Asymmetric monetary policy rules in Australia

ABSTRACT

In this study, the hypothesis that the Reserve Bank of Australia (RBA) implements an asymmetric monetary policy rule is tested. We estimate both linear and asymmetric monetary policy reaction functions for the period before inflation targeting was adopted, for the period when inflation targeting was explicitly adopted and for the full sample period. The results of the linear monetary policy rules are consistent with the estimates reported from other studies that estimate linear monetary policy rules for Australia. On the other hand, the results of estimating the asymmetric monetary policy rules for the pre-inflation targeting period shows that the RBA had reacted symmetrically, suggesting that it had acted with the same aggressiveness towards both inflation and output gaps of the same magnitude, over both phases of the business cycle. However, for the inflation targeting period, the results show that the RBA had reacted asymmetrically in its policy response to the inflation gap, output gap or both. A similar result is found for the full sample period. This asymmetric response supports the view that a non-linear monetary policy rule emanated from asymmetric preferences, rather than from the existence of a non-linear Phillips curve.

KEYWORDS:

• Monetary policy
• asymmetric monetary policy rule
• asymmetric preferences
• non-linear Phillips curve

JEL CLASSIFICATION:

• E52
• E58

View correction statement:

Erratum

Acknowledgements

The author would like to thank Benjamin Durance for excellent research assistance. Work on this project was financed by a generous grant from the Australian Centre for Financial Studies, which the author gratefully acknowledges. Any remaining errors are the responsibility of the author.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ The neutral nominal rate of interest, $\bar{i}$, is defined as the sum of the neutral real interest rate and the previous period’s inflation rate, so that $\bar{i}=\bar{r}+{\pi }_{t-1}$.

² For more on the phenomenon of a deflationary bias, see Mishkin (1999, 2000, 2002) and Mishkin and Posen (1997), who document this effect for a number of central banks around the world, particularly for the Bank of Canada and the Bank of England.

³ Interestingly, Orphanides and Van Norden (2005) highlight the problems of using final data to estimate monetary policy rules. However, Croushore and Evans (2006) show that the results obtained using final data are very similar to those obtained from real-time data, suggesting that there is no serious limitation in using final data.

⁴ This database is available at http://fbe.unimelb.edu.au/economics/macrocentre/artmdatabase/database_and_documentation.

⁵ Grenville (1997) states that implementation of monetary policy was secondary to the incomes and wages policy used by the Australian Government for achieving price stability.

⁶ For more on the checklist approach to monetary policy in Australia, see Morgan (1990) and Stemp (1991).

⁷ Other studies use similar values. For instance, De Brouwer and Gilbert (2005) impose an inflation target of 4.70 per cent for the 1980s, and a value of 9.30 per cent for the full sample period, while Leu and Sheen (2006) use an arbitrary value of 5.40 per cent for the pre-inflation targeting period, and a value of 2.50 per cent for the inflation targeting period.