agreement over trade policy using the Nash Bargaining Solution.
a two-player international game without any domestic politics, where P is the executive
negotiating in the home country (call her the president or prime minister) and F is the foreign
country. No veto players exist in either country except the executives. This game assumes
complete and perfectly symmetric information. The ideal points of P and F (p and f) are
common knowledge, as is the position of the status quo (q). The payoffs to the players decline
linearly as the distance of the outcome from each player’s ideal points rises. Each executive is a
unitary, rational actor. We assume P and F have no domestic political considerations that are not
factored into their ideal points; these ideal points maximize their political support.
horizontal axis depicts all the possible values of the status quo (q). The vertical axis represents a
continuum of policy outcomes on a single issue. In the arena of trade policy, this could be the
difference between the home country’s and the foreign country’s trade barriers. Negotiating a
trade agreement means choosing a pair of trade barriers for each country that is different than the
status quo and usually one in which the level of barriers for the members is lower than the status
quo.
The figures capture both the ideal points of the actors and the outcomes of the
negotiations (a). The dark line represents the policy choices chosen given the value of the status
quo, as shown along the horizontal axis. There is never a failure to agree in the complete
information model used here; it only occurs with incomplete information. The figures are drawn
2
See Milner and Rosendorff (1996, 1997) and Milner (1997) for a derivation of the results and discussion of the use
of the Nash Bargaining Solution (NBS). We use the NBS because it imposes minimal structure on the bargaining
game, which best replicates the international environment in contrast to the domestic one. In addition, when offers
are made suitably fast, other bargaining games, such as Rubenstein’s, converge to the NBS.
3
We assume that the preferences of interest groups are contained in the ideal points of the political actors. These
actors use the preferences of the interest groups that are important to them to construct their ideal points.
8