سیاست های پولی برای غفلت اقتصاد به نحوی عقلانی با تنظیم قیمت مبهم
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|27944||2014||25 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Economic Dynamics and Control, Volume 38, January 2014, Pages 184–208
The paper examines the optimal monetary policy when firms are constrained by information processing capability and infrequent price adjustments. Firms' information processing limit gives rise to imperfect knowledge about macroeconomic aggregates and endogenous information learning contingent on the monetary policy. Staggered price setting introduces the observed price duration and additional policy tradeoffs resulting from the interactions between nominal rigidities and imperfect information processing. The integrated model implies an optimal policy that commits to complete price stabilization in response to natural rate shocks but not in response to markup shocks. In the presence of markup shocks, it is optimal for the central bank to focus on price stabilization in the initial periods following markup shocks and shifts the emphasis to output gap stabilization later. Moreover, larger information capacity, stronger complementarities and more persistent shocks require more aggressive price stabilization in the short-run.
This paper studies the optimal monetary policy in a Taylor staggered-pricing economy in which firms pay limited attention to aggregate variables. What policy should a central bank follow? It is a classic question in macroeconomics. The answers reflect economists' belief about how the economy works, what should be valued, what monetary policy is feasible, and hence what explicit assumptions and models should be used. To account for the real effect of monetary policy, the literature mainly focuses on two types for models for monetary policy analysis: imperfect information models and models with short-run nominal frictions. The widely used New Keynesian models, such as Goodfriend and King (1997), and Clarida et al. (1999), assert that firms' price adjustment is infrequent and time-dependent. Despite their ability to generate monetary neutrality and appealing microfoundations, it is found that such models encounter difficulties in generating plausible effects of monetary policy, such as counterfactual costless deflations (Ball, 1994) and failure to generate a delayed and persistent effect on inflation (Mankiw and Reis, 2002). The basic New Keynesian model thus has been modified by incorporating a backward-looking component into the Phillips curve (e.g. Gali and Gertler, 1999). However, the modification is ad hoc, and it has been criticized for not able to explain the lack of inertia in early U.S. monetary regime when the economy operated under a gold standard and for ignoring the Lucas critique similar to the conventional Phillips Curve (Ball et al., 2005). Another part of the literature focuses on the information structure in monetary models. Slowly disseminating or incomplete information can produce the observed sluggish inflation response to a monetary shock and account for monetary nonneutrality. For example, Lucas (1972) assumes that producers solve a signal extraction problem to infer aggregate prices. However, this idea has been criticized since macroeconomic data is widely available, and such information should be considered by a rational person. Mankiw and Reis (2002)'s sticky information model assumes frictions in updating information rather than updating prices. It successfully generates the inflation-output tradeoff and monetary nonneutrality, but suffers similar problems as Lucas (1972). The main mechanism for generating inertial inflation behavior is firms' price setting decision based on outdated information, but it is puzzling why some firms would use very old information when most macroeconomic variables are publicly available with little delay. Rational inattention models, as in Sims (2003), may be able to solve this dilemma. The model claims that agents' inability to process, not obtain, the data perfectly prevents them from obtaining full information. Economic agents have a limited capacity to process information “attention”, and they rationally choose to be less attentive to some available information and fail to incorporate full information in their decision-making. An information processing constraint is an appealing way to model information frictions for monetary models for several reasons. The method, discussed in Sims (2003) and initiated in information theory, is able to approximate information processing procedure while stepping away from the psychological details. Furthermore, optimality with limited information processing resources leads to information endogeneity. Agents' information learning and expectation are policy contingent, making the model robust to Lucas critique. This paper makes progress towards determining the optimal monetary policy by integrating staggered pricing models with an information friction resulting from limited information processing capability. Firms have multi-period overlapping price contracts as in Taylor (1979), and they optimize their information processing to obtain the most accurate information given their constraints. Both information processing limits and infrequent nominal adjustment are empirically relevant features.1 The integrated model is capable of generating a delayed price response to shocks and an amplified output response to monetary disturbances. There are two types of shocks that cause aggregate fluctuations in the model: natural rate shocks, which are driven by technology innovations and change the efficient level of output, and markup shocks, which cause fluctuations without affecting the efficient level of output. The aggregate fluctuations lead to two inefficiencies. First, aggregate output may deviate from the efficient output as the shocks disturb firms' pricing and production process (real effect of shocks). Second, price dispersion arises, which leads to inefficient allocation of the production resources. In the model, the monetary authority solves a Ramsey problem and chooses a money supply rule to minimize these two inefficiencies. The effectiveness of monetary policy depends on how the nominal and information frictions affect firms' pricing decisions. Monetary policy affect output gap variation through three channels. First, fundamental shocks directly lead to changes in output gap, the magnitude of which can be dampened by policy response. With incomplete information and price adjustment, a change in the money supply leads to a partial adjustment in output gap. Second, staggered pricing introduces an intertemporal link in firms' prices, allowing for any policy response to have a persistent effect on the output gap. Third, variations in economic fundamentals raise uncertainty. Tracking errors arise as firms' information processing is imperfect. A policy that facilitates learning improves the quality of information received by firms and leads to more efficient pricing and production decisions. Furthermore, inefficient price dispersion can also be traced to the two frictions in the model. Staggered price contracts lead to price difference across firms depending on their time of adjustment. Information heteorogeneity also contributes to price dispersion as firms base their pricing decision on their own information set. A monetary policy that emphasizes price stabilization and facilitates information processing thus can minimize the inefficiency caused price dispersion. I find that an optimal policy implied by this integrated model should fully accommodate natural rate shocks and stabilize prices. However, when markup shocks are present, there are conflicting policy effects. Offsetting the markup shocks leads to price stabilization, which facilitates learning, reduces relative price distortion, and eliminates persistent shock effects. Intuitively, price stabilization allows for better knowledge of the aggregate price overall and more efficient pricing. It also helps firms that lack the ability to update their prices when a markup shock occurs, since the optimal reset price remains unchanged. Therefore, a price stabilization objective serves the tasks of (1) minimizing inefficient price dispersion and (2) minimizing the inefficient output gap variation due to tracking errors and intertemporal transmission of shock effects through staggered prices. However, markup shocks lead to inefficient fluctuations in output, which requires the central bank to accommodate the shock and let prices fluctuate. Therefore, a careful balance between the effects is necessary, and complete price stabilization is no longer optimal. The tradeoff between the objectives is determined by firms' information processing capacity, strategic complementarities in pricing, and shock persistence. I find that it is optimal for the monetary authority to focus more on offsetting markup shocks and stabilizing prices in the short-run, and shifts to accommodating the shock and stabilizing output in the long-run. Larger information capacity, stronger complementarities, and higher shock persistence all lead to more aggressive price stabilization in the short-run. Larger attention also leads to a more rapid shift of policy emphasis to stabilize output in later periods. The paper is mainly related to three strands of literature. It is related to the rational inattention literature, such as Sims (2003), Mackowiak and Wiederholt (2009), and Woodford (2009). This paper models information structure following the line of this literature, but none of the above studies consider monetary policy problems. My paper also relates to the literature that studies the monetary policy using time-contingent pricing models, such as Fukunaga (2007) and Wolman (2001). Fukunaga (2007) considers a similar Taylor economy, but it studies the effect of an exogenous monetary policy when agents have higher order beliefs. Wolman (2001) studies the optimal monetary policy a Taylor staggered price setting, but does not consider the role of imperfect information. More closely, my paper is related to the literature that considers the policy implications of imperfect information, such as Ball et al. (2005), Adam (2007), and Paciello and Wiederholt (2013). Ball et al. (2005) study optimal monetary policy in a sticky-information setting and predicts price level targeting to delivery the best outcome. Adam (2007) and Paciello and Wiederholt (2013) analyze the optimal monetary policy in a flexible-price economy when agents have limited information processing capacities. A standard rational inattention model with flexible prices has difficulty explaining why prices are fixed for some time (Mackowiak and Wiederholt, 2009). The flexible price assumption, without the aid of an additional mechanism such as higher order beliefs or persistent shocks, also gives rise to implausible prediction to inflation dynamics, such as an immediate and transient response, though the magnitude is dampened. My paper considers Taylor staggered pricing to incorporate the observed fact that firms do not adjust prices frequently.2 Not only does incorporating nominal rigidities account for the observed price duration and delayed response, it also introduces additional concerns in monetary policy implementation and policy effectiveness when there are interactions between nominal rigidities and rational inattention. The rest of the paper is organized as follows. Section 2 presents the baseline model and derives key equations regarding firms' policy-contingent pricing and information processing. Section 3 discusses analytically the effect and tradeoffs in monetary policy implementation, and derives the optimal monetary policy in the simple baseline model. Section 4 extends the analysis to consider a more general policy rule and shock dynamics. Section 5 concludes.
نتیجه گیری انگلیسی
This paper examines the policy implication for a rationally inattentive economy with nominal frictions in updating prices. The optimal monetary policy for this economy should fully offset the expected impact of natural rate shocks and completely stabilize prices, but complete price stabilization is not optimal in the presence of markup shocks. The central bank should carefully balance the need to accommodate the real effects of markup shocks and to offset the shock in order to maintain a stable price path, taking into account the coordination effect and the reduction of cross-sectional price dispersion due to information dispersion and price rigidity. In the short-run, the emphasis on price stabilization outweighs the emphasis on output stabilization after the shock, as stabilizing prices initially improves firms' knowledge and reduces the inefficiency caused by price stickiness and information dispersion.