This paper studies optimal fiscal and monetary policies in a stochastic economy with imperfectly competitive product markets. Presuming that the government cannot commit to past policy plans, we characterize time-consistent policies in both flexible and sticky price environments.
The economy we consider features dynamic inconsistency, and the properties of optimal policies, therefore, generally depend on whether the government is assumed to have access to a commitment technology that binds its future actions. Assuming commitment, and hence adopting a Ramsey approach, both the long-run and the cyclical properties of optimal macroeconomic policies have been studied extensively in the literature and are by now well-understood. In this paper, we seek to learn about the properties of optimal policies absent a commitment technology. Under this alternative assumption, existing studies so far have focused on the long-run properties of optimal policies within the context of relatively simple economies. In particular, attention has largely been restricted to deterministic settings. Consequently, our knowledge about the cyclical properties of optimal time-consistent policies in the face of macroeconomic shocks is limited. By analyzing optimal policies without commitment in a stochastic economy, our paper tries to address this gap in the literature.
The framework we employ for our analysis is the stochastic production economy developed by Schmitt-Grohé and Uribe (2004a,b). In this economy, technology and government spending shocks are key in driving macroeconomic fluctuations. We devote attention to both environments of flexible and sticky prices, and we allow for varying degrees of imperfect competition. The government’s task in the model is the classical public finance one of using monetary and fiscal instruments in order to finance an exogenously given stream of public expenditures in the least distortionary way. This is a non-trivial dynamic problem, as the government does not have access to lump-sum instruments: Fiscal income taxes reduce the incentive to supply labor, whereas positive nominal interest rates generate distortions due to their role as opportunity costs of holding money. Moreover, by allowing to shift these distortions over time, government debt can potentially play an important role in absorbing macroeconomic shocks. Under discretion, the policy problem is best described as a dynamic game between successive governments. Assuming that reputational mechanisms are not at work, we analyze Markov-perfect equilibria of this game. Hence, policy choices are restricted to depend on history only via a set of layoff-relevant state variables. Being an indicator for the tightness of the government’s intertemporal budget constraint, the real level of public liabilities is such a state variable.
Our focus on a particular model economy is motivated as follows. First, and most importantly, the economy studied by Schmitt-Grohé and Uribe (2004a,b) constitutes a relatively simple and tractable framework. At the same time, however, it embraces several interesting features that are deemed important for the analysis of business cycles. Specifically, the model incorporates monopolistic competition, and therefore allows for positive profits in the economy; this provides scope for inflation as an indirect tax on monopoly rents. Furthermore, the economy features nominal rigidities which are introduced via quadratic price adjustment costs; this modeling approach implies resource costs associated with inflation, but avoids relative price distortions. Finally, monetary non-neutrality is generated via a transaction costs motive for holding money; this turns out to be attractive from a computational perspective, because, unlike cash-in-advance restrictions, it does not give rise to occasionally binding constraints in a stochastic economy. Apart from these considerations, the second motive behind our choice of model economy is to facilitate the comparison of our results with the Ramsey framework. In this respect, the existence of a readily available and well-established benchmark allows us to focus our attention on time-consistent policies when assessing the qualitative and quantitative implications of alternative assumptions concerning the government’s intertemporal commitment capacity.
A first key insight of our analysis is that, for the model under consideration, the steady state corresponding to a Markov-perfect equilibrium is (at least locally) unique and asymptotically stable. This is not true for the model’s Ramsey equilibrium, which features a continuum of neutrally stable steady states and where the initial conditions determine which of these steady states is approached in the long-run. Intuitively, the steady state that emerges as the Markov-perfect equilibrium outcome has the property that the commitment problem faced by the policy maker disappears. In particular, we find that the Markovian government accumulates a large net asset position in order to finance its outlays via the associated interest earnings. In this respect, our paper relates to the results in Aiyagari et al. (2002). However, the mechanism stressed here is not the precautionary savings motive of a Ramsey planner under incomplete markets, but the Markovian planner’s intertemporal commitment problem which provides incentives to manipulate the economy’s endogenous state variables.
Download
Optimal fiscal and monetary policy without commitment
