Sammendrag
The importance of monetary policy is unchallenged in modern macroeconomics and its prominent place in the theoretical discourse is unchallenged. Economists have offered several attempts to define the determinants of monetary policy in order to offer an optimal solution to a complex problem. Some well know examples of this in the literature are the Friedman rule, the Taylor rule and most recently central bank independence and inflation targeting (Bernanke and Mishkin, 1997). The main goal in the recent decades for monetary policy has been to achieve low and stable prices by stabilizing inflation. However, research has shown that a strict adherence to this policy can potentially have a destabilizing effect on the economy. In addition to the goal of price stability, an increasing number of central banks are clear in announcing their attempt to stabilize output. These tandem goals of monetary policy are often referred to as flexible inflation targeting. According to Svensson (2002) there is a general agreement that this is how central banks in industrialized countries conduct their policy.
With the main goals of stabilizing inflation and the output in mind, this thesis aims to investigate how these goals determine monetary policy. How much emphasis does a central bank put on each of these goals, and has the relative importance of them changed over time? Furthermore, if a central bank put more emphasis on the main goal of stabilizing inflation, would this lead to more volatility of the output?
To address these questions, the thesis will compare the actual policy interest rate with the outcome of two simple policy rules for 13 OECD member countries between 1990 and 2005. The first policy rule is based on a strict inflation-targeting approach. Here the central bank only attempts to reach the inflation target in the short-term, so the policy interest rate is according to this. The second policy rule is based on the well-known Taylor rule (Taylor, 1993). This rule suggests that the central bank puts some weight on both inflation variability and output gap variability, making it a flexible inflation-targeting policy rule (Svensson, 2002). The policy interest rates and the output-gaps of the policy-rules will then be compared to the outcome of the actual policy represented in graphs. The outcome of the strict-inflation targeting rule is calculated for all 13 countries, but due to restrictions in the data set the Taylor policy-rule is only calculated for 7 of these countries.
The approach outlined for this paper poses some challenges in how to construct the policy rules. For the policy rules to be effectively compared to the actual policy, they will have to be based on information that was available at the time when the central banks made their policy decisions. Knowing the exact information set of the central bank is out of the scope of this paper. In order to solve this problem, real-time data obtained from Consensus Economics Inc. (2010b) is used as forecast of the expected inflation with unchanged policy. The advantage with the data from Consensus Economics Inc. is that it is real-time and thus not revised ex post. These forecasts should be good proxies for the central bank forecasts, even if they are not likely to be identical.
The analysis implicates that monetary policy in the period between 1990 and 2005 appears to be best explained as following a flexible inflation-targeting approach. Comparing the actual policy with the two policy-rules shows that the Taylor-rule interest rate follows the actual interest relatively closely for the countries in the sample. The strict inflation-targeting rule is rather different from the actual monetary policy apart from when expected inflation is close to target. However, as the strict inflation-targeting rule exclusively focuses on controlling inflation the effect is an increase in output gap variability. Furthermore, the relative weight on the targets of the monetary policy appears to have shifted somewhat during the sample period, allowing for an increased focus on output stabilization.