II SIX DECADES OF MULTILATERAL TRADE
COOPERATION: WHAT HAVE WE LEARNT?
C THE DESIGN OF INTERNATIONAL TRADE AGREEMENTS
WORLD TRADE REPORT 2007
156
Enforcement is a function of “enforcement capacity” and “enforceability”. Enforcement capacity is the ability
to reciprocate credibly against a violation of the terms of the contract. Enforcement can be exercised by the
affected party itself (self-enforcement), by a neutral third-party,
111
by society at large or through collective
enforcement by a circle of affected or interested parties (such as the membership of a multilateral contract).
Enforcement instruments can vary from physical (incarceration), economic (penalty fees) to emotional
(reputation loss) measures. Enforceability has three components: observability, verifiability and quantifiability.
Observability means that infringements can be detected in the first place – either by the affected party itself
or by a third party. A contract violation is verifiable if the affected party can point to a clause in the contract
and prove its violation. This presupposes that such a clause is contained in the contract and/or that the
violation can be determined by a neutral third party. Finally, quantifiability, implies that the aggrieved party
(or a court) can quantify the damage incurred as a result of the breach of the contract.
112
(i) The level of cooperation and the level of enforcement
Every contract is driven by the desire to cooperate. Parties enter into contractual relationships with the aim
of minimizing costs, engaging in risk transfer and/or reaping transaction efficiencies. Cooperation may
not be a binary issue, but a matter of degrees. In Chart 3 various degrees of cooperation (C) are plotted
on the horizontal axis, where (C
max
) refers to full cooperation and (C
N
) to the absence of cooperation.
The Chart shows a contract between two players,
113
assumed to cover a long-term relationship with
repeated interaction (such as a trade agreement). Each contracting party has an incentive to cheat by
deviating from the terms of the agreement. The short-term benefit from defection is called hit-and-run
advantage H&R.
114
The vertical axis depicts the utility gain in excess of a situation without a contract,
i.e in excess of U
N
.
115
The hit-and-run benefit (if seized by the potential defector) is by definition an
opportunistic, that is, inefficient redistribution of welfare to the detriment of the affected party.
Chart 3 displays two mechanisms to enforce continued cooperation.
116
In case of self-enforcement (i.e. no
superior enforcement body exists), the affected party exits the agreement (the threat of reverting to the
pre-contractual non-cooperative “Nash equilibrium” is also called a “grim trigger” strategy of enforcement).
The curve
S-E (“self-enforcement”) in Chart 3 represents the injurer’s opportunity costs of reprisal, that is,
the discounted value of cooperation.
117
It is the sum of future benefits from cooperation that the injurer
foregoes from having defected and prompted the grim trigger response. The potential defector balances
the short-term incentive to cheat with the long-term cost from an infinite suspension of cooperation. As the
level of cooperation increases, the costs of reprisal exceed the hit-and-run gains from one-time defection
until the two curves intersect at C
SE
, which can be defined as the “most cooperative” cooperation level that
can be sustained through self-enforcement. Beyond this point, the gains from one-time infringement of the
agreement exceed all the compiled future gains of cooperation. Beyond C
S-E
it is irrational for the injuring party
111
Third-party enforcement may be called “court-and-copper” enforcement, since constitutional states require that a
jurisdiction (judge) determines a legal infringement and an executive (police) enforces the law.
112
Quantifiability may be important where there is no prima facie violation of the rules, or in seeking retaliation.
113
For simplicity, parties are assumed to be symmetrical. Therefore, only one contractor needs to be examined, since the incentives
and actions by the other are identical. This is without loss of generalization: In a model with multiple actors, the enforcement
can always be represented as a two-player game, namely between a player “X” and a player “rest of the world”.
114
The hit-and-run gain is the additional utility the injurer enjoys from defecting over cooperating as promised. It is a
short-term benefit, since it merely stretches from the moment of defection until the time the violation is detected.
115
The convex curvature of line H&R is intuitive (increasing marginal return from defection), but not necessary. Bagwell and Staiger
(2002: 102) provide some arguments in favour of this curvature for the case of the “contract” being a trade agreement. The
curve is flat and equal to zero at C
N
, where the contracted cooperation is equivalent to the situation without agreement.
116
Other kinds of enforcement are possible, but are not considered here, since they are just variants of the two explained
mechanisms.
117
The discounted value of cooperation is the sum of all future per-period extra-gains from having a contract (vis-à-
vis having no contract at all). Again, the concave curvature of the S-E curve is intuitive (more cooperation in every
round is beneficial up to some optimal point C
opt
). After that point more cooperation has declining, possibly negative
returns, for example because of a loss of freedom and sovereignty. See Bagwell and Staiger (2002: 102) for the case
of a trade agreement. The
S-E curve also must go through the origin: The more the contracted cooperation level
approaches the no-contract (Nash-cooperation) level, the smaller are the future gains from cooperation.