تجزیه و تحلیل کمی از شوک های قیمت نفت، سیاست پولی سیستماتیک، و رکود اقتصادی

Are the recessionary consequences of oil-price shocks due to oil-price shocks themselves or to the monetary policy that responds to them? We investigate this question in a calibrated general equilibrium model in which oil use is tied to capital utilization. The response to an oil-price shock is examined under a variety of monetary policy specifications. Under our benchmark calibration, which approximates the Federal Reserve's behavior since 1979, monetary policy contributes about 40 percent to the drop in output following a rise in oil prices. Moreover, none of the commonly proposed policies we examine completely offsets the recessionary consequences of oil shocks.

Are the recessionary consequences of oil-price shocks due to oil-price shocks themselves or to the monetary policy that responds to them? Fig. 1 plots the federal funds rate and oil-price increases over the postwar period. The shaded areas in the graph are NBER recessions. Oil prices generally rise prior to recessions but so does the funds rate, suggesting that contractionary monetary policy may play a role in downturns as well. Recent work by Bernanke et al. (1997) argues that monetary policy plays the larger role: a monetary policy different from that in place during the 1970s, i.e., a fixed nominal interest rate policy, could have largely eliminated the negative output consequences of the oil-price shocks on the U.S. economy. This view has, in turn, been challenged by Hamilton and Herrara (2000), who argue that Bernanke, Gertler, and Watson's (BGW) empirical results are driven by model misspecification. Hamilton and Herrara reimplement the BGW experiment using a different model specification and find that it is the increases in the price of oil that lead directly to contractions in real output: contractionary monetary policy plays only a secondary role in generating the downturn. Thus, while there is widespread agreement that oil-price shocks have been an important factor for the volatility of real output in the postwar period, there is less agreement on the channel of transmission.We explore the role of alternative monetary policies in amplifying or dampening the economy's response to oil-price shocks in the framework of a dynamic, stochastic, general equilibrium model. This approach contrasts with much of the existing analysis that has examined interaction between oil-price shocks and monetary policy using reduced-form vector autoregression (VAR) models. These empirical models are largely silent on the channels through which oil-price changes affect real output. Further, any VAR-based analysis of the reaction of the economy to oil-price shocks under alternative monetary policy specifications runs squarely into the Lucas critique: It is problematic to assume that reduced-form coefficients are stable across different policy regimes. Sims (1997), in his discussion of the BGW paper, points out some of the difficulties of basing alternative policy simulations on reduced-form estimates, especially when the alternative policies considered are far from historical experience. Indeed, as Sims points out, the fixed-interest-rate rule simulation in BGW imply explosive behavior in prices. Our contribution is to use a model based on the primitives of preferences, technology, and the stochastic processes governing shocks to examine economic responses to oil-price shocks under a variety of specifications for monetary policy rules. We quantify the relative importance of both oil-price shocks and monetary policy as contributing factors to economic downturns under these different policy rules in a model calibrated to match features of the U.S. economy. Our findings suggest that while the monetary policy rule in place can contribute to the magnitude of the negative output response to an oil-price shock, the direct effect of the oil-price increase is the more important factor. We do not find rules that can replicate BGW's result that monetary policy can eliminate the recessionary consequences of oil-price shocks. Our results do suggest that better economic outcomes are likely if policymakers focus on offsetting the inflationary consequences of oil-price shocks rather than focusing on stabilizing output. We find that the recessionary consequences of oil-price shocks are smallest when the central bank targets the price level. The model generates worse outcomes for both output and inflation when policy is weighted toward output stabilization. For example, we investigate the economy's response to oil-price shocks under pre- and post-1979 monetary policy rule specifications. The output and inflation responses are larger under the pre-1979 rule, which places relatively more weight on output. Our model has households and monopolistically competitive firms operating in a limited participation environment. We examine cases with and without nominal rigidities. Oil usage is tied to the variable rate of capital utilization. If capital is used more intensively, the energy requirement for production rises. The price of oil is assumed exogenous with oil-price increases acting as a tax on the economy: none of the expenditure on oil is recirculated to the economy. As a consequence of the model environment and the way in which oil use is introduced, oil-price shocks do have symmetric effects in our model, a symmetry that is not evident in empirical studies of the economic consequences of oil-price shocks. Those studies find that oil-price increases have a much larger effect on output than do oil-price decreases. However, our focus is on the recessionary consequences of positive oil-price shocks and on how monetary policy interacts with those shocks to amplify or dampen output and inflation responses. While we are unable to capture the asymmetric response of output to oil-price shocks, the model does generate what look like recessions following an oil-price increase. Though we focus on the interaction of oil-price shocks and monetary policy, our exercise is related to a literature that examines how oil-price shocks affect economic performance in non-monetary economies. Kim and Loungani (1992) and Finn (1995) investigate the contribution of energy price shocks to business cycle dynamics. They conclude that energy price shocks account for up to 20 percent of real output volatility at business cycle frequencies. Rotemberg and Woodford (1996) study the decline in output and response of real wages to an oil-price increase. They conclude that imperfect competition is important in generating model responses that match those found in the data. We find that the relative contribution of oil-price shocks and monetary policy to economic downturns is similar in both the flexible and sticky price versions of the model. We also find that TFP shocks are not always a good proxy for oil-price shocks. Under a monetary policy rule that targets prices, the response of output and inflation to oil-price shocks is different from the response to TFP shocks because the latter lead to smaller changes in relative prices. Hence, the explicit introduction of an oil sector is important in quantifying the respective contributions of oil-price shocks and monetary policy responses to economic downturns. The rest of the paper is organized as follows. Section 2 presents the model, and Section 3 describes its calibration. Section 4 presents the simulation and sensitivity results, while Section 5 concludes.

The model examined in this paper suggests that central banks cannot fully insulate their economies from the consequences of oil-price shocks, but that the way monetary policy is conducted plays a significant role in how the consequence of oil-price shocks play out in the economy. We examine a variety of monetary policy rules in a calibrated general equilibrium model and find none of the rules is able to completely offset the negative output consequences of positive oil-price shocks. Easy inflation policies are seen to amplify the impacts of oil-price shocks on output and inflation while policies that place more weight on stabilizing inflation or prices lead to more damped output and inflation responses. We find that the recessionary consequences of oil-price shocks are smallest when the central bank targets the price level. A version of the model that uses an interest rate rule calibrated to match that followed by U.S. monetary authorities in the post-1979 era implies that nearly 40 percent of the decline in output following a positive oil-price shock can be attributed to the way monetary policy responds to the shock. A rule calibrated to the pre-1979 era, which places relatively more weight on the output gap, results in up to 75 percent of the drop in output being attributed to the monetary policy response and only 25 percent to the direct effect of the oil shock. Somewhat improved outcomes were obtained under rules that peg either prices, interest rates, or inflation. Our finding that output and inflation are less volatile when policymakers place relatively more weight on inflation stabilization is shared by a number of other studies. Clarida et al. (2000) analyze alternative Taylor rule specifications in a sticky-price model and find that self-fulfilling inflations are less likely to occur when the monetary authority concentrates on stabilizing inflation rather than stabilizing output. Orphanides (2001), who estimated forward-looking Taylor rules for the U.S. economy, attributes the poor inflation outcomes of the 1970s to two factors: the Federal Reserve placed too much weight on an output gap and that output gap proved to be mismeasured. In this sense, according to Orphanides, the Federal Reserve was too activist in responding to the output gap rather than too weak in responding to expected inflation. Although we focused primarily on the output consequences of oil-price shocks under alternative policy rules in an environment without self-fulfilling expectations or mismeasurement, our conclusions are broadly similar to those in the other studies—policies that focus on stabilizing inflation lead to better outcomes for inflation and output.