2
1. Introduction
Recent research on the causes of war has focused on problems of private information and
incentives to misrepresent to explain what prevents leaders from reaching ex ante bargains
that would avoid the costs and risks of war. This literature concludes that, if both parties in a
dyad have complete information about the other’s capabilities and resolve, war is never
rational, as there is always a bargain that at least one of the parties prefers to fighting (Fearon
1995, Powell 1999). Actors therefore settle on a bargain that achieves the same division of
benefits as the war would have, without incurring its costs and risks: ‘‘with complete
information, bargaining never breaks down in war’’
2
(Powell, 1999).
This conclusion relies, however, on an assumption that is largely unstated in most existing
theories and empirical studies of bargaining, namely that agreements can be updated at any
time without any cost. The costs of reaching and enforcing an agreement are usually
understood as negligible, such that the addition of a variable to a model of crisis bargaining
would not be justified. Authors tend to assume a Coasian world in which adaptation would be
frictionless. Yet, agreements are not and often cannot be re-negotiated continuously and
uninterruptedly at every possible time t
i
. Rather, agreements are ‘‘sticky’’ and extend for
discrete periods of time. Negotiations and the necessary concessions they imply are difficult,
costly and take time. As a consequence, agreements tend to exhibit some degree of rigidity,
such that they frequently remain in force even though the initial conditions that led to their
adoption have dramatically changed. Thus, most of the debates in France and Holland over
the recent referendum on the treaty instituting a constitution for Europe clearly revolved
around the anticipated difficulty to update the agreement in the future.
Discrepancies between
reality and the content of the agreement are easily conducive to increased tensions.
In fact, the mere anticipation of such discrepancies can increase the difficulty to reach
agreements. Thus, an important and largely unaddressed puzzle is why certain deals are ever
2
“[...] If both states are satisfied, neither can credibly threaten to use force to revise the status quo and the status
quo goes unchanged. If one of the states is dissatisfied, the satisfied state offers the dissatisfied state control over
an amount equivalent in value to its payoff to fighting. The dissatisfied state accepts this offer, and the status quo
is peacefully revised in its favor”, in Powell 1999.
Convention