credible government policies.pdf

Chapter 23 Credible Government Policies, I23.1.IntroductionKydland and Prescott (1977) opened the modern discussion of time consistencyin macroeconomics with some examples that show how outcomes differ in other- wise identical economies when the assumptions about the timing of government policy choices are altered.1In particular, they compared a timing protocol in which a government chooses its (possibly history-contingent) policies once and for all at the beginning of time with one in which the government chooses sequentially. Because outcomes are worse when the government chooses sequen- tially, Kydland and Prescott’ s examples illustrate the value to a government of having a “commitment technology” that requires it not to choose sequentially. Subsequent work on time consistency focused on how a reputation can substi- tute for a commitment technology when the government chooses sequentially.2 The issue is whether constraints confronting the government and private sec- tor expectations can be arranged so that a government adheres to an expected pattern of behavior because it would worsen its reputation if it did not. A “folk theorem” from game theory states that if there is no discountingof future payoffs, then virtually any first-period payoffcan be sustained as a reputational equilibrium. A main purpose of this chapter is to study how dis- counting might shrink the set of outcomes that are attainable with a reputational mechanism. Modern formulations of reputational models of government policy use and extend ideas from dynamic programming.Each period, a government faceschoices whose consequences include a first-period return and a reputation to1Consider two extensive-form versions of the “battle of the sexes” game described by Kreps(1990), one in which the man chooses first, the other in which the woman chooses first.Backward induction recovers different outcomes in these two different games. Though theyshare the same choice sets and payoffs, these are different games. 2Barro and Gordon (1983a, 1983b) are early contributors to this literature. See KennethRogoff(1989) for a survey.– 937 –938Credible Government Policies, Ipass on to next period. Under rational expectations, any reputation that thegovernment carries into next period must be one that it will want to confirm. We shall study the set of possible values that the government can attain withreputations that it could conceivably want to confirm. This and the following chapter apply an apparatus of Abreu, Pearce, and Stacchetti (1986, 1990) (APS) to reputational equilibria in a class of macroeco- nomic models. APS use ideas from dynamic programming.3Their work exploits the insight that it is more convenient to work with the set of continuation values associated with equilibrium strategies than it is to work directly with the set of equilibrium strategies. We use an economic model like those of Chari, Kehoe, and Prescott (1989) and Stokey (1989, 1991) to exhibit what Chari and Ke- hoe (1990) call sustainable government policies and what Stokey calls credible public policies.The literature on sustainable or credible government policies in macroeconomics adapts ideas from the literature on repeated games so that they can be applied in contexts in which a single agent (a government) behaves strategically, and in which all other agents’behavior can be summarized in terms of a system of expectations about government actions together with competitive equilibrium outcomes that respond to the government’ s choices.43This chapter closely follow Stacchetti (1991), who applies Abreu, Pearce, and Stacchetti (1986, 1990) to a more general class of models than that treated here. Stacchetti also stud- ies a class of setups in which the private sector observes only a noise-ridden signal of the government’s actions. 4For descriptions of theories of credible government policy, see Chari and Kehoe (1990),Stokey (1989, 1991), Rogoff(1989), and Chari, Kehoe, and Prescott (1989). For applications of the framework of Abreu, Pearce, and Stacchetti, see Chang (1998), and Phelan and Stacchetti (1999).Introduction93923.1.1. Diverse sources of history dependenceThe theory of credible government policy uses particular kinds of history depen- dence to render credible a sequence of actions chosen by a sequence of policy makers. Here credible means an action that a government decision maker wants to implement. By way of contrast, in chapter 19, we encountered a distinct source of history dependence in the policy of a Ramsey planner or Stackelberg leader. There history dependence came from the requirement that it is neces- sary properly to account for constraints that dynamic aspects of private sector behavior put on the time t choices of a Ramsey planner or Stackelberg leader that at time 0 makes once-and-for-all choices of intertemporal sequences. In this context, history dependence emerges from the requirement that at time tthe Ramsey planner must confirm private sector expectatio。