An Example of Conflicts Rather Than Passive Subordination and the Need for a Theory of Conflicts
In order to illustrate the existence of a conflict that can be represented in terms of both REs and strategic games, think of the Barro-Gordon model. At the time of Barro and Gordon (1983), the emphasis of the policy debate was still far from the search for conditions of existence of an instrument vector that could guarantee satisfaction of some fixed targets (Tinbergen’s fixed-target approach) or of an optimal policy that could minimise a given loss function (Theil’s flexibletarget approach).
In fact, the Lucas critique was usually deemed to negate the possibility that the policymaker could control the system at all. The discussion therefore concentrated instead on issues of effectiveness or neutrality of specific instruments when the private sector has some specific target(s) and instrument(s), thus continuing, in a new setting, the debate that had started in the preceding two decades.Barro and Gordon in fact studied a situation of conflict where both players, the central bank and the private sector are active. This can be represented simply by introducing the assumption of REs by the private sector in a model where the government wants to minimise a loss function depending on employment and price deviations from some target values. Alternatively, the model can be stated as a Stackelberg game between the private sector (acting as the leader) and the public sector. It is also interesting to note that in this case, one player, the private sector, whose loss depends on deviations of employment from its target value, has one instrument (price expectations). It thus has one target and one instrument. By contrast, the government’s loss depends on two targets, employment and price deviations from some target values (the one for employment is higher than that of the private sector), but the government has only one instrument by which it can control inflation only.
By solving either the game or the system with REs, one can show that the private sector fully crowds out monetary effects on real output. A superior solution, for the public sector, would be to commit to a certain rule. However, having induced favourable private-sector expectations, the policymaker would always be tempted to cheat and renege on his or her commitment, the classic time-inconsistency argument (Kydland and Prescott 1977), in an attempt to achieve yet better outcomes. Being aware of this possibility, in selfdefence, the private sector would anticipate worse results; these can be avoided only if the policymaker’s temptation to cheat is balanced by a fear that he or she might lose his or her reputation and no longer be able to act effectively if this game of interactions with the private sector is repeated.
Thus, with Barro and Gordon, we have a result of policy neutrality. The conflict is solved favourably to the private sector. However, their result is specific to the assumptions of their model, as we see below. They certainly do not have a general theory of policy neutrality, even if much of the literature has acritically accepted it as such. A part of later studies has tried to elaborate such a theory in different ways, without making any reference to Tinbergen’s contribution.[52]
What we need, however, is a theory of conflicts within which this result can be generalised. Such a theory should first show when and to what extent a conflict arises and what the terms of the conflict are. In this respect, a conflict arises every time the players’ targets differ, at least in terms of target values. No conflict can arise where two agents aim to obtain the same desired target values for all their targets, as there is a convergence of interests between the players. Enlarging the focus of our analysis to multiple-player (non-cooperative) games, only some target values of the shared targets might coincide for some players, whereas each of them also pursues other targets, different from those of other players.
The possibility of a partial convergence of interests arises in this case. In fact, ‘implicit’ coalitions can derive from situations where some players share the same target values of some common target variables. We would thus move from a situation of conflict among all players to a situation where conflict is absent among some players, but there is a residual opposition or rivalry among certain groups of them. Conflicts can refer to the relationship between a public and a private sector or among different public bodies, national (e.g. fiscal and monetary authorities) or international (think of decisions of different governments on monetary or trade issues).In addition, a theory of conflicts should state when and how these can be solved, defining the features of the equilibrium solution (e.g. if there is or is not policy neutrality), if any, and how they depend on some critical assumptions of the game, in particular, as to numbers of targets and instruments of each player as well as his or her propensity to negotiations and agreements with other players. More generally, it should show how other possible outcomes of the game, such as no equilibrium or multiple equilibria, might depend on those assumptions - or others.
4.4