The importance of output for the monetary policy rules

Abstract

We examine the importance of variable output (Gross Domestic Product (GDP) growth rate) on the preferences of the policy makers. We do so by examining the effect of output regimes on the form of monetary policy implemented by the central banks. We use simple monetary policy rules to characterize the monetary policy of the central bank. Regime switching models are utilized to model output and monetary policy regimes separately. The importance of output is then examined for the UK and the USA by corresponding the output and monetary regimes together. We find some evidence supportive of the fact that the phase of output existent can change the preferences of the policy makers.

Notes

¹As McCallum (2000) argues that reliance of a policy rule upon any output gap measure is risky for different measures give quite different values and there is no professional consensus on an appropriate measure or even the concept itself.

²The AR(2) process for the UK (as opposed to AR(4)) was selected on the basis of various diagnostic test statistics, including Akaike Information Criterion (AIC) and Schwartz Information Criterion (SIC) and the significance of the t- and F-statistics.

³The data utilized in the study have been extracted from DataStream International.

⁴Filtered probabilities refer to inferences about s_t conditional on information up to t (Pr[s_t  = jψ _t ]), whereas smoothed probabilities refer to inferences about s_t conditional on all information in the sample (Pr[s_t  = jψ _T ]).

⁵The interest rate used for the UK is the interbank overnight interest rate and for the USA is the Federal Funds Rate. The data on the interbank overnight interest rate for the UK are only available from 1975 onwards in the Organization for Economic Co-operation and Development (OECD) database, therefore the estimation sample for the reaction function is different from the sample utilized in the previous section. The same is true for the USA as well.

⁶In Ox the software used in the estimation of Markov-switching processes, models such as Equation 4(4) are characterized as ‘MSIA’ i.e. Markov-switching process in intercept term and autoregressive parameters. See Krolzig, H. M. (1998) Econometric modelling of Markov-switching vector autoregressions using MSVAR for Ox, Working Paper, Oxford University.

⁷The coefficient on the inflationary gap should theoretically be much greater than one for an inflation-targeting central bank and at least this is what we expect for an inflation-targeting regime like the Bank of England. A coefficient of less than 1 on inflationary gap is, however, also reported in some other empirical studies for the UK as well. See, for instance, Clarida et al. (1998).

⁸For some of the regimes, for the two countries, the coefficient of the output growth variable is negative which is intuitively a bit hard to digest. It is, however, plausible if the central bank has a target growth rate of real GDP or it could be because of the inflationary bias inherent in the discretionary monetary policy, as documented in the literature.

⁹The early 1990s correspond to a regime 1 reaction function, however as mentioned earlier our model for the business cycle misses a small recessionary period in the early 1990s as well.

¹⁰One possible explanation for insignificant coefficients on the output for some regimes in the reaction functions could be the use of growth rate of output as opposed to output gap.