European Journal of Law and Economics, 15: 251–261, 2003

c 2003 Kluwer Academic Publishers. Manufactured in The Netherlands.

The Role of Interest Groups in the European Union

Decision-Making Process
ANTONIO GARCÍA LORENZO aglec@udc.es
Department of Economic Analysis and Business Administration, Faculty of Economics, University of A Coruña

Abstract

Most of the economic models that analyse the behaviour of interest groups in the policy making process uphold
the idea that there are many organized groups that compete, with the aim of achieving their individual goals. The
adoption of decisions is the result of a complex system of strategic interactions, and since different groups have
different resources at their disposal this makes it easier for one or more groups to influence whether or not a policy
is adopted. This research demonstrates that an institutional system such as the European Union (EU) should be
able to protect us from the potential manipulation, which accompanies these channels of influence.

Keywords: European Union, decision-making, interest groups

JEL Classification: D72

Over recent years a great deal of attention has become focussed on information theory, both
on a theoretical level and as a tool to analyse specific problems. Agency theory has enabled
us to understand certain phenomena, which would have been impossible to understand using
models in which all the participants possessed the same information.
The situations in which asymmetric information constitutes an integral feature of a con-
tract, i.e., a situation in which one party knows something that the other does not, are
frequent. Agency models have allowed researchers to explain a number of important situ-
ations of this type within the fields of public choice—Williamson (1990), McMaster and
Sawkins (1996) and Mashaw (1997), in law—Spiller (1990) and Privileggi, Marchese, and
Cassone (2001), and in economics—Bamberg and Spremann (1987), Molho (1997) and
Wright, Mukherji, and Kroll (2001).
Given that governmental structures within the system of political organization constitute
a chain of agency relationships, they may be legitimately studied within this perspective.
The works of Dixit (1998), Spiller and Tommasi (2000) and Majone (2001), form the basis
of an important program of research whose objective is to explain the evolution of political
relationships in terms of an agency relationship, in markets which are characterized by
imperfect information.
Most of the economic models that analyse the behaviour of interest groups with respect
to the political decision-making process put forward the idea that there are many organized
groups all competing to achieve their individual objectives. As a result policy decisions are
taken as a consequence of the interaction among these groups. We should ask ourselves
however, whether the channels via which society takes part in the political decision-making
252 LORENZO

process are equally available to all groups. Does the imbalance in the distribution of re-
sources allow some of the more powerful groups to condition the outcome of the policy
making process and thus strengthen the somewhat incestuous nature of the process?
By the very nature of the European Union (EU), certain interests may be looked upon
favourably to the detriment of transparency. Given that our main concern should be to
protect the political system against certain forms of influence, we should ask ourselves if
this goal remains viable given the institutional changes that have come about as a result of
the integration process. To this end we set up a model, which allows us to formalize the
participation of interest groups in the EU decision-making process, and this constitutes the
main objective of this research.
The basic role of the European Commission (EC) within the context of this process is to
collate information and anticipate the reactions of member states to its proposals. In carrying
out its objectives the EC has always been the most permeable of the European institutions
with respect to the opinions of organized groups. We will attempt to study the possibility
that the EC promotes the formalization of a contract designed to provide an incentive for a
specific interest group to act as a producer of information, as means of assuring the viability
of its proposals. Such a contract is formalized as an agency relationship, given that, the
imperfect information that the EC possesses with respect to the group’s intervention, is
reflected in the design of the contract. Logically, from a political point of view, there are
many potential pit-falls in the creation and consolidation of such a system, the crux of
which may be stated thus; could a system of checks and controls for such a relationship be
established that is rigorous enough to protect the decision-making process from the threat
of the abuse of this form of symbiosis?
This paper is set out in the following way: Section 1 contains a brief description of the
EU decision-making process. Section 2 formally presents a principal-agent model whose
aim is to facilitate a payment system that provides the incentive necessary to induce the
interest group to realise the level of effort desired by the EC. In Section 3, the theoretical
model is applied in order to explain a particular case study, and finally in Section 4, we
offer our main conclusions.

1. The political decision-making process in the EU

Over recent years attention has become focussed on the relevance for EU citizens of the
political system that underlies the EU itself; although democracy offers citizens diverse
channels via which to make their demands, it would seem that, in general, they are far from
content with the efficacy of the system. The so called Euro-barometer, has allowed us to
confirm this lack of confidence and reveals that there are growing doubts as to the legitimacy
of a process which involves making and taking excessively opaque decisions.
We find ourselves faced with a political system, which is practically impossible to classify.
The reason for this is that although the EU has the authority necessary to adopt legislative
decisions, which under the guise of policies are imperative for society, there are no political
parties capable of articulating their interests on a European scale. The reader interested in
a more exhaustive analysis of this subject should read George (1996), Stone and Sandholtz
(1997), Armstrong and Bulmer (1998), Peterson and Bomberg (1999) and Peterson (2000).
EUROPEAN UNION DECISION-MAKING PROCESS 253

The peculiarities of this governmental structure have ramifications for decision-making

procedures. The key word here is “negotiation,” which is usually a complex process in
which the member states tend to fall back upon ideologies rooted in national sovereignty,
and this tends to create networks of political influence, which range from the anarchic to the
highly structured. EU political networks, as suggested by Peterson (1995), are based on the
interests of a number of actors on the political stage: politicians, bureaucrats, interest groups
and the lobby or pressure groups who, on a European level, constitute a highly organised
strata of the European political bed-rock, since they have the information and resources at
their disposal that are needed to attain a desirable consensus.
The EC, as a means of advancing its policy initiatives, promotes informal contacts with a
wide range of interest groups. The EC’s objective is two-fold; to find and collate information
in order to improve its proposals and, to overcome the resistance of national administrations.
National governments, on the other hand, which have at their disposal information that is
relevant to the potential outcome of a policy decision of which the EC is ignorant, will
attempt to condition the EC’s decision by restricting information, thus making the EC
dependant on the data provided by pre-designated bureaucrats and specialists acting as
experts or assessors. An important part of pertinent literature has focused on an analysis of
the agency relationship that exists between those moving the wheels of political power and
the bureaucracy. The real underlying problem inherent in this relationship is the fact that
certain aspects of the level of effort are difficult, or impossible to measure with exactitude.
For a more exhaustive analysis of this subject see, Steunenberg and Van-Vught (1997), Cox
and McCubbins (1999), Vaubel (1999) and Moravcsik (2001).
The rational decision-making process, based on the existence of sufficient information
is short-circuited, precisely because of a lack of information. The modeling of the agency
relationship provides us with a means by which may analyze the informational asymmetries
that exist between the participants.
Of course the EC would be able to solve these problems by carrying out its own search for
information. One means of reducing the costs of information searches, when these appear
to be excessively high, might be to use agents who are specialized in the production of
information. These agents would benefit from having greater accessibility, and their spe-
cialization would make them more productive. In short, agents specialized in the production
of information would effectively reduce the transaction costs normally incurred when other
means are used.
This work focuses on an analysis of the incentives, which arise from the optimum design
of contracts with respect to the making and taking of political decisions within the sphere
of the European Union. We approach the science of contract formulation, specifically with
within the context of interest groups in a novel way and, without recourse to more traditional
analysis.

2. The model

It might be said that transition costs act as a way for the EC to justify the participation
of interest groups in the decision-making process but it should be underlined that the
information produced is imperfect. We should ask ourselves, therefore whether the interest
254 LORENZO

groups, who enjoy important advantages when it comes to acquiring information, should
receive certain incentives from the EC. These incentives would be regulated contractually
and would allow the interest groups to become more profoundly integrated into the decision-
making process acting in the role of information producers.
It would be of advantage to the interest groups to participate in policy making since
they would be able to augment their credibility by transmitting all the information at their
disposal whilst at the same time advancing documented opinions or preferences. The fact
that this possibility becomes very remote once it arrives at the Council of Ministers stage
serves to reinforce the benefits of such a system.
It is assumed that each interest group needs to exert an effort (e), which may have a value
of between 0 and 1. In other words if the problem is to be evaluated correctly and, as a
consequence, the information obtained reliable and veracious, the effort that the information
producing group would need to exert, would be 1. Were the group to exert an effort of a
value of below 1, the result would be an inexact evaluation of the problem. In short:

(e = 1): Maximum effort, implying the correct evaluation of the problem.

(e < 1): Indicating that the effort is below maximum, thus leading to an inexact evaluation
of the problem.

Further, it is supposed that even though the effort made by the interest group is not
observable by the EC, and as such cannot be included in the terms of the contract offered
to the group, the result of the effort1 can be reflected by means of a noisy indicator (β).2
This indicator can take a value of 0 or 1 and is affected by a source of noise if and when a
maximum effort achieves the correct evaluation of a problem and therefore veracious and
reliable information. This does not imply that the estimator must be exactly equal to one for
it to act deterministically, but rather that its value must approach 1 with a certain probability
according to the effort made.3
We may therefore define the probabilities associated with the indicator which are condi-
tioned to the effort carried out by the interest group, in the following way:

p[β = 1 | e < 1] = q
p[β = 0 | e = 1] = r
p[β = 0 | e < 1] = 1 − q
p[β = 0 | e = 1] = 1 − r

If it is assumed that the probability of the indicator being correct is greater than the proba-
bility of it being incorrect, then the following expression must hold true:

1>r >q >0

Formally, we suppose that the preferences of the information-producing group may be

expressed in the form of a utility function given by:

V (w, e) = U (w) − eK (1)

EUROPEAN UNION DECISION-MAKING PROCESS 255

Where (w) represents the payment that the group receives for its collaboration with the
EC,4 (K ) is a positive real valued scalar, and U (·) is twice continuously differentiable and
bounded, by U  (·) > 0 and U  (·) < 0. In the same way the inverse function may be written
as:

ψ(·) ≡ U −1 (·)

In accordance with the outline we have given, the general set up of the game we propose
to deal with is as follows: First, the EC decides on the contract that they are going to offer
the interest group. The interest group then accepts or rejects the contract depending on its
terms. In order to understand the nature of the difficulties with which the EC is faced prior
to designing the contract, it should be pointed out that, when the conditions for obtaining
information and the effort required to obtain it are asymmetric, it is possible for the EC to
propose the amount of effort it would wish the group to exert, but it is not possible to reflect
this in the terms of the contract. As a consequence, the EC must take into consideration the
fact that once the contract has been signed, the interest group can choose the level of effort,
which is most advantageous from their point of view given the nature of the contract. In
short, the EC may propose a certain level of effort, but should ensure that this corresponds
to the level that the interest group really wants to exert.
Given that the effort made by the interest group benefits the EC, it is highly likely that
the EC will offer a contract that induces it to exert the maximum effort (e = 1). In doing
so it achieves two goals; on the one hand it allows the EC to ensure that the information
used in the elaboration of policy is reliable and thus helps to prevent the proposals from
being discredited at the Council of Ministers, and on the other, it clears up any doubts that
might remain with respect to the credibility of the information-producing interest groups.
The contract therefore may be expressed in the following way:

(β) = W H , if β = 1
(β) = W , L
if β = 0

Where (W H ) and (W L ) represent the payments made by the EC to the interest group.5
Formally, the expression for the expected utility of the interest group as a function of
effort may be written as follows:

EU (e = 1) = r w H + (1 − r )w L − K (2)
EU (e < 1) = qw H + (1 − q)w L − eK (3)

Where U (W H ) = w H and U (W L ) = w L .
In order to calculate the optimum contract,6 the EC must find a solution to the following
programme:

Min r ψ(w H ) + (1 − r )ψ(w L ) (4)

[w H ,w L ]
256 LORENZO

subject to (5) and (6).

r w H + (1 − r )w L − K ≥ V (5)
(r − q)(w H − w L ) ≥ K (1 − e) ∀ e ∈ [0, 1] (6)

Where (5) ensures that an individually rational (IR) group accepts the contract. Inherent in
this equation is the fact that the interest group could refuse the contract if the expected utility
that the group obtains from the relationship is not at least equal to that of the reserve. It
follows therefore that EU (e = 1) > V . And Eq. (6) ensure incentive compatibility (IC) in
order to motivate the group to select (e = 1). This equation reflects the fact that the interest
group is willing to choose the option that the EC proposes (e = 1) if that option maximizes
its objective function, which is EU (e = 1) ≥ EU (e < 1). Since the (IC) constraint will be
most binding for (e = 0), we can rewrite it as: (r − q)(w H − w L ) ≥ K .
With Lagrange multipliers µ and λ appended to (5) and (6) respectively, the first-order
optimality conditions may be expressed in the following way:

r −q
r ψ  (w̄ H ) − r µ − (r − q)λ = 0 ⇔ ψ  (w̄ H ) = µ + λ (7)
r
r −q
(1 − r )ψ  (w̄ L ) − (1 − r )µ + (r − q)λ = 0 ⇔ ψ  (w̄ L ) = µ − λ (8)
1−r

Where (w̄ H ) and (w̄ L ) represent the utilities that the interest group hopes to obtain, in the
form of payments received in accordance with the optimum contract, if the indicator reveals
(β = 1) and (β = 0) respectively.
Given that (1 > r > q > 0) and (K > 0), Eq. (6) implies that (w̄ H ) > (w̄ L ). Further,
if we take into account that ψ(·) is increasing and convex, we have ψ  (w̄ H ) > ψ  (w̄ L ).
Finally, Eqs. (7) and (8) imply specifically that (λ > 0) and (µ > 0). Consequently, both
(5) and (6) are binding equalities.
Solving (5) and (6) as simultaneous equations gives us:

1−q
w̄ H = V + K (9)
r −q
q
w̄ L = V − K (10)
r −q

The solution to the programme resolved by the EC in order to design the optimum contract
may be defined in the following way:
 
1−q
W̄ = ψ V + K
H
(11)
r −q
 
q
W̄ L = ψ V − K (12)
r −q

Where (W̄ H ) and (W̄ L ) represent the optimum values for the payments made by the EC to
the interest group.
EUROPEAN UNION DECISION-MAKING PROCESS 257

From this result, the following conclusions may be drawn:

(1) The interest group’s expected utility, considering the payments made by the EC and
Eq. (1), is equal to the reserve utility: V (w, e) = V .
(2) As the source of noise that affects the estimator disappears, so the payment received by
the interest group approaches the payment that covers the reserve utility, and this also
compensates for the unused available effort. In the extreme case in which the total effort
exerted by the information producing group may be estimated exactly, the optimisation
of risk distribution, will lead the EC, which is neutral, to assume the whole of the risk,
and thus it completely ensures that the group that receives a payment:

W̄ H = ψ[V + K ]

(3) Finally, we might say that when the effort exerted by the information-producing group
is perfectly observable, the expected cost for the EC will be less than when there is noise
in the verification of the level of this effort. This is because under such circumstances
the payment for uncertainty disappears:
   
1−q q
ψ[V + K ] < r ψ V + K + (1 − r )ψ V − K
r −q r −q

This conclusion is easily obtained by means of the Jensen inequality.

3. The application of the model to a case study

In spite of the difficulties inherent in applying agency theory to specific problems, the
model, which has been presented above, is capable of explaining a real case, one which
contemplates the participation of an interest group in the EU’s decision making process.
This case study is the objective of this section.
The emotive case of meat destined for human consumption bearing excessive quantities
of certain types of hormone, first came to light in Europe in the 1980’s. The press managed to
provoke widespread public concern by publishing disturbing images of pupils from various
different schools with abnormally large chests. The EC, following normal procedure in
such circumstances and with the objective of compiling and collating information in order
to improve their proposals, set up a committee, in this case the Lamming committee. The
role of this body was to investigate the use of both synthetic and natural hormones in the
feeding of European cattle.
The hormone producing companies, which form a part of the European Federation of
Animal Health, reported to the Lamming committee and a number of other bodies including;
The Committee of Animal Nutrition, The Scientific Committee of Alimentation, and The
Veterinary Committee. They claimed that there are no scientific reports, or indeed evidence
to support that adding hormones to animal feed has negative effects on the individuals who
consume the meat. The Lamming Committee concluded that the natural hormones and
anabolic steroids used to fatten the animals were harmless.
258 LORENZO

In the case study we wish to carry out, it is not possible to observe the effort exerted by
the information-producing group directly, which means that our “principal-agent” model
constitutes a suitable means by which to focus on, and analyze the problem. The model
allows us to incorporate the fact that it is impossible to observe with exactitude the amount
of effort exerted by the interest group. The actual result of this exertion however, can be
known via the use of an indicator. Any contract containing some kind of control mecha-
nism, which facilitates information about the behavior of the group would be beneficial to
the EC.
The analysis is based on the following three variables: the decision, the amount of effort
exerted, and the payments offered. The second of these, the amount of effort exerted,
represents an indicator of the human and material resources utilized by the interest group in
the elaboration of the reports, which are subsequently given to the EC. Whilst it should be
remembered that the measurement of the effort exerted by the interest group constitutes an
inherently problematical component of information analysis, we assume that in this case it
is clearly defined, that is, the effort itself may be measured with precision. The payment the
group receives is related to its reputation, which in turn is taken into account by the group
when it determines the value of its expected utility.
Finally, The EC promotes the formalization of the contract in an attempt to assure its
acceptance by; this part of the process is known as “the decision.” It is assumed that there is a
relationship between the effort exerted by the group and the extent to which the proposal that
is elaborated by the EC is acceptable to the Council of Ministers. This is a realistic situation
given that the effort exerted by the information-producing group, whilst being somewhat
difficult to control, benefits the EC in that it allows them to improve their proposals.
Within this context therefore, the EC (the principal), wishes to design a contract that
provides an incentive to the interest group (agent) to exercise a given level of effort, the
effort necessary to obtain reliable information on the subject of the proposal, a proposal,
which is intended to provide an efficient solution to a specific problem. It should also be
noted that the given level of effort, a level, which is decided by the EC, must satisfy the
additional condition that it has, by necessity, to be accepted by the information-producing
group. The level of effort therefore, has to be such that it maximizes the objective function
of the agent.
In formal terms, this implies that the problems that the EC has with respect to the
maximization process, which will dictate the optimum contract, is subject to two constraints:
the constraint of participation and the constraint of compatibility of incentives. The first
condition requires that the interest group (FEDESA) accepts the invitation to participate
in the elaboration of the proposal. To this end it is, of course, necessary to guarantee a
reserve utility, that is, the option to not participate, a choice that is assumed to be constant
and known. The incentives condition, on the other hand, must take into consideration that
the interest group acts once the contract establishing the conditions under which they must
work has been defined. This means that the contract must be designed in such a way that
the group chooses the level of effort desired by the EC.
Finally, in the last stage of the game, the “decision” is taken, the results are observed by
the EC and the corresponding payments are made. Given that a credibility factor is taken
into consideration when the value of the interest group’s expected utility is determined, the
EUROPEAN UNION DECISION-MAKING PROCESS 259

renewal, or repetition of the contract will induce the interest group to adopt the behavior
desired by the EC.
In the case study, the final stage of the game took place in 1984, when the EC made
a proposal, based on the findings of the Lamming report, which was subsequently sent
to the European Parliament where a resolution was then taken. Following the European
Parliament’s recommendations, the Council of Ministers adopted directive 85/649/CEE,
which bans all types of hormones as additives in the composition of animal feed. In the end
the Council of Ministers approved a proposal, which was quite distinct from the one drawn
up by the Lamming committee. In our opinion this was due to the fact that, although the
report itself was based on strictly scientific arguments, it failed to take into consideration
the fears that lay at the heart of this particular food crisis, namely; the extent to which
consumers were protected or, “unprotected.” The control system that we have designed,
which has been based on the need for consensus among a series of institutions, would have
worked correctly and prevented the party possessing the informational advantage from
exploiting this advantage.
This particular case continues to be the source of much debate, given that FEDESA’s
industries declare significant annual losses, a fact that will be exacerbated by the threat of
a trade war with the U.S., particularly since the directive affects imported mean in exactly
the same way.

4. Conclusions

The paper herein expounded sets out a new way of analysing the way in which interest
groups participate in the policy making process. It establishes the conditions via which
interest groups, acting as information-producers, are provided with incentives to participate
in the policy making process.
The model we have designed contemplates a relationship between interest groups and
the EC, which is governed by contracts that induce honesty and transparency. It focuses
therefore, on an analysis of the incentives that arise as the result of the design of these
contracts. This paper foresees the need to impose a system of control in order to generate
the correct level of incentives, given that the level of effort exerted is not directly observable.
The efficacy of this control or limitation system will depend upon the amount of information
made available concerning the group acting as the information producer. Any information
about the level of effort exerted by the interest group will be valuable as this will allow the
minimization of the of the risk inherent in the relationship and thus reduce the expected
cost for the EC. In the extreme case in which the source of noise, which affects the indicator
disappears and the effort exerted can be verified with precision, the party in possession of
an advantage in terms of information would not be able to benefit from said advantage.
This system of control therefore constitutes an effective means of protecting the pro-
cess from the risk which is latent in the EC-interest group dynamic, and this is possible
thanks to the way the EU is structured, a structure based on the cooperation of a series
of institutions that share diverse responsibilities and various sources of legitimacy. What
is certainly true is that, in an institutional system in which the powers of the institutions
themselves are interwoven and disperse, an increase in consensus should mean that there
260 LORENZO

is less manipulation taking place with respect to all the groups taking part in the policy
making process.
The model, which we have designed here may be developed further when the control
system itself is modified. The control system would be affected as the number of countries
in the EU increases and the need for reform grows. The construction of the political union
would also create the need for a profound reform of those factors affecting the institutional
equilibrium of the decision-making process and as a consequence a modification of the
control system. The efficacy of the control system that we have designed will ultimately
depend upon the way in which the institutional equilibrium changes over time, specifically
with respect to two scenarios. The first of these envisages the EC as a dynamic technocracy
with the Council of Ministers ratifying the EC’s decisions whilst the second foresees the
EC as a federal government with a bicameral parliament in which the Council fulfils the
role of the upper chamber, representing the interests of the member States.
Finally, a natural extension of the model would involve an analysis of reputation models
in which longer-lasting relationships were studied by means of introducing repeated games.

Notes

1. The utility from introducing the result into the contract, is derived from the information it provides concerning
the effort exerted by the group. This constitutes its only value given that we consider that the EC is neutral with
respect to the risk, and the interest group is adverse to it.
2. It will be in the EC’s interest to include within the contract any mechanism which might provide information
about the behaviour of the interest group taking part.
3. The extent to which the Council of Ministers’ proposals are found to be acceptable, may be used as an indicator
of the effort which the interest group exerts, on evaluating the problems that arise within the framework of the
contractual relationship with the EC. Thus if (β = 1), then the proposal has been accepted and this reveals the
correct evaluation of the problem. Conversely if (β = 0), then the proposal is rejected which means that the
evaluation of the problem was inexact. Further, the indicator is sometimes affected by a source of noise because
of the lack of political motivation or simple inertia affecting those governments that make up the Council and
this is what sinks the EC’s proposals.
4. The payment that the interest group receives for acquiring and processing the information might be defined
in terms of the esteem in which the EC holds the interest group. Thus, attaining and maintaining a healthy
reputation should represent a natural incentive for the said group, and will condition the extent to which the
group participates in the future.
5. As previously stated, the EC will use the result as a source of information about the behaviour of the interest
of the group. As a consequence payment will depend upon the information produced, and this payment will
increase in accordance with the result, which in turn will mean that the interest group will tend to increase
the effort exerted. Finally we use the letter (H ) to indicate a high level of effort and the letter (L ) in order to
indicate low effort.
6. We will study the case when the EC is neutral with respect to the risk and the interest group is averse to it.

References

Armstrong, K. & Bulmer, S. (1998). The Governance of The Single European Market, Manchester: Manchester
University Press.
Bamberg, G. & Spremann, K. (eds.) (1987). Agency Theory, Information and Incentives, New York: Springer.
Cox, G. W. & McCubbins, M. D. (1999). “The Institutional Determinants of Economic Policy Outcomes.” Annual
Conference 1999 of the International Society for New Institutional Economics. Washington D.C.
EUROPEAN UNION DECISION-MAKING PROCESS 261

Dixit, A. K. (1998). “Transaction Cost Politics and Economic Policy: A Framework and a Case Study.” In M.
Baldassarri (ed.), Institutions and Economic Organization in the Advanced Economies. New York: St. Martin’s
Press.
George, S. (1996). Politics and Policy in the European Union, Oxford: Oxford University Press.
Majone, G. (2001). “Nonmajoritarian Institutions and the Limits of Democratic Governance: A Political
Transaction-Cost Approach.” Journal of Institutional and Theoretical Economics. 157(1), 57–78.
Mashaw, J. L. (1997). Greed, Chaos and Governance: Using Public Choice to Improve Public Law, New Haven
and London: Yale University Press.
McMaster, R. & Sawkins, J. W. (1996). “The Contract State, Trust Distortion and Efficiency.” Review of Social
Economy. 54(2), 145–167.
Molho, I. (1997). The Economics of Information: Lying and Cheating in Markets and Organizations, Oxford:
Blackwell.
Moravcsik, A. (2001). “Despotism in Brussels? Misreading the European Union.” Foreign Affairs. 80(3), 114–122.
Peterson, J. (1995). “Decision-Making in the European Union: Towards a Framework for Analysis.” Journal of
European Public Policy. 2, 69–93.
Peterson, J. (2000). Power, Decision and Policy in the European Union, Oxford: Oxford University Press.
Peterson, J. & Bomberg, E. (1999). Decision-Making in the European Union, London: Macmillan.
Privileggi, F., Marchese, C., & Cassone, A. (2001). “Agent’s Liability Versus Principal’s Liability When Attitudes
Toward Risk Differ.” International Review of Law and Economics. 21(2), 181–195.
Spiller, P. T. (1990). “Politicians, Interest Groups and Regulators: A Multiple-Principals Agency Theory of
Regulation, or Let Them Be Bribed.” Journal of Law and Economics. 33(1), 65–101.
Spiller, P. T. & Tommasi, M. (2000). “The Institutional Fundations of Public Policy: A Transactions Approach
with Application to Argentina.” Annual Conference 2000 of the International Society for New Institutional
Economics. Tübinben.
Steunenberg, B. & Van-Vught, F. (eds.) (1997). Political Institutions and Public Policy: Perspectives on European
Decision Making, Boston and London: Kluwer Academic.
Stone, A. & Sandholtz, W. (1997). “European Integration and Supranational Governance.” Journal of European
Public Policy. 4(3), 297–317.
Vaubel, R. (1999). “Enforcing Competition Among Governments: Theory and Application to the European Union.”
Constitutional Political Economy. 10(4), 327–338.
Williamson, O. E. (1990). “A Comparison of Alternative Approaches to Economic Organization.” Journal of
Institutional and Theoretical Economics. 146(1), 61–71.
Wright, P., Mukherji, A., & Kroll, M. J. (2001). “A Reexamination of Agency Theory Assumptions: Extensions
and Extrapolations.” Journal of Socio Economics. 30(5), 413–429.