عدم خنثی بودن سیاست پولی در بازی های سیاست

This paper investigates the sources of non-neutrality in policy games involving one or more trade unions. We use a simple open economy model to demonstrate basic mechanisms that also arise in other frameworks. There are common roots in the non-neutrality results obtained in such apparently different contexts as, for example, an inflation-averse union playing against the government; a union sharing some other common objective with a policymaker; and when more than one union interacts with monopolistic competitors in the goods market and a policymaker.

The interaction between monetary policy and wage setting was analysed in the 1970s and 1980s in terms of policy games with a particular focus on questions of time consistency, central bank independence and the like. A related aspect of such interaction, that of the non-neutrality of money (i.e., the possibility for the monetary authorities to control the rate of output growth), was first investigated by Gylfason and Lindbeck (1994). They make use of a rather simple game between government and organised labour and show that ‘monetary expansion stimulates output and employment despite the optimal reaction of the unions as long as they care about inflation’ (Gylfason and Lindbeck, 1994, p. 43). The property noted by Gylfason and Lindbeck has been widely used in the literature to derive a number of unconventional results. For instance, Jensen (1997) shows how the Rogoff's result of counter-productiveness of international co-ordination is not robust when trade unions are introduced as players. However, Jensen's result no longer holds if the assumption of an inflation-averse union is removed. Cukierman and Lippi (1999) derive a Calmfors and Driffill hump-shaped relationship between the degree of centralisation and employment. Their result also collapses into a monotonic relationship if the assumption of an inflation-averse union is removed. Moreover, their result is not robust if an information setting where players simultaneously interact (Nash equilibrium) is considered rather than a game where the unions are able to pre-commit to their wage policies (Stackelberg equilibrium). The reason the inflation-aversion assumption does not provide the same results in the Nash case is not completely understood (see Ciccarone and Marchetti, 2001).1 The fruitfulness of the results obtained in policy games between the central bank and one or several unions, together with the criticism of the assumption of an inflation-averse union,2 has motivated several studies where non-neutrality comes out not from the union inflation-aversion but from the interaction between goods and labour markets Soskice and Iversen, 1998, Soskice and Iversen, 2000, Coricelli et al., 2000, Coricelli et al., 2001 and Lippi, 2001. However, even in these cases, the non-neutrality result is not robust with respect to the elimination of the assumption of either a multiplicity of unions acting in the labour markets or monopolistic competitors in the goods markets. While the literature on policy games and unionised economies has moved several steps beyond the pioneering models of the 1970s, as we have indicated, not all the results are completely understood. In particular, although many studies have based their results on some sort of non-neutrality proposition, only a few have risen to the challenge of investigating its roots.3 We shall investigate the sources of non-neutrality in policy games involving trade unions in a simple model in order to clearly expose the basic mechanisms, which also apply to more complex frameworks. We then show that there are common roots in the non-neutrality results obtained in apparently different contexts. Finally, we show that there are other cases where these results can arise. Section 2 clarifies the definition of neutrality and the propositions that have been advanced to state the conditions for non-neutrality. Section 3 presents a simple general model that is used in Section 4 to discuss the mechanisms that drive non-neutrality. Section 5 states necessary and sufficient conditions for non-neutrality to arise. Section 6 summarises our findings and draws general conclusions.

6. Concluding remarks By using a rather general but simple model, we have investigated the sources of non-neutrality in policy games involving one or more trade unions. We have highlighted the basic mechanisms at work and found necessary and sufficient conditions for non-neutrality to hold in the particular class of models analysed. In the manner of Tinbergen, neutrality is finally determined by a particular specification of the policy game in terms of relations between instruments and targets. In our closed-economy specification, if the union does not care about inflation, both the union and the policymaker have two apparently independent arguments in their preference functions and one instrument. However, the union's arguments are not truly independent and, therefore, neutrality necessarily arises since the union's two objectives can be reduced to one. By contrast, when the objectives of each player are truly independent, which happens when the union—in addition to caring about the real wage and output—also dislikes inflation, the players are forced to share their payoffs. Neutrality can then arise only as a special case. Moreover, the possible existence of a real wage-wedge plays an important role, since it is a way to introduce inflation indirectly into the union's preferences and so to break down the one-to-one correspondence between the real wage relevant for the union and output. Specifically, we have shown that a necessary, albeit not sufficient, condition for non-neutrality to arise is that the union take account of prices in its preference function, either directly or indirectly. Two further qualifications are required in order to have sufficient conditions: either the marginal rate of substitution between output and prices in the union's ‘inflation-augmented preference function’ depends on prices or the union should be able to pre-commit to its wage policy. This perspective allows us to understand the common roots in the non-neutrality results obtained in such apparently different contexts as: an inflation-averse union playing against the government, a union sharing a common objective with a policymaker, and several unions interacting with a policymaker and with monopolistic competitors in the goods market.