The model

We focused on a public organization or government agency that depends politically and economically on the central government authority. Inside this office there are two kinds of civil service officers. The first ones are the dishonest individuals (labelled δ), who benefit from and support the illegal structure inside the government agency office. On the other hand, there are honest people (labelled σ), who receive a normal wage for working in the government agency. They are homogeneous within their own type.

The government agency is receiving a transfer from the central government in order to be distributed between the civil service officers. This transfer is totally independent of the wage received by the officers (this is fixed and set by the central government as a part of the total budget); it consists on an economic compensation, which is going to depend on the budget policy implemented by the government agency authority.

The transfer should be distributed taking into account the productivity of the employees or, at least, as an equal compensation for all of them. Both are budget strategies that could be implemented by the government agency authority. However, dishonest people may affect the budget distribution by offering a bribe to the agency authority. In other words, dishonest people lobby the agency authority in order to obtain an extra amount of income through a bribe. Honest people do not lobby the government and they are expecting a fair policy.

On the other hand, the government agency authority presumes the existence of corruption inside the office, and the distribution policy is going to depend on the level of institutional clearness. The amount of transfer to the honest and dishonest people would depend on the efficiency of the government agency institutional framework. The institutional framework is understood as the legal environment set by the government agency in order to properly regulate the political and economic activities. In this sense, an efficient institutional framework strengthens the control against illegal activities. An inefficient institutional framework means a weak control over illegal activities⁶.

The institutional framework is set by the government agency through a credible political reform of the legal system. For simplicity, this reform is the result of a legal and political process with no direct ex ante economic cost attached. This process will produce a parameter, which measures the efficiency of the institutional framework, and consequently the distribution policy. This parameter will be set between 0 and 1 and will be called institutional level. When the institutional level α, the policy instrument of the government, is close to one the institutional framework is more efficient, when α is closer to zero the institutional framework is inefficient.

Taking into account the considerations mentioned above, some utility functions are defined for the honest people, dishonest people and the government: these functions will be used to determine the optimal institutional level. Assuming quasi-linear preferences, the utility of the honest people can be defined as⁷

In (1), the first term is the legal wage paid by the central government to honest people. This wage is constant and it does not depend on the policy implemented by the government agency. The second term is the economic compensation made by the government agency multiplied by the institutional level. This economic compensation, which will be part of the income given to honest people, is transferred in a lump-sum fashion from the government agency.

The income conferred to dishonest people is given by

In (2), the first term is the legal wage paid by the central government to dishonest people. Once again, this wage is fixed and does not depend on the policy implemented by the government agency⁸. The second term is the economic compensation made by the government agency multiplied by the opposite of the institutional level.

Clearly it is assumed that the wage is the same for both types of employees (w _d=w _h=w). An equal distribution policy will be given when the institutional level is 0.5. Under this policy, and from (1) and (2), it is clear that the income for honest and dishonest people will be the same. Another policy rewarding the employees honest behavior would imply the institutional level to be larger than 0.5. In this case the income received by honest people will be larger than the income received by dishonest people.

At the end, the optimal policy is the institutional level. According to the New Institutional Economics (NIE), an institution is reflects the set of rules of the game that the society agrees to play in order to reduce transaction costs. These rules reduce the cost of negotiation between the agents involved in an organization (^{Klein, 2000}). Of course, these rules reflect a behavioral component from each agent in the organization, and they are the result of the interactions of these agents with a variety of interests (^{Rokeach, 1979}).

Under the argument mentioned above, the institutional parameter is more than a policy instrument, is the expression of an interaction between all the agents involved in an organization. The institutional level reflects the power of corrupted people, the weakness of honest people and the particular interest of the government agency authority. The institutional parameter is the organizational policy implemented by the government agency authority and it is determined by an organizational political equilibrium. It has been closely followed by ^{Dixit, Grossman & Helpman (1997)} and Grossman & Helpman (1994) so as to specify this equilibrium. Honest people do not lobby the government agency, but dishonest people make political contributions to influence the government agency’s decisions. The political contribution scheme settled for the dishonest is denoted byC( α ). So the government agency’s objective function is given by

whereρ> 1 is a constant parameter called corruption level⁹. Reorganizing the equation, it is deduced thatG= (ρ- 1)C+I ^δ+I ^σ. Hence, the government attaches a positive weight to contributions provided thatρ> 1. Equation (3) states that the government agency considers the total welfare of its employees (the terms in parentheses), as well as the total amount of political contribution that it receives (the first term on the right-hand side of Eq. 3).

The organizational political equilibrium is the outcome of a two-stage game. In stage one, dishonest people choose their contribution schedule. The government agency then sets its institutional policy behavior in the second stage. An organizational political equilibrium is given by (i) a contribution functionC(α), in order to maximize the welfare of all the dishonest people given the political behavior anticipated by the government, and (ii) a policy behavior variable,α, that maximizes the government agency’s objective function given by (3), taking the contribution scheme as given.

According to ^{Dixit et al. (1997)}, there is a refinement equilibrium known as truthful organizational mechanism. This equilibrium implements Pareto-efficient outcomes to be taken into account in this analysis. However, as said before, the framework is focused on the specific aspect of the organizational political equilibrium; in such a case it will also follow closely the original ^{Grossman & Helpman (1994)} approach in modeling this equilibrium. Stated formally, let (C° (α°,I ^δ° ), α°) be a truthful equilibrium in which I^δ° is the per-capita utility level of dishonest people at equilibrium. Then (C°(α°,I ^δ°), α°,I ^δ°) is characterized by

where ϕ is defined by

and

Equations (4) and (7) state the truthful contribution scheme. It is set to the level of compensating variation relative to the equilibrium utility level of the dishonest. The definition of ϕ is the basic concept of the compensating variations. Under a truthful equilibrium organizational function, for any change in α, the change in the contribution received by the government agency will exactly equal the change in the dishonest income, provided that the payment, both before and after the change, is strictly positive. Equation (5) is self-explanatory: the government takes the utility level of the dishonest as a given value and chooses the institutional level so as to maximize its objective function. Equation (6) (together with [8]) complete the characterization of the truthful equilibrium and tie down the equilibrium utility level of dishonest people, which is derived from the premise that the dishonest would pay the lowest possible contribution to induce the government to pursue the organizational equilibrium policy given in (5). For this to be the case, the government agency must be indifferent between (i) implementing the organizational equilibrium policy and receiving contributions from the dishonest, and (ii) implementing a policy by accepting no contribution at all. Equation (5) states precisely that¹⁰.

According to ^{Grossman & Helpman (1994)}, in the case of one lobby group there is no opposition from competing interests, and the lobby group captures all of the surplus from its political relationship with the government. In this specific case, at organizational political equilibrium, the government agency derives exactly the same utility as it would have achieved by allowing no contribution. An interesting example with one lobby group can be found in ^{Rama & Tabellini (1998)}. Now, the backbone of the analysis is established.

Some analytical comparative static

It is quite straightforward to see, from (1) and (2), that an increase in the institutional level will benefit honest people and will harm dishonest people in the same proportion. From (1) and (2) it is deduced that

a linear relationship exists between the transfers. Of course, when money is going towards one type of employees, it is leaving the other one. This flow of transfers is made by the government agency authority.

An interesting parameter to be considered is the corruption parameter, which is the level of sensibility of the government agency authority with respect to the contribution made by the dishonest people. In other words, when the corruption parameter is larger, the weight attached to the political contribution is larger too. From (3) it results

An increase in the corruption parameter will increase the government agency objective function in the same proportion than the amount of bribe or political contribution.

On the other hand, when ρ = 1 there is no political relationship between the government and the dishonest people. The weight that the government attaches to social welfare is normalized to one and the government agency objective function is defined as:

In here, there is not government agency authority corruption and, independently of the bribe or contribution, the goverment will apply the fairest policy addressed to benefit honest and dishonest people. It is not an institutional policy because there is not corruption. The corruption of dishonest people does not have any impact on the government agency preference.

Optimal institutional level

Having described the properties of the organizational political equilibrium, in this section we will be analyze the optimal institutional level and its effect on welfare. The first step to determine the optimal α is to obtain the first order condition for the optimization problem given in (5) and (7). The following result is implicitly obtained from (1) to (10):

This is unequivocally negative. It means that the optimal policy will be to set the softest institutional level (α = 0), since when the government agency tries to set a strict institutional level, the loss given by the reduction in the contribution policy plus the reduction in the income of dishonest people is larger than the income received by honest people.

None of the institutional policies seem to be the best organizational policy option and corruption becomes relevant for the government agency. Probably the institutional policy addressed by the government agency does not intend to reduce the flow of income to the honest people, but the consequence of income deviation is grounded by a previous corruption structure. Independently of the government intention, corruption becomes a meaningful variable in the implementation of an organizational policy.

Central government enforcement policy and institutional level

Corruption is not only a kind of social cancer but also an immoral economic practice. It is a complex phenomenon with deep psychological and cultural roots. Corruption is neither an occasional matter, nor exclusive to some countries or political systems. Corruption is supported by social crises affecting all kind of governmental and private organizations.

In this sense, most of the institutional environments inside organizations include, implicitly or explicitly, a moral position against illegal activities. These illegal activities are strictly prohibited because either there is a social pressure faced by one (or all) of the members of the government, or there is recognition that corruption may affect considerably the objective of the government organizations. Independently of the origin and nature of corruption, this phenomenon implies a set of penalties or anti-corruption measures to take place against the corrupted government entity.

Independently of the anti-corruption measure faced by the corrupted government agency, it is assumed that the central government will set an anti-corruption strategy due to social and political commitments.

Even when the government agency benefits from political contributions given by a corrupted civil service body, there are social commitments to hold on, and the central government is forced to set an anti-corruption policy or enforcement policy. This plan is addressed to fight against the government agency corruption trough a credible strategy implemented by the central government.

It will be assumed that the transfer of income in the budget granted from the central government to the government agency is going to depend on the commitment of the government agency authority to increase the institutional level, in order to reduce the impact of corruption. This transfer is the economic compensation given to honest and dishonest people. This distribution may change with the enforcement policy. The transfer from the central government to the government agency is a function of the institutional level, so

and

In other words, when the government agency increases the institutional level, the amount of transfer from the central government to the government agency increases. With the highest institutional level, the government agency is receiving the largest possible transfer; with the lowest institutional level, the government agency is receiving no transfer at all. The central government is rewarding the honest behavior of the government agency authority. However, this is the enforcement policy set by the central government. What would be the optimal organizational policy set by the government agency authority?

There is a conflict faced by the government agency authority. On the one hand, a strict institutional policy will reduce the income of dishonest people and, consequently, the amount of the contribution. On the other hand, a strict institutional policy will increase the amount of transfer and the benefit of honest people. The government agency will transfer the economic compensation from dishonest to honest people.

However, there is a moral dilemma: a strict institutional policy may produce an unfair result, as the economic compensation of honest people may be larger than the economic compensation of dishonest people. The strictest institutional policy (α = 1) may be seen as a punishment to the dishonest. If we intend to set a fair compensation we need to restrict the institutional policy. In other words, the government agency authority may not be interested in changing dishonest people into honest people; the authority may be committing to have a fair compensation policy.

First of all, it is important to determine the optimal organizational policy including now the enforcement policy set by the central government in (14) and (15). For this to be the case, the optimal will be obtained by the first order condition for the optimization problem given in (5) and (7). From (1) to (10), (14) and (15), the following result is implicitly obtained:

Gα=-ρTα+ρT´α-αT´αρ-1+T(α)

Solving implicitly, the result is

It is possible to see that (16) is an ambiguous expression and its value depends on the functionT(α). Proposing an expression holding the conditions in (14) and (15) it can be assessed that

such thatk=T. In this case, the transfer made by the central government will be linearly increasing in α. Taking (17) in (16) the result is:

From (18) the optimal institutional level will be unequivocally positive (α > 0), independently of the level of corruption in the government agency authority. It means that the bribe or contribution offered by dishonest civil service officers in order to be benefited from the economic compensation is negligible. The bribe offered to the government is negligible because the amount of transfer is reduced with a lower institutional level and the bribe is small, independently of the sensibility to the contribution. The government agency authority is not given an important weight to the bribe and it makes a more fairly distribution in the economic compensation. With a strict institutional policy, the benefit obtained by honest civil service officers is larger than the reduction in the benefit of the dishonest, taking account of the smallness of the bribe received by the government agency authority.

However, the question mentioned before is: how large should the institutional level be? The strictest organizational policy (α = 1) means that dishonest people will have a punishment for being corrupted and they are going to receive only their current wage but no compensation at all. On the other hand, honest people will have a large economic compensation not only because the distribution policy of the government agency authority favors them with full assignation, but also because the amount of the transfer made by the central government is the largest possible.

It seems this organizational policy could be an incentive to change the behavior of dishonest people. The rewarding behavior of the government agency authority in favor of the honest may change the attitude of the dishonest in order to turn into honest individuals. The rewarding behavior of the authority is a powerful conductive resource through which the authority of the organization may conduct some changes in the operational rules, eliminating corruption in the organization. However, it is an unfair organizational policy since dishonest people do not receive any economic compensation and disturb the role of the government agency authority as equalizer, which is the main role of the authority according to the model set before.

Alternatively, the optimal organizational policy may be set at the level in which there is not discrimination at all about the type of civil service officers (α = 0.5). Both are receiving the same amount of economic compensation independently of their behavior. In this case, the government agency authority is not discriminating between corrupted and honest civil service officers. This is an efficient behavior because the role of the government agency authority is to make an efficient compensation distribution, leaving the presumption that it plays a moral role about the civil service officers’ behavior. Basically, it is a fair policy because it assumes the role of equalizer and moral considerations (inducing a possible change in the behavior of dishonest people) are not part of the government agency’s duties. However, this organizational policy allows the existence of some corruption in the organization.