guppta et al. (1999)). Regarding to the tax policy, in
the condition that the government faces many difficulties and
obstacles in the initial stages of economic development, the major
measures of reducing tax evasion and corruption, such as reducing
tax burden, eliminating preferential treatments, increasing
incentives for tax inspectors, and setting a system of heavy
penalties on tax evasion and corruption, have certain limits in
their real impacts. On the other hand, the checking system might be
too expensive and biased in the condition of limited budget
resources and asymmetric information in these economies. Therefore,
the system of organization and management of tax administration
would play an important role in reducing disastrous behaviors in
the process of tax compliance, including tax evasion and
corruption, in the developing countries.
Regarding to the topic of organizational system in general, as
well known in the theory of industrial organization and the
practice of the Japanese economy, the introduction of the job
rotation system might play an important role in improving the
overall efficiency of the whole system. Although job rotation has
obvious disadvantages in some important aspects, of which, the two
most important ones are the deficiency of breaking down
specialization and the difficulty in application, it is noted for
its main benefits from positive effects on career development,
motivations/renovations, flexibility, and reductions in working
stresses, boredom, and monotony, among others (Triggs and King
(2000), Cosgel and Miceli (1998)). This also explains why the
application of the system has been still limited in certain fields,
sectors, or in some countries. When choosing between two
organizational schemes of job specialization (specific/assignment
jobs) and rotation, there would be usually trade-offs between
advantages and disadvantages of these schemes that need to be
seriously considered. Generally, job rotation is supposed to be the
superior choice in cases of works requiring broad experiences,
highly pressured works, or not so high concerned costs in both
material and mental aspects.
In this paper, we explore a special advantage of job rotation in
reducing tax evasion and corruption and consider the trade-off
between the efficiency and the deficiency of job rotation in tax
administration in a simple theoretical framework of tax evasion and
corruption for developing countries. Based on the fact of these
economies, we assume that the tax inspectors probability of finding
out tax evasion and the probability of cooperation between the
taxpayer and the inspector are increasing functions of the time
that the inspector does his job with the same taxpayer. As the time
increases, the possibility of finding out tax evasion increases,
and therefore, the expected government revenue would be increased.
However, the probability of cooperation between the taxpayers and
the inspector also increases along with the time. This would reduce
the expected government revenue. The solution of this trade-off
relation would be the optimal level of the time when its positive
revenue effect is offset by its negative cooperation effect. This
theoretical model clearly illustrates the superiority of job
rotation in tax administration. However, it would be not easy to
apply the system in the developing countries, although its
advantage in reducing tax evasion and corruption is firmly
recognized in the theoretical model. Therefore, we also need to
analyze and discuss the framework of job rotation in order to make
the job rotation system feasible and efficient in these
economies.
The remaining of the paper is then organized as follows. In
section II, we make a brief literature survey on job rotation and
tax evasion and corruption. The theoretical model illustrating the
important role of job rotation in reducing tax evasion and
corruption is presented in section III. Then, we will discuss the
framework of job rotation in tax administration for developing
countries in section IV. Finally, section V makes conclusions.
II. Literature survey
Job rotation and corruption in tax administration are usually
considered as two separate topics in the recent economic
literature. Job rotation is mostly considered in the context of
industrial organization at the firm level. The relation between job
rotation and corruption has been rarely discussed. On the other
hand, in the literature on tax compliance, most of studies concern
the aspects of tax burden, penalties, and incentives. There has
been also not much attention on the job rotation system and its
effects on reductions in tax evasion and corruption. Therefore,
before combining these two issues into an integral framework, it is
necessary to make a brief outline of recent studies in these
fields.
1. Job rotation
There are many definitions of job rotation; however, there might
be divided into two main types: the narrow- and the wide
definitions of job rotation, as presented in the following
figure.
Figure 1: Job rotation: the narrow and wide definitions
The inside rotation
The outside rotation
Organization A
Organization B
People Works
People Works
A1 X1
B1 Y1
A2 X2
B2
Y2
A3 X3
B3
Y3
Narrow definition
Narrow definition
Wide definition
Regarding to the narrow definition, job rotation is considered
as transfers of employees between jobs in an organization. In this
case, job rotation is happened only inside the organization, then,
it might be called the inside rotation. However, there are also the
case of exchanging employees in certain fields/jobs between
different organizations, the outside rotation, such as the case of
rotations between government agencies and between government
agencies and private organizations, as observed in the Japanese
economy. Then, the wide definition of job rotation includes both
the inside rotation and the outside rotation - the case of crossing
transfers of employees between different organizations. Of course,
the difference between these definitions is only relative because
job rotation between divisions/departments in an organization or
between corporations in a holding group is the inside rotation for
the whole organization but the outside rotation for smaller units
inside the organization.
In organizations applying the system of job rotation, employees
are usually not remained in the same jobs/positions permanently but
are to be transferred to do other jobs/positions at some
frequencies. They usually do not return to former jobs. There are
many arguments on the merits and demerits of job rotation.
Regarding to the merits of job rotation, most of the studies
concern theories of industrial organization and management system
in the aspect of relationship between rotation and career
development. Studies on career development have recognized the
importance of work experiences that the career motivation can be
influenced by experiences gained through rotation (London (1983),
Morrison and Brantner (1992)). Job rotation is also argued to play
important roles to job learning that can be viewed as an
environmental strategy for employees career development. The
process of employees human capital accumulation is believably
dependent on their working experiences: the more they are rotated,
the more they learn. Hence, in the corporate management strategy,
job rotation is also considered as an important instrument for
development (Sonnenfeld and Peiperl (1988)).
For managers who are presumed to perform a variety of different
roles, rotation would be very important in the context of
developing managers as generalists who can understand and do jobs
concerning a vast of issues in the concerned fields. Empirical
studies also find that work experiences were related to executives
career opportunities, learning, and changes in abilities,
attitudes, and motivation (Guthrie and Olian (1991)). Practically,
large organizations, such as multi-national corporations,
frequently use rotation to develop their networks of managers.
There is also another aspect that organizational socialization is
considered as a process of information acquisition though
rotations, as showed in Ostroff and Kozlowski (1992). Job rotation
may enhance career development because of the adjustments and
knowledge acquisition that new jobs require. On-the-job factors,
such as co-workers, supervisors, and experiences are considered
more important than off-the-job factors such as formal training and
social activities. Therefore, setting up the organizational
framework of varying work experiences through rotation might be
considered as one important component for the process of career
development.
Studying the cost and benefit of job rotation, there are also
discussions concerning the issue of job matching. In the job
matching studies (Jovannovic (1979), Miller (1984)), jobs and
employees should be matched in the best way possible because of the
job specificity of human capital. However, the proficiency of
employees in different jobs could be only known if they actually
try such jobs. Therefore, job rotation would be an effective
instrument for optimally allocating labor forces in case of
asymmetric information in the labor market. Looking at each
employee performing different jobs, the employer would find the job
that is most suitable for him. Concerned the aspect of job
specificity of human capital, there are also some striking
arguments that job rotation would improve productivities of working
more than specialization. Considering job rotation as a learning
mechanism in a clear theoretical framework of the job specificity
of human capital, Ortega (2001) compared job rotation to an
assignment policy, and also showed that job rotation would be a
better learning mechanism than specialization when there is little
prior information about the relative profitability of different
jobs/ works. Relative gains from job rotations are higher than
specialization when the initial uncertainties about employees and
technologies become larger. Empirical works also support the view
that firms using new technologies are likely to use job rotations.
Beside the above economic merits, there are also other advantages
of job rotation in the social and welfare aspects. They are
reductions in boredom and monotony, work stresses, and CDT
(cumulative trauma disorders) It is widely believed that a
proportion of employees would feel bored and monotonic with the
same works for long periods. On the other hand, some of them,
especially the young, want to challenge new jobs in new working
environments and enjoying new working relations with new people. In
many cases, working in the same place and maintaining long
relations with the same colleagues would cause stresses and CDT for
employees, which are negatively affect their utilities. Therefore,
job rotation could be considered as a possible choice to improve
employees welfare that might not concern economic incentives.
On the demerits of job rotation, they have been considered in
many aspects. The biggest and foremost disadvantage of job rotation
concerns specialization. As widely accepted in the economic
literature, from classical economists like A. Smith or K. Marx,
specialization is one of the major engines for renovations and
enhancing working efficiency. Job rotation clearly restrains the
degree of specialization. Therefore, cost-benefit analyses of
choices between rotation and specialization would be very
important. Job rotation should be only applied when it produces
benefits that are higher than losses from reductions in
specialization (Cosgel and Miceli (1998, 1999)). In the fields
requiring specialists with deep knowledge or working skills and
experiences in the same jobs, such as some works in high-tech
industries and scientific researches, job rotation at relatively
frequent levels might be questioned.
Besides, job rotation has close relations with the work
structure, workplace relation and culture, and employees behaviors.
As usual, experienced employees do not want to do new types of jobs
or handover their currently good conditions or share their
knowledge and experiences of working to newcomers. It is also very
difficult to identify appropriate job for each people in the
organization in the process of rotating. On the other hand, there
are usually not small costs of rotating, including both
material-and non-material costs, such as changes in the living
environments, customs and cultures, especially for the employees
side (MacLeod and Kennedy (1993), Triggs and King (2000)). This is
why the system of job rotation has not been popular in the world
economy, and only relatively big organizations with a
well-performed management system can apply the system efficiently.
Generally, considering the job rotation system to apply, all
aspects of cost-benefit analyses, the organizational structure, the
system of work evaluation, as well as the period and timing of
rotation should be carefully reviewed.
It is also necessary to consider the job rotation system in
Japan as a fairly successful and unique example in the economy-wide
scale of application. The Japanese economy started to apply and
develop the system in the 1950s. There are many studies indicating
important roles of the system to the Japanese miracle in the high
growth period. The system of lifetime jobs with rotations is one of
the most distinguished features of the Japanese economy. Not only
the private sector, the system of job rotation has been applying in
government agencies, and up to present, it has generally been still
rather functioned. Why does the job rotation system go well in
Japan but not in somewhere else in the world? As pointed out in
Dutton (1987), The strong traditional permanent employment in
larger Japanese firms, especially for the managerial group, makes
job rotation possible and feasible. Historically, the Japanese
economy, especially the labor market, has been different from other
economies in many aspects. A typical Japanese employee usually
stays in the same organization until their retirements. During his
period of staying in the same organization, he experiences many
jobs, working positions, and moves towards the top of the pyramid
of the management structure in that organization. This is so-called
the lifetime job system, and considered as unique in the world.
Ouchi (1981) also certified the non-specialized career path as one
of the fundamental features of Japanese firms.
Beside the management structure, there are many other factors
supporting the job rotation system in Japan. Among others, the wage
system in Japan is mainly based on the length of time working at
the organization (employees ages), therefore, the system of job
rotation would not cause much troubles concerning employees
earnings. On the other hand, job rotation has become a
custom/culture of business administration, and been widely accepted
by the public. In addition, Japanese people are noted for their
very diligences at the work. Their attitudes to works and
responsibilities for their works are extremely high. The
organizational socialization of the Japanese people is also notably
respected so that the implementation process is quite smooth and
functioned.
Overall, we see that almost studies analyzed the job rotation
system in the aspect of industrial organization and management
structure. At the firm level, although having disadvantages because
of the negative effect on specialization and the difficulty in the
implementation process, job rotation is noted for its advantages in
career developments/ working experiences, flexibility, reductions
of working stresses, boredom, among others. In practice, the
Japanese economy has been successfully applying the system due to
her special and unique socio-economic conditions at both micro and
macro levels. However, looking at the concerned literature, the
relation between job rotation and corruption, specifically
corruptions in tax administration, has generally still not
attracted much attention of studies on the field.
2. Tax evasion and corruption
Tax evasion and corruption in tax administration has been a
regular topic in the economic literature, as noted above. Regarding
to the theoretical aspect, there are many studies exploring the
principal-agent relation in tax administration and cooperative
games between taxpayers and inspectors. Historically, the first
study on tax evasion might be the seminal paper of Allingham and
Sandmo (1972). As showed in this study, it is necessary to audit or
investigate the taxpayers reported income in order to deter or
reduce evasion. The tax authority (auditor/inspector) is
responsible for making justice in tax compliance by finding out the
real taxable income and imposes a penalty rate (higher than the tax
rate) in case of tax evasions. However, in a hierarchical structure
where the principal delegates a task to agents, there would be the
cases these agents cooperate with taxpayers to lower their tax
liabilities. Then, after the collusion in a hierarchy was
formalized by using the principal-agent framework (Tirole (1986),
Kaufman and Lawaree (1993)), there have been emerged a bulk of
studies on tax evasion and corruption in tax administration.
In these studies, the possibility of cooperation between
taxpayers and inspectors would be appeared when the concerned cost
is relatively low. The problem of optimal tax scheme is usually
regarded in the aspects of tax schedule (burden), penalties and
incentives for taxpayers and inspectors, and schemes of auditing
and investigations. Besley and McLaren (1993) considered a
framework with both honest and dishonest agents, compared three
distinct remuneration schemes of reservation, efficient, and
capitulation wages that provide different incentives to the
inspectors. They characterized the conditions under which each
scheme generates the greatest amount of net tax revenues after
abstracting administration cost. On the other hand, in Mookherjee
and Pngs (1995) paper, they considered corruptible agents with a
moral hazard problem that inspectors have to exert costly
non-observable efforts to disclose evasions, and found that
although bribes encourage the inspector to monitor, it should be
better when compensation policies are adjusted to eradicate them.
Then, Hindriks et al. (1999) studied a model consisting of all
dishonest agent with the distinction of setting flexible tax
schemes and allowing extortion that the inspector might overstate
the taxpayer income. They also proved that imposing progressive
taxes without measures of presenting evasion and corruption would
make bribes on high-income reports. Recently, Polinsky and Shavel
(2001) analyzed a framework of bribery, distortion, and framing,
and found two ways to reduce corruption, consisting of imposing
sanctions and rewards for reporting violations with optimal
penalties and rewards schemes. Now, we consider the situation of
tax compliance and tax administration in the developing countries
for looking stylized facts to build the theoretical model in the
next section. Tax administrations in the developing countries have
arguably been mostly corrupted. As presented above, difficulties
and obstacles in the socio-economic conditions in these economies
make the picture different from that in the developed economies. In
these economies, beside reasons of tax evasion and corruption
concerning the tax policy and the checking system, the weak tax
administration system itself has been also an important background
for developments of tax evasion and corruption. There would be two
important issues concerning the tax administration: the
organizational structure and management system, and the officers
and taxpayers attitudes. Regarding to the organizational structure
and management system, shortages of modern management frameworks
and modern technologies assistances make the system out-of-date,
less flexible and transparent, and inefficient. The reporting
system and the information processing system, which play major
roles in tax enforcement, are usually weak and inadequate. For the
tax administration purposes, both systems of inside relations and
outside relations are always questionable in all aspects. Tax
inspectors are usually authorized to work with their taxpayers for
long periods without a strictly reporting system and frequently
investigations. This might be considered as one of the main reasons
for cooperation between inspectors and taxpayers that lead to tax
evasions and corruptions. Even in the case of investigation by
higher-level officials, there are usually further
bargaining-cooperation steps in the hierarchical organization. That
is why it would be not exaggerated to say that tax evasion and
corruption in the developing countries is systematic and
serious.
Regarding to the tax officers and taxpayers attitudes, because
average levels of per capita income and living standards in the
developing countries are relatively low, most of taxpayers and tax
officers would have motivations for tax evasion and corruption at
some certain levels. Moral hazards and adverse selection might be
happened in tax compliance in both sides of taxpayers and tax
officers. For taxpayers, under the condition of popularity of tax
evasion, they would be forced out of the game if being honest in
the competitive market. In the tax officers sides, because of low
wage, by touching with opportunities to make money, they would take
their chances. The honest officers would be also forced out of
their works, partly because of low income, partly because of being
strange to their colleagues. Even in the case of initial honest tax
officers, after a long time touching with a possibility of making a
large amount of money in a relatively safe way in the condition of
low salary, they would be engaged in corruptions.
Looking overall, we see that the organization and management
aspect would be very important in reducing tax evasion and
corruption in the developing countries. The model developed in the
next section is then based on the theoretical arguments of job
rotation and tax compliance in the recent economic literature and
the stylized fact in tax administration in these countries that the
structures of tax administration are relatively static in terms of
rotating and exchanging their officers. This is considered as one
of the major reasons for cooperation between taxpayer and tax
officers to engage in tax evasion and corruption in the difficult
socio-economic conditions. This model is different with the current
literature on job rotation in the aspect that it focuses on
analyzing the aspect of cooperation between taxpayers and tax
officers. Comparing with the existed theoretical studies on tax
evasion and corruption, we concentrate on the aspect of the time
(period) that tax officers touch the same taxpayers, which
considered as a necessary of cooperation for engaging in tax
evasion and corruption.
III. Theoretical Model
1. The framework
Based on the stylized facts of the developing countries, we use
a simple framework of studying tax evasion and corruption and
integrate the factor of the time required for cooperation between
agents into the framework to analyze effects of job rotation on tax
evasion and corruption. There are four types of agents in the
model: the government, tax inspectors, taxpayers, and auditors. The
government announces the tax scheme, consisting of the tax
schedule, penalties for the case of reporting income below the real
levels for both taxpayers and inspectors, incentives for the
inspectors to encourage them to do their jobs. The representative
taxpayer is subjective to the tax scheme announced by the
government. She is to maximize her payoffs, given the tax scheme.
We also assume that she is tax-evasion-driven, as popularly
observed in the developing world.
The inspectors are responsible for investigating whether the
taxpayers report precisely their real income levels or not. The
probability of finding the real income is assumed to be dependent
on the time that the representative inspector does his job with the
same taxpayer. This reflects the advantage of specialization in tax
investigations. On the other hand, the inspectors are
corruption-driven, which reflects the popular situation of
corruption in tax investigations in the developing countries. The
inspectors and taxpayers may negotiate to engage in tax evasion and
corruption. However, the process of negotiation is to depend on
mutual beliefs between them. This is also a distinguished feature
of this study. Besides other factors affecting the process of
negotiation, especially the profitability of cooperation, we
supposed that the length of the time working with the same partners
would be the main reason to strengthen the spirit of cooperation
between them, which lead them to negotiate to engage in tax evasion
and corruption (evasion for the taxpayer, corruption for the
inspector). This is the disadvantage of specialization in tax
investigations. We also already noted that even in the case of
honest inspectors, to work with the same taxpayers in the long time
would make them gradually becoming the dishonest.
Beside the tax inspectors being responsible for investigating
the taxable income, an independent auditing agency enters the
picture to make the justice of tax compliance. One this agency
audits tax reports, they will find the real income. However, the
cost of auditing is assumed quite large so that the probability of
auditing is rather small. On the other hand, because they do not
know which case is tax evaded and corrupted, the choice of auditing
tax reports would be incidental.
Looking at the framework, we see that job rotations for
inspectors are able to resolve the conflict between the advantage
and disadvantage of specialization in tax investigations. Job
rotation would significantly affect tax inspectors in many positive
and negative aspects, as the recent literature on industrial
organization suggested. However, in this paper, we focus on
analyzing its effects on reducing tax evasion and corruption under
the context of social economic environment in developing countries.
As regarded above, tax evasion and corruption is popular in these
countries. The tax policy and the systems of penalties and
incentives have limits in their real effects under the
socio-economic conditions of the developing world. Therefore,
besides other policies measures, the government can also apply the
system of job rotation to cut off mutually deep relations between
taxpayers and inspectors, which are considered as one of the main
reasons for tax evasion and corruption. In the model developed in
this section, we prove that job rotation would reduce the
possibility of cooperation between taxpayers and inspectors,
although it would have a natural problem concerning the inspectors
working proficiency.
2. The process
Based on the facts of tax administration, the process of tax
compliance is assumed to consist of three stages, as follows.
Stage 1: The government announces the tax scheme,
where are the tax rate, real (taxable) income, reported income,
penalties on the representative inspector and the representative
taxpayer (i = I, T), fixed wage and incentive rate for tax
inspector, respectively. This tax scheme, including the functions
of penalties and incentive, is popularly applied in the world
economy nowadays, especially the developing countries.
With a given level of the real income, the penalty functions are
strictly concave on the level of the reported income, as
follows,
Because the evasion amount is the gap between the real income
and reported income, the fine would be zero when the amount of tax
evasion is zero, and strictly increasing with the amount of tax
evasion. The fine level is higher than the tax liability for every
amount of tax evasion.
Stage 2: Income reporting, investigating, and negotiating
The representative taxpayer reports the level of income r, .
This level is set up to maximize her net income, given the tax
scheme and the inspectors probability of finding out the real
income. The representative inspector investigates to find the
reality of the reported income with the probability of finding out
the real income as follows,
Of which t is the length of the time that the inspector does his
job with the same taxpayer (called the time for short, hereafter).
The function (2) means that the inspector needs time to find out
the real income. The probability of finding out the real income is
an increasing function of t when . When this probability is equal
to unit. In other words, is the length of time for surely finding
out the real income,.
Now, we consider the process of negotiation between the taxpayer
and the inspector. There are two conditions for making negotiation
between them successfully. The first is that the cooperation would
be profitable for both of them. This means when they agree to
report some level of income and accept some level of bribe, both of
them are better off in terms of net income. On the other hand, the
negotiation process between the inspector and the taxpayer is also
supposedly dependent on the time, t. This is because they only
cooperate when the relation between them reaches certain levels of
closeness, deepness, and mutually believing and understanding. As
we noted above, they are all human being; therefore, the more the
time to work together, the higher the level of mutual belief and
close relation between them. The probability of negotiation is then
as follows,
We assume that which means that the limit of time for surely
cooperating is shorter than that for surely finding out the real
income. Besides, it is supposed that both the taxpayer and the
inspector are risk-neutral so that we can concentrate on analyzing
the effects of the time in the model. Stage 3: Auditing
The independent auditing agency is responsible for auditing the
preciseness of tax compliance. We assume that the probability that
the agency audits a tax report is , a given parameter. When
auditing, the agency always find out the real income. If the
reported income is less than the real income, both the taxpayer and
the inspector will be fined according to the announced tax scheme,
as presented above. The probability of auditing is assumed rather
small because the heavy cost of auditing, as regularly cited in the
economic literature and the real world.
3. Analyses
Based on the above process, we can describe the game in the tax
administration as that in Figure 1. The left hand is the case of no
cooperation with the probability of 1-p(t); meanwhile, the case of
cooperation with probability of p(t) is presented in the right
hand.
In this figure, we use the following signs: U is the net income
of the taxpayer, V is the net income of the inspector. For upper
signs, NA is for the case of not being audited; meanwhile A is for
the case of being audited. In the lower signs, NN is for the case
of no cooperation and the real income not recognized by the
inspector, NR is for the case of no cooperation and the real income
recognized by the inspector, and C is for the case of cooperation
between the taxpayer and the inspector.
Figure 1: The game in tax administration
1-p
p No cooperation
Cooperation
Considering the payoffs for agents in the game, namely, the
representative taxpayer, the representative inspector, and the
government, as presented in Figure 1, we need to analyzes both two
cases: (i) no cooperation between the taxpayer and the inspector,
and (ii), cooperation between the taxpayer and the inspector. In
the first case, the taxpayer faces the inspectors probability of
recognizing the real income,; meanwhile in the second case, they
mutually agree on the level of reported income. In each case, the
probability of auditing is the same, , as regarded above.
3.1. The case of no cooperation For the inspector, supposing
that the taxpayer reports the income r,, as presented above (r is
called the taxpayers reported income), in the case of not being
audited, because the probability of finding out the real income is
, the inspectors expected income would be as follows,
In the case of being audited, his expected income would be,
Therefore, with the probability of auditing , the expected
income of the inspector in the case of no cooperation is then,
We can see that this expected income in the equation (4) is
dependent on the time t and the reported income r.
Now, we consider the expected income of the taxpayer. Similarly,
for the case of not being audited, the expected income would
be,
Meanwhile, in the case of being audited, it would become,
With probability of auditing to find out the real income , the
taxpayers expected income would be as follows,
We can see in equation (5) that the condition of engaging in tax
evasion is,
whereis the expected gain for the taxpayer from tax evasion. In
the right hand of (6), the first expression is the expected gain
from evasion, and the second expression is the expected loss.
Therefore, tax evasion is happened only if the expected gain is
larger than the expected loss.
Considering the behavior of the taxpayer, she is to report the
income level that maximizes her net income (the after- tax and
penalties income). Then, the maximization problem for the taxpayer
would be . From (5) and (6), we have,
From the maximization problem of the taxpayer and the characters
of the function in (1), we have the optimum choice for the taxpayer
as follows,
The optimization level of the taxpayers reported income is
determined as described in the following figure.
Figure 2: The optimal level of the reported income
In equation (7), when is rather small. Then we have the rule of
1/2 as stated in the following proposition.
Proposition 1. In case of no cooperation, if the function of
penalty for the taxpayer is strictly convex (the penalty is always
higher than the tax revenue), when the probability of auditing is
rather small, the taxpayer will report the real income when the
inspectors probability of finding out the real income is from about
1/2.
Intuitively, because the penalty is always higher than the tax
liability, the rational taxpayer would report the real income when
she knows that the probability of the real income being recognized
by the inspector is from about 1/2. Otherwise, she would face the
expected loss from the tax evasion. If the probability is smaller
than this level, she will choose to report some level of income
less than the real income, as showed in (7).
Now, we consider changes in the optimal level of the reported
income with respect to t. From (1) and (7), we have:
where is determined by.Looking at (8), the level of reported
income would be increased with t until when t is high enough such
that the taxpayer would report the real income (Figure 3).
Figure 3: The dependence of the optimal level of the reported
income on t
From (8), at the optimum choice of we have,
of which . Equations (9a) and (9b) mean that whenthe expected
income of taxpayer is only negatively affected by the direct effect
of t on the inspectors probability of finding out the real income.
The indirect of t on the reported income is zero because is at the
optimal level of the reported income. On the other hand, whenthe
taxpayer would report the real income, and then, the reported
income would not be affected by t.Regarding to the trend of
changing in the inspectors expected income, we see that it is
dependent on the time, t, and the taxpayers reported income, as
regarded above. Then, when the reported income is , from (4), (7),
and (8), we have,
From equation (10), we see that whenthe inspectors expected
income would increase in t. Here, both the direct- and the indirect
effects of t on the income are appeared. When t increases, the
probability of finding out the real income increases; therefore,
the expected incentive income would increase. On the other hand,
the expected penalty would be also reduced. Whenthe taxpayer
reports the real income, and then, the inspectors expected income
would be not affected by t.
The expected (net) government revenue in this case is,
therefore, as follows,
The evolution of the expected government revenue is then,
Whenwe assume that,
In (13), the first expression in the right hand is positive, but
the second expression is negative. However, because the incentive
rate would be not so high and the probability of auditing is rather
small, as usual observed, the revenue effect is then higher than
the payment effect. Intuitively, increases in the government
revenue due to increases in the reported income are to be higher
than increases in the net payments to the inspector and the
taxpayer. This is also reasonable for the real world of tax
administration. The increase in the tax revenue would be not used
totally for paying the inspector because it would largely affect
the inequality in income distribution in the society.
3.2. The case of cooperation
Supposed that when the taxpayer and the inspector cooperate to
report some level of income r and agree on the level of bribe is b,
the taxpayers expected income would be dependent on the probability
of auditing, as follows,
Similarly, the inspectors expected income would be,
As usual, the maximization problem of cooperation is then to
maximize the expected joint income, as follows,
Then we have the optimal value of reported income in case of
cooperation,
When the probability of auditing is small and the incentive rate
is not so large, there would exit an optimal level of the reported
income, as determined in (17). We have the following
proposition.
Proposition 2. In case of cooperation, the optimal level of the
reported income is unique and not dependent on t.As showed in (16),
and (17), at low levels of r the gain from tax evasion would be
higher that the loss from penalties. However, because the functions
of penalties are convex, the amounts of penalties would increase
rapidly when r increases, and after some certain level of r,
increases in the gain from tax evasion would be less than losses
from increases in penalties. Because both functions of penalties
are assumedly not intersected, there would be a uniquely optimal
level of r that makes the expected (net) joint income maximized.
This level is not dependent on the time, t. Intuitively, when the
taxpayer and the inspector cooperate, they would not care about
anything else but the expected joint income.
The expected government revenue in the case of cooperation is as
follows,
As we can see in the above equation, there would not be the
revenue on penalties in the case that the inspector recognizes the
real income of the taxpayer because they cooperate to report an
optimal level of income that maximizes their expected joint
income.
The premium from cooperation would be divided based on
bargaining powers of each partner in the cooperative game. Assuming
that the bargaining powers are equal, the level of bribe would be
as follows,
Generally, the level of bribe is considered as the solution of a
Nash bargaining problem between the taxpayer and the inspector.
However, for the purpose of studying, we only concentrate on
analyzing corruption; therefore, we assume that their bargaining
powers are equal so that the premium from cooperation would be
equally divided between them.
As noted above, there are two factors deciding the process of
cooperation. One is that cooperation is beneficial for both
players. The other is the time that the inspector works with the
same taxpayer. Because we are dealing with the case of cooperation,
we only consider the later. Supposing they cooperate, regarding to
the first condition, the expected joint income from cooperation
should be higher than that from no cooperation. Therefore, the
condition for cooperation would be as follows (P is the expected
joint profit),
Otherwise, they would never cooperate. In case of cooperation,
because the value of the first parenthesis in (19) doest not depend
on t, we have,
Therefore, we can see that there might be two cases depending on
the parameters of the model: (i) cooperation is always beneficial;
and (ii), cooperation is only beneficial when t is large
enough.
For the second case, we assume that at some small level of t,
cooperation might not be beneficial (), especially when the
inspectors probability of discovering the real income is relatively
low at the initial period. When t increases, the inspectors
probability of discovering the taxpayers real income increases, and
therefore, cooperation will become beneficial (). Then, there might
exist some at which . When , the cooperation would not be
beneficial; meanwhile when , cooperation becomes beneficial.
Therefore, only when , the cooperation between the taxpayer and the
inspector is possible. We also assume so that there would always be
the possibility of cooperation.
3.3- The expected government revenue
As regarded above, there would be two possible cases of
cooperation: cooperation is always beneficial, and cooperation is
only beneficial when t is large enough. Then, it is necessary to
consider the expected government revenue for both these cases. The
expected government revenue should be understood as the
(comprehensively) expected government revenue when considering the
possibility of cooperation between the taxpayer and the inspector.
We suppose that the government pursues the objective of maximizing
the expected government revenue. The only way they can do that is
to change the time that the inspector does his work with the same
taxpayer. This could be implemented through introducing the system
of job rotation. Therefore, by considering the evolution of the
expected government revenue, and through the results of analyses,
we can perceive effects of the rotation system in pursuing the
objective of maximizing the expected government revenue.
Case 1: Cooperation is always beneficial
The expected government revenue is determined in the following
formula,
Looking at equations (22), when the change of G would depend on
the slope of and. Based on the evolution of these functions, we
have the following proposition.
Proposition 3. If cooperation is always beneficial, is concave,
p(t) is linear or convex, and when ,
There exists a unique value of that maximizes the expected
government revenue. This optimal value of t is determined where
Proof. Differentiating G when by t in equation (22), we have
The first term in the right hand of (23) reflects the positive
revenue effect of changing in, meanwhile the second term is the
negative cooperation effect of changing in p. Because cooperation
is always beneficial,
We also have,
When , if , from (26) we see that . This means that increases in
t near 0 from the right hand would have positive effects on G.
On the other hand, when , we have,
Therefore, there would be some value such that and changes the
sign from positive to negative when t passes across , which is
called the optimal time that maximizes the expected government
revenue. Because we assume that is concave and p(t) is linear or
convex, the first term in the right hand of (23) is decreasing,
meanwhile the second term is non decreasing, therefore, the value
of is unique (point E in Figure 4).
We can also describe the evolution of the expected government
revenue as that in Figure 5. We can see that when the probability
of cooperation is between zero and one, , the evolution of G would
depend on the evolution of and . Because and have the same sharp of
increasing in t, when if the condition in the proposition is
satisfied, the slope of would be higher than that of . On the other
hand, when , the slope of would be higher than that of . Therefore,
there would be some point that the slope of these curves is equal
(). Because is concave and p(t) is linear or convex, there would be
a unique point E at which the slopes of and are equal. The locus of
the expected government revenue would start from , reach the
maximum level at E (), and then, reduce to the value of from point
F (). When , the government revenue remains constant at because the
probability of cooperation is one.
Figure 4: Determination and uniqueness of the optimal time in
Proposition 3.
E
In Proposition 3, there would be the cases when the conditions
that is concave and p(t) is linear or convex are violated. In these
cases, the uniqueness of the optimal level of time would be
questioned. There might be several points/regions of t that make
and the sign of changes from positive to negative. They called the
locally optimal points/regions. In this case, the globally optimal
point/region would be chosen between these locally optimal
points/regions where the expected government revenue is maximized.
Figure 5: The evolution of the expected government revenue in
Proposition 3.
0 t* t
When the condition concerning t at near 0 from the right hand in
Proposition 3 is not satisfied, i.e., at ,
there might be two following possibilities:
(i) The gap between and is non-increasing in . That means,
In this case, the expected government revenue would be decreased
from to when t increases from 0 to . Hence, in this case, the
higher the frequency of rotation, the more the government revenue
could be collected. Even in this case, it could be possible to find
the optimal level of t by two possible methods. The first method is
to set up a threshold for p(t), as follows,
where is the threshold level from which there would be a
possibility for cooperation. On the other hand, the second method
is to add some cost for rotation at the beginning periods for new
jobs. It is easy to see that for the first case, is the optimal
level of t; meanwhile, in the second case, the optimal time would
be determined when the concerned cost of rotation become less than
the expected loss in revenue due to the possibility of
cooperation.
(ii) The gap between and is increasing in at some ranges in .
This case is similar to the case we considered above that the
conditions of is concave and p(t) is linear or convex in
Proposition 3 are violated. In this case, there might be several
points/regions of t that make and changes the sign from positive to
negative. They are the locally optimal points/regions. In this
case, the globally optimal point/region would be chosen between
these locally optimal points/regions where the expected government
revenue is maximized.
Case 2: Cooperation is only beneficial when t reaches a certain
level ()In this case, the expected government revenue is as
follows,
We have the following propositions.
Proposition 4. If cooperation is only beneficial when t is large
enough (), is concave, and p(t) is convex or linear, there always
exists a unique value of that maximizes the government revenue.
This optimal value of t is determined where
Proof. From equation (24), at two polarized point ,
Because when , we have,
On the other hand, we assume that is concave and p(t) is convex
or linear. Therefore, similarly to the case stated in Proposition
3, there would be a unique value of t* at which, or,
and changes the sign from positive to negative when t crosses
t*.
In this case, the expected government revenue can be presented
in Figure 5. Looking at this figure, we see that the curve is
always below the curve . Therefore, the expected government revenue
would be initially increased and then decreased in the determinant
region . Because is concave and p(t) is convex or linear p(t) is
convex or linear, the expected government revenue G would be
uniquely maximized at point E () where , where the slope of the
curve is equal to that of the curve (also equal to the slope of the
curve ). When , we have , the locus G is that of . At , it reaches
point A, and then, goes to point E () where the expected government
revenue is maximized. After that, it is gradually downward to point
B (), where the government revenue is . Beyond this point, the
expected government revenue remains at because the probability of
cooperation becomes unit (the taxpayer and the inspector cooperate
surely).
Figure 5: The evolution of the expected government revenue in
Proposition 4.
It could be easy to see that the optimal level of t is always
existed in this case (cooperation is only beneficial when t is
large enough). However, regarding to the uniqueness of the optimal
time, there would be several cases. Beside the case in Proposition
4, there would also be the cases at which the conditions that is
concave and p(t) is convex or linear are violated. The uniqueness
of the optional time is dependent on the sharp of these functions.
Similarly to the cases regarded above, there might be several
locally optimal points/regions of t that make and the sign of
changes from positive to negative. In this case, the globally
optimal point/region would be chosen between these locally optimal
points/regions where the expected government revenue is
maximized.
Looking overall, we can see that, in general, job rotation in
tax administration would have a positive effect on the expected
government revenue. The optimal time for rotating is dependent on
the parameters of the model. Specifically, it would be not
difficult to see that it depends on the functions of the
probability of cooperation,, and the inspectors probability of
finding out the real income, . Regarding to the former, when this
function is highly convex, the optimal time for job rotation would
be lengthened. By contrast, when it becomes relatively concave, the
optimal time would be shortened. By other words, the optimal length
of time would be dependent on the slope of this function, the more
the flatness of the function at the initial time, the higher the
optimal length of time is. For a given sharp of the possibility of
finding out the real income of the inspector, if the probability of
cooperation is low at the beginning time, the expected government
revenue will increase gradually and reach its optimal level at a
relatively large t.
On the other hand, the function of probability of finding out
the real income is also play an important role in determining the
optimal time that the inspector should do his job with the same
taxpayer. The higher the slope of the curve at the beginning time,
the shorter the optimal time would be. In this case, the inspector
would be able to find the real income at the beginning stage at a
high probability. Therefore, the expected government revenue would
be increased rapidly at the beginning stage but with a decreasing
speed, and early reaches its optimal level.
I. Discussions on the framework for developing countries
Before considering the framework for applying the job rotation
system in tax administration in the developing countries, we need
to consider the job rotation scheme in the Japanese public finances
administration system as a unique successful example in the world.
Similarly to almost other Japanese government agencies, job
rotations in the public finance system, including the tax
administration system, have been implemented regularly. The basic
time for rotation is about two years. There are usually two times
for rotating officials and staffs (the director class and under) in
April and July every year. The classes of higher-ranking officials
are also subject to this rotation scheme, although there might be
different in frequencies. The rotation system has also been
designed to ensure the continuousness of works such that there are
always newcomers and experienced people at the offices, relatively.
The jobs/works before and after rotation might be in the same
field, but it is not a rule.
It should be noted that the administrative system in Japan focus
on training public servants in their practice works. After several
times of rotation, they are believed to perceive a vast of
knowledge and experiences concerning the fields of their agencies.
There is also a distinguished feature that the rotation system
usually makes the rotated people moving following upwardly
oscillated orbits, from the bottom to the top of the organizations
management system. Although it seems that there is no relation
between job rotation and promotion, looking at the whole
working-life of a public servant, it would be easy to see the
relation between them, not only in positions/ranks, but also the
salary levels.
For a public official who passed the Type I of National
Examination for public servants and work in government agencies,
so-called type I officers, he usually have the working life as
follows.
Stage 1: Working at some divisions/department in the agency for
around two years, and then, promoted to the section-chief level. In
this stage, he is to work at the office from a position as a staff
from the beginning to gain various experiences, including the
mental aspect, at the lowest level in the administration system.
Therefore, it would be easy for him to understand how the works are
processed when he promotes to manager positions.
Stage 2: Going to abroad for studying in two years, rotated in
the agency, or going to an agencys local branch. In this period, he
is usually sent to study in a master course in public
administration or concerned fields. In case of going to an agencys
local branch, he is supposed to learn comprehensively working
experiences in practice at the local level.
Stage 3: Going back to the same agency or going to other
government agency, and being promoted to the rank of deputy
director. In case of going to other government agency, he would be
back to the initial agency after around two years for learning
out-side experienced and creating connections between agencies. In
this stage, he is rotated every two-year period in different
divisions/ departments in the agency and faced various tasks
concerning the agencys fields.
Stage 4: Promoted to the rank of director and rotated within the
same agency. After around ten years of working as a deputy director
in stage 3, he is promoted to the rank of director. He is also
subject to rotation between different divisions in the agency at a
frequency of around two years.
Stage 5: The period before retirement. As a director, after
several times of rotations, he would be promoted to higher
positions/ranks, or after some years of working, retired to work at
some public/private companies as a member of the board of directors
or as a senior advisor to the board. There is also the possibility
that he would go a local finance office to resume the position as
the general director.
Therefore, the typical public servant in Japan is generally a
generalist, not a specialist. In recent years there have been
increasing trends in job rotations between government agencies and
between public and private sector. The Japanese government intends
to make their administration system more flexible and efficient in
a close relation with the public. However, with the system of
top-down administrative orders, it would be very difficult for
creating a flexible system, at least in the near future. There have
been recently some issues emerging because of fast changes in the
socio-economic conditions in Japan. First, as a culture of working,
the system of job rotation seems to be changed little in responding
to rapid changes in the environment. Under the new context of
socio-economic conditions and the process of globalization, the
abortion or revisions of the rotation system, especially in the
frequency aspect, might be appropriate in some fields/sectors;
meanwhile it is necessary to enhance or introduce in others. It is
very difficult to set up a general rule for rotation, as in present
system; there should be discretions over the rule. The job rotation
system needs to be applied flexibly case-by-case. On the other
hand, the currently seniority wage system based on the seniority of
ages also needs to be changed towards or mixed with the system of
wage based on the working performances/productivities. Therefore,
the topic of job rotation for each field/sector should be carefully
checked and revised with relevant reforms in the wage system. This
is also an important issue for the process of stabilization,
liberalization and development of the Japanese labor market.
For tax administration in the developing countries, it is not
easy to apply the job rotation system such that the system is
immediately well functioned like the Japanese case, although we
know that it might play an important role in reducing tax evasion
and corruption. As we noted above, one of the major disadvantages
of job rotation is the difficulty in application. Based on the
socio-economic conditions of developing countries, there would be
three main difficulties. First, the job rotation system would face
strong oppositions of tax officers and taxpayers and increases in
costs of tax compliances. Regarding to tax officers, the job
rotation would reduce income from corruption and cut off the deep
and preferential relations with the taxpayers. Especially, for the
old generation, they would protest the system even stronger because
their conservative manners. There would also be social and welfare
issues for the tax officers who are rotated into the jobs they do
not like or they are not able to do well. For the taxpayers, they
might also oppose the system because, beside the possibility of
increases in the tax liability, the job rotation also means that
the cost of compliance increases relatively because they face new
tax officers more often and need to spend time and other concerned
costs to meet their requests.
Second, there might be distortions in the management system and
organizational aspects. In this aspect, there would be three main
problems: (i) the allocation issue: the work of the personnel
department in charge of rotating is likely to be biased. The rule
of the right person in the right place would be easily violated,
especially in the socio-economic conditions of the developing
countries where rules seem less affective than discretions; (ii),
the job evaluation system and concerning policies such as the wage
policy and promotion schemes would be significantly affected; and
(iii), the possibility of cooperation between the current officers
and the next officers in charge of the same taxpayers. If they
cooperate, there would be the issue of new bottle, old wine, and
then, the effect of job rotation would be less meaning in the
aspect of protecting tax evasion and corruption.
Third, it should be noted that nowhere in the world has such a
wonderful socio- economic conditions like Japan so that it would be
easy to introduce the system of job rotation. The social aspects
concerning the rotation, such as culture of working, customs,
routines, and especially, the peoples attitudes to the job rotation
system would be rather stiffened. It seems not easy to change these
aspects if there are not relevant measures and breakthrough
explanations.
Although facing these difficulties, job rotation is believably
indispensably necessary in tax administration in the developing
countries because tax evasion and corruption has been becoming more
and more serious in these countries. The introduction of the job
rotation system would be beneficial for the economy as a whole if
there is a relevant and feasibility framework in combination with a
process of preparing necessary conditions for successfully
introducing the system. Regarding to the feasible framework, there
would be following issues that should be seriously considered.
(i) Which fields of works/jobs/regions should be rotated? It
should be noted that job rotation is usually at the cost of
specialization; therefore, at the beginning stage job rotation
might be only suitable in the works/jobs where the level of
corruption is relatively high or the jobs require less levels of
specialization. The decision on the scope and regions for applying
the system is also needed to consider carefully because it would be
very difficult to send some tax officers living in the urban areas
to do some jobs at the remote mountain areas. Rotating people
within the local network would be relevant.
(ii) Who is subject to rotations? This is very important because
it is impossible to rotate everyone in tax administration at the
beginning. It is also very important when regarding to the aspect
of the efficiency of rotations in reducing corruption. At the
initial stage, it would be advisable to rotate only tax officers
who are directly in charge of investigating or determining the
reported income of the taxpayers because they are believably the
most powerful and corruptible.
(iii) How often to rotate? This is an extremely difficult
question at the initial stage of application. The optimal time
would be dependent on the jobs characters, objectives, and
socio-economic environments. However, the time should not be too
short that would cause high costs and make the works of rotated
people less effective. The frequency should be also not too long
such that the possibility of cooperation between taxpayers and tax
officers become relatively high.
It would also be advisable for the government that the rotation
system should be applied in some pilot programs with the approach
of trying, seeing, and revising. One the system is successful in
the level of these pilot programs; it would be easy to multiple
these positive effects to expand the application of the job
rotation to the whole tax administration system as well as other
relevant fields/sectors of the economy.
Regarding to the necessary conditions for applying the job
rotation system, it is necessary to prepare all concerned issues
carefully, especially the systems of job evaluation, wage, and
promotion, the program of training concerned officials, and the
attitude of the government in the process of introducing the
system. First, regarding to the systems of job evaluation, wage,
and promotion, it is necessary to set up new regulations concerning
job rotations, especially schemes of penalties and incentives.
Without a relevant system of job evaluation, the efficiency of the
job rotation system would be not precisely assessed so that
necessary revisions and adjustments would not be implemented. On
the other hand, the issues of wage and promotion are also very
important, which should be considered comprehensively along with
the job evaluation system. Second, training officials to meet the
requirements of job rotations is also very important. Because the
system is rather new for developing countries, the concerned people
in tax administration need to know well about the nature of the
system and study both work-related knowledge and skills and mental
aspects. This also an effective measure to avoid shocks to the
rotated people because they used to do their jobs in very narrow
fields. With the rotation system, they need to understand broader
fields of works, which need a relevant system of training for
rotation.
Finally, a strong stance of the government towards the job
rotation system is very important. As presented above, the system
would be strongly opposed by both the taxpayers and the tax
officers, costly, and rather strange to the public at the beginning
stage. Therefore, it would be not implemented successfully without
a strong leadership of the government. Beside their strong stance,
the government also needs to make the system clearly to the public,
especially to explain its important effect on reducing corruption.
Because in the developing countries, most of the people are alarmed
by the popularity of corruption, one they understand the positive
effect of the rotation system in reducing corruption, they would
support the government to pursue the objective of introducing the
job rotation in public administration in general and tax
administration in particularly.
Besides, another important condition is to enhance internal
control, which is considered as a foundation for enhancing
accountability in the administration system. The framework of
establishing and maintaining effective internal control would be
very important for successfully introducing the system of job
rotation. There is also another aspect should be noted. The job
rotation system would change the ways of working and thinking of
jobs and careers of public servants. It would also affect the
working environment, relation and culture. Therefore, it would take
time to introduce and adjust. The introduction of the system would
be only successful when its advantages and positive effects on the
economy finally recognized by a large number of the people. When
introducing successfully in the public sector, the prospect for the
private sector to apply the job system would be also opened. For
the small- and medium enterprises, it might be not necessary to
introduce the system. However, for the large firms and
organizations, the system of job rotation would be necessary and
important due to its noted advantages, especially positive effects
on career development, productivity, flexibility, and reductions in
working stresses, among others.
II. Conclusion
In the developing world nowadays, preventing and protecting tax
evasion and corruption in the tax administration have been become
one of the most important tasks for both policy-makers and
economists. Based on the stylized facts of these economies, we
analyzed the effect of job rotation in reducing corruption in a
simple model of tax compliance for developing countries with the
assumption that the tax inspectors probability of finding out tax
evasion and the probability of cooperation between the taxpayer and
the inspector are functions of the time that the inspector works
with the same taxpayer. In this framework, the time would have two
distinct effects on tax compliance. On the one hand, the time would
make the expected government revenue increasing because the
possibility of finding out tax evasion increases along with the
time. On the other hand, the probability of cooperation between the
taxpayer and the inspector also increase as the time increases,
which would reduce the expected government revenue. The solution of
this trade-off relation would be the optimal level of the time
where the positive revenue effect is offset by the negative
cooperation effect. This theoretical model clearly illustrates the
superiority of job rotation to reduce tax evasion and corruption in
tax administration in the developing countries. The idea of
considering the job rotation system in the framework of tax evasion
and corruption is the distinguished feature of the theoretical
model developed in this paper because job rotation and tax evasion
corruption have been usually considered as two separate topics in
the recent economic literature. Job rotation is usually considered
in the context of industrial organization at the firm level; the
relation of job rotation and corruption has been rarely discussed.
On the other hand, in the literature on tax compliance, most of
studies concentrated on analyzing aspects of tax burden, penalties,
and incentives.
It should also be noted that the theoretical framework in this
paper might be widely extended in the manner that introductions of
working efforts of tax inspectors, honest and dishonest taxpayers
and inspectors, and changes in assumptions and parameters would
have no significant effects on the results and arguments of the
model the job rotation system would still have a positive effect in
reducing tax evasion and corruption, especially in the condition of
much biased and distorted information and difficulties in socio
economic conditions in the developing countries. In this paper, we
restrained analyses in the case that job rotation would reduce the
inspectors probability of finding out the real income. However,
when this restraint is lifted, it is even more beneficial for the
government to introduce the system in tax administration. In the
field of tax investigations, rob rotation might also enhance the
working proficiency of the inspectors in some cases. Among others,
one of the most possible cases is that there would be heterogeneous
taxpayers with different types of income in an economy with
asymmetric information. The taxpayers structure of income is
changed rapidly such that a rotated inspector with a vast of
experiences concerning investigations of these types of income
would do his job better than a non-rotated inspector who are
specialized in specific taxpayers.
We also discussed the framework for applying the job rotation
system in tax administration in the developing countries. It is
clear that it would be not easy to apply the system in the
developing countries because of difficulties and obstacles in the
socio-economic conditions in these countries, although we know its
positive effect in reducing tax evasion and corruption. The job
rotation system in the Japanese administration is very important
for the developing countries when considering the system to
introduce because it has been quite successful and stable. However,
the socio-economic conditions in the developing countries are much
different from that in Japan. Therefore, it is necessary to make
careful considerations on possible difficulties in introducing the
system, build the feasibly framework, and prepare necessary
conditions for the process of implementation in order to make the
job rotation system feasible and efficient in tax administration in
these countries.
References
Acconcia, A., DAmato,M., and Martina, R., 2003, Tax Evasion and
Corruption in Tax Administration, Mimeo, University of Napoli
FedericoII, Italy.
Aoki, M., 1990, Toward an Economic Model of the Japanese Firm,
Journal of Economic Literature, Vol. 28, 1-27.
Allingham, M. , and Sandmo , A., 1972, Income Tax Evasion: A
Theoretical Analysis, Journal of Public Economics, Vol. 1,
323-38.
Asher, M., 2001, Design of Tax Systems and Corruption,
Conference on Fighting Corruption: Common Challenges and Shared
Experiences organized by the Konard Adenauer Stiftung, and the
Institute of International Affairs (SIIA) Singapore, 10 ~ 11 May
2001Bardhan, P., 1997, Corruption and development: a review of
issues, Journal of Economic Literature, Vol. 35, 13201346.Besley,
T., and McLaren, J., 1993, Taxes and Bribery: the Role of Wage
Incentives, Economic Journal, Vol. 103,119-41.
Buchanan, B., 1974, Building Organizational Commitment: the
Socialization of Managers in Work Organizations, Administrative
Science Quarterly, Vol. 19, 533-46.Campion, M., and Cheraskin,
Lisa, 1994, Career-Related Antecedents and Outcomes of Job
Rotation, Academy of Management Journal, Vol. 37, 6, 1518-43.
Chander, P., and Wilde, L., 1992, Corruption in Tax
Administration, Journal of Public Economics , 49:333349.
Cho, I., and Kreps, D., 1987, Signaling Games and Stable
Equilibria, Quarterly Journal of Economics, Vol. CII, 179221.
Cho, I., and Sobel, J, 1990, Strategic Stability and Uniqueness
in Signaling Games. Journal of Economic Theory, Vol. 50,
381-413.
Cosgel, M., and Miceli, T., 1998, On Job Rotation, University of
Connecticut, Department of Economics Working Papers, 1998-02.
Cosgel, M., and Miceli, T, 1999, "Job Rotation: Costs, Benefits
and Stylized Facts." Journal of Institutional and Theoretical
Economics, Vol. 155, pp. 301-20.Das-Guppta, A., Engelschalk, M.,
Mayville, W., 1999, An Anticorruption Strategy for Revenue
Administration, PREM Notes (Worlbanks Poverty Reductions and
Economic Management Network), No. 33.
Dutton, R., 1987, The Japanese Manager: the Coordinator of the
Network, Business Forum/Spring 1987, 21-24.
Fisman, R., and Svensson, J., 2002, Are Corruption and Taxation
Really Harmful to Growth? Firm Level Evidence, Mimeo, Columbia
Business School.
Gill, J, 2000, A Dianogstic Framework for Revenue
Administration, World Bank Technical Paper 472, the World Bank.
Gittleman, M., Horrigan, M., Joyce,M., 1998, Flexible Workplace
Practices: Evidence from a National Representative Survey,
Industrial Labor Relation Review, Vol. 52, 1, 99-115.
Guthrie, J, and Olian, J., 1991, Does Context Affect Staffing
Decisions? the Case of General Managers, Personnel Psychology, Vol.
44, 263-92.
Hazzard, L., Mautz, J., Wrightsman, D., 1992, Job Rotation Cuts
Cumulative Trauma Cases, Personnel Journal, Vol. 71, 29-32Hellman,
J., Jones, G., Kaufmamn, D., Schankerman, M., 2000, How Firms and
Bureaucrats Shape the Business Environment in Transition Economies,
Policy Research Working Paper 2312, the Work Bank.Hindriks, J.,
Keen, M., and Muthoo, A., 1999, Corruption, Extortion and Evasion,
Journal of Public Economics, Vol. 74, 395-430.
Jovannovic, B, 1979, Job Matching and the Theory of Turnover,
Journal of Political Economy, Vol. 87, 6, 972-990.
Jones, G, 1986, Socialization Tactics, Self-Efficacy, and
Newcomers' Adjustments to Organizations, Academy of Management
Journal, Vol. 29, 262-79.
Kaufmann, D., 1997, Corruption: The Facts, Foreign Policy,
Summer 1997
Kaufman, F., and Lawaree, 1993, Collusion in Hierarchical
Agency, Econometrica, Vol. 61, 629-56.
Kreps, D. M., and R. Wilson. (1982). Sequential equilibria.
Econometrica, 50:863894.
Koike, K., 1984, Skill Formation Systems in the U. S. and Japan:
A Comparative Study in M. Aoki (ed.), The Economic Analysis of the
Japanese Firm, North Holland: Amsterdam, 47-75.
Kindra, G., and Stapenhurst, R., 1998, Social Marketing
Strategies to Fight Corruption, The Economic Development Institute
of the World BankKuroda, A., 2001, The wage-labour system in Japan,
the effect of globalization, ERC/METU International Conference
2001Kusunoki, K., and Numagami, T., 1998. "Interfunctional
Transfers of Engineers in Japan: Empirical Findings and
Implications for Cross Sectional Integration." IEEE Transactions on
Engineering Management, Vol. 45, No. 3, 250-62.
Lehtinen, K., and Lehtinen, P., 2001, Job Rotation European
Models, JobRotation International AssociationsLondon, M.,1983,
Toward a Theory of Career Motivation, Academy of Management Review,
Vol. 8, 620-30.
MacLeod, D, Kennedy, E, 1993, Job Rotation System: Report to XYZ
Co., Dan MacLeod Ergonomics website,
http://www.danmacleod.com/Miller, R., 1984, Job Matching and
Occupational Choice, Journal of Political Economy, Vol. 92, 6,
1086-1120.
Mookherjee, D., and Png, I., 1995, Corruptible Law Enforcers:
How Should They Be Compensated, Economic Journal, Vol. 105,
145-59.
Morrison, R., and Brantner, T., 1992, What Enhances or Inhibits
Learning a New Job? A Basic Career Issue, Journal of Applied
Psychology, Vol.77, 926-40.
Mourdoukoutas, P., and Roy, U., 1994, Job Rotation and Public
Policy: Theory with Applications to Japan and the USA,
International Journal of Manpower, Vol. 15, 6, 57-71.
Ortega, J, 2001, Job Rotation as a Learning Mechanism,
Management Science, Vol. 47, 10, 1361-70.
Ostroff, C., Kozlowski, S., 1992, Organizational Socialization
as a Learning Process: The Role of Information Acquisition,
Personnel Psychology, Vol. 45, 849-74.
Ouchi, W., 1981, Theory Z: How American Business Can Meet the
Japanese Challenge, Reading, Addison-Wesley, MA.
Polinsky, A., and Shavell, S., 2001, Corruption and Optimal Law
Enforcement, Journal of Public Economics, Vol. 81, 1-24.
Rose-Ackerman, S., 1997, High-Level Rent Seeking and Corruption
in African Regimes: Theory and Cases, WB Working Paper WPS 1780,
the World BankSanchez, I., and Sobel, J., 1993, Hierarchical Design
and Enforcement of Income Tax Policies, Journal of Public
Economics, Vol. 50, 345-69.
Shleifer, A., and Wishny, R., 1993, Corruption, Quarterly
Journal of Economics, Vol. 108, 599-617.
Sonnenfeld, J., Peiperl, M., 1988, Staffing Policy as a
Strategic Response: A Typology of Career Systems, Academy of
Management Review, Vol. 13, 588-600.
Triggs, D., and King, P., 2000, Job Rotation - An Administrative
Strategy for Hazard Control, Professional Safety, February
2000.
Tirole, J., 1986, Hierarchies and Bureaucracies: on the Role of
Collusion in Organizations, Journal of Law, Economics and
Organization, Vol. 2, 181-214.
Transparency International, 2003, Global Corruption Report 2003,
the Transparency International
Wane, W., 2002, Tax Evasion, Corruption, and Remuneration of
Heterogeneous Inspectors, WB Working Paper WPS 2394, the World
Bank.World Bank, 2000, Helping Countries Combat Corruption, Poverty
Reduction and Economic Management (PREM) Network, The World Bank.
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
EMBED Equation.3
( The author would like to thank Hosen Mitsuo, Kashiwagi Shigeo,
Mutoh Shizuki, Satoh Masayuki, Kuzumi Masayuki, and other members
participating in a workshop on tax policy and administration in
developing countries at the Policy Research Institute of the
Ministry of Finance, Japan. He is also grateful to Ukai Yukio from
the National Personnel Authority and Shimazaki Masao from the
Ministry of Finance Secretariat for providing useful information on
the job rotation scheme in the Japanese administration system. The
views expressed in this paper are, however, those of the author and
do not necessarily reflect the views of affiliated
institutions.
See Global Corruption Report 2003 (Transparency International
(2003). Also see Bardhan (1997), Rose-Ackerman (1997), Hellman et
al. (2000).
Also Campion and Cheraskin (1994).
For theoretical models of tax evasion and corruption, in
general, see Besley and McLaren (1993), Hindriks et al. (1999),
Polinsky and Shavel (2001), among others.
See Aoki (1990) and Campion and Cheraskin (1994). It is also
necessary to recognize the European style of job rotation, which
focuses on training employees (off-the-job training) and pre-
training for the unemployed to work as substitutes (See Lehtinen
and Lehtinen (2001)).
Also see Buchanan (1974) and Jones (1986).
Gittleman et al. (1998) showed the probability of job rotation
is 22.3 percent for the firms introducing technology (compared with
the average ratio of 9.6 percent) in a sample if over 5,800 US
firms in all industries in 1993.
See Hazzard et al. (1992), Triggs and King (2000), among others.
It should be noted that Cosgel and Miceli (1998) considered an
organization model of analyzing costs and benefits of job rotation
and recognized the importance of job satisfaction and
attributes.
Also see Koike (1984), Aoki (1990), among others. For recent
empirical studies on job rotation in Japan and US, see Kusunoki and
Numagami (1998) and Campion and Cheraskin (1994)
It should also be noted that, as showed in the study of
Mourdoukoutas and Roy (1994), job rotation is higher in Japan than
in the US partly because anticipatory job mobility is higher in
Japan than in the US, due to the fact that the culture and labor
policy in Japan encourages anticipatory job rotation, meanwhile the
US labor policy provides assistances after an employees lost his
job.
Also see Shleifer and Wishny (1993), Wane (2002), Acconcia et
al. (2003)
The results that the expected government revenues might be
reduced when there are increases in the fine rate or the tax rate
can be also seen in Graetz et al. (1986), Chander and Wilde (1992),
among others.
Regarding to the solution and distribution of premium from
engaging in tax corruption, as in most of the literature on
corruption, the bribe the taxpayer pays the inspector is the
solution of a Nash bargain and they mutually agree to report the
income that maximizes their joint profit (Kreps and Wilson (1982),
Cho and Sobel (1990)).
Also see Kaufman (1997), Gill (2000), Asher (2001), Fisman, R.,
and Svensson (2002)
The tax rate might be varied, as a function of income. However,
for simplicity, we assume that the tax rate is flat. The case of
varied tax rates would have no significant impact on the results of
the model. Furthermore, excluding the individual income tax, most
types of taxes, such as the corporate income tax, are usually
flat-rated taxes.
There would be also the case of considering the inspectors
effort; however, we assume that it is given in order to concentrate
on analyzing the times effects. The case concerning the inspectors
effort can be analyzed by introducing a variable of inspectors
effort in this function. However, the results of the model would
not be changed significantly.
We can also examine the case when the limit of time for surely
finding the real income is shorter than that for surely cooperating
( EMBED Equation.3 ). At some certain level of the gap between
these two limits, the results of the model would not be
changed.
We also assume that there is no relation betweens the auditing
agency and the inspector and taxpayer so that they are well
functioned: there would be not further bargaining processes.
Also see Choo and Krep (1987), Hindriks et al. (1999).
From now, we assume that EMBED Equation.3 . The case of EMBED
Equation.3 can be similarly described without significant effects
on the results.
It should be noted that this possibility happens only when p(t)
is relatively concave. In fact, this function is rarely relatively
concave. However, we discuss here to ensure the comprehensiveness
of the model.
For the case of type II and type III officers who passed the
Types II and III of National Examination for public servants, they
are also usually rotated to different jobs in the same or different
offices in the agency, and also subject to promotion up to the rank
of deputy director.
Also see Kuroda (2001).
Also see Das-Guppta et al. (1999), WB (2000), Gill (2000), Asher
(2001).
Also see Kindra and Stapenhurst (1998).
PAGE 5
_1132398114.unknown
_1132399627.unknown
_1132399811.unknown
_1132485390.unknown
_1132485584.unknown
_1132759457.unknown
_1132993166.unknown
_1132485601.unknown
_1132485867.unknown
_1132485480.unknown
_1132400968.unknown
_1132401132.unknown
_1132402759.unknown
_1132483521.unknown
_1132400992.unknown
_1132400351.unknown
_1132400817.unknown
_1132400081.unknown
_1132399715.unknown
_1132399750.unknown
_1132399793.unknown
_1132399727.unknown
_1132399738.unknown
_1132399640.unknown
_1132399646.unknown
_1132399633.unknown
_1132399108.unknown
_1132399517.unknown
_1132399571.unknown
_1132399580.unknown
_1132399538.unknown
_1132399160.unknown
_1132399199.unknown
_1132399377.unknown
_1132399130.unknown
_1132399018.unknown
_1132399070.unknown
_1132399088.unknown
_1132399061.unknown
_1132398991.unknown
_1132399005.unknown
_1132398158.unknown
_1131963538.unknown
_1132233139.unknown
_1132318253.unknown
_1132321939.unknown
_1132322749.unknown
_1132323118.unknown
_1132398075.unknown
_1132323131.unknown
_1132322795.unknown
_1132322402.unknown
_1132322576.unknown
_1132322595.unknown
_1132322543.unknown
_1132322383.unknown
_1132321062.unknown
_1132321100.unknown
_1132319843.unknown
_1132320933.unknown
_1132321000.unknown
_1132319894.unknown
_1132319817.unknown
_1132234787.unknown
_1132240731.unknown
_1132317858.unknown
_1132317938.unknown
_1132318012.unknown
_1132242238.unknown
_1132317694.unknown
_1132242221.unknown
_1132235738.unknown
_1132236061.unknown
_1132237977.unknown
_1132236015.unknown
_1132235456.unknown
_1132233309.unknown
_1132233438.unknown
_1132233287.unknown
_1131969583.unknown
_1132042367.unknown
_1132042766.unknown
_1132158580.unknown
_1132042421.unknown
_1131971055.unknown
_1132041803.unknown
_1132041844.unknown
_1132041900.unknown
_1132041517.unknown
_1131970092.unknown
_1131968904.unknown
_1131968922.unknown
_1131964850.unknown
_1131965158.unknown
_1131968842.unknown
_1131964891.unknown
_1131964064.unknown
_1131898644.unknown
_1131957957.unknown
_1131958388.unknown
_1131958459.unknown
_1131958382.unknown
_1131956694.unknown
_1131957401.unknown
_1131957946.unknown
_1131956990.unknown
_1131900857.unknown
_1131900938.unknown
_1131900880.unknown
_1131900802.unknown
_1131900821.unknown
_1131793660.unknown
_1131880510.unknown
_1131886414.unknown
_1131893727.unknown
_1131895076.unknown
_1131895827.unknown
_1131896856.unknown
_1131894542.unknown
_1131894390.unknown
_1131894361.unknown
_1131887783.unknown
_1131888787.unknown
_1131893643.unknown
_1131889258.unknown
_1131888071.unknown
_113188