RDP 9807: Inflation Targeting in a Small Open Economy 7. Conclusion

This paper analyses inflation targeting in a small open economy. In line with more recent literature we emphasise the importance of the exchange rate channel, in conjunction with the conventional aggregate demand channel. Transmission of policy via the exchange rate works with greater speed, reducing the control lag with which the central bank is able to achieve its inflation target. However, targeting inflation at a shorter horizon may induce considerable volatility in both output and interest rates.

To investigate these issues more thoroughly, we developed a model for an open economy that produces traded and non-traded goods. We argue that this distinction is the relevant one for a small open economy, such as Australia, and is consistent with empirical evidence. We compared the behaviour of relevant macroeconomic variables under two regimes – aggregate inflation targeting and non-traded inflation targeting.

A feature specific to our analysis is our ability to isolate the impact of different shocks. In line with Svensson (1998), we present unconditional variances calculated based on the combination of all shocks hitting the economy. However, we also present the variance of key variables, conditional on individual shocks to the exchange rate, non-traded inflation and demand. This provides valuable information, particularly in the case where the relative size of different shocks is uncertain (that is, the average size of exchange rate shocks relative to supply and demand shocks).

Compared with aggregate inflation targeting, targeting non-traded inflation allows the central bank to tolerate more extreme movements in the exchange rate. For an exchange rate shock, targeting aggregate inflation implies a larger interest rate response. This reduces the volatility of the exchange rate, at the cost of greater variability in output and non-traded inflation. In contrast, the policy response to supply or demand shocks is greater in the case of non-traded inflation targeting. This allows non-traded inflation and output to return to equilibrium more rapidly, at the cost of greater variability in both the exchange rate and aggregate inflation.

Our results are robust to alternative specifications of the basic model. We examined models with forward-looking inflation expectations, and gradual exchange rate pass-through to domestic prices of traded goods. Also, we considered the case of discretionary policy where the central bank is unable to pre-commit to the path of its instrument. These variations did not alter the main findings of the paper.

We are unable to provide an answer as to whether aggregate inflation targeting is strictly preferable to non-traded inflation targeting. We have considered an institutional framework in which the central bank sets policy according to either one of two simple objective functions. However, it is plausible that society has preferences for avoiding excessive volatility in product and financial markets which are not adequately captured by either of these simple objective functions. Although it is beyond the scope of this paper to formally model social preferences, we could imagine that if we had an explicit form for the social welfare function then we would be able to evaluate whether or not aggregate inflation targeting is preferable to non-traded inflation targeting. In this paper we have demonstrated that the result of this comparison will depend not only on the exact form of the social welfare function, but also on the nature, and relative size of the shocks hitting the economy.

An inflation target with a more medium-term perspective would be one way of avoiding the extreme outcomes that can be associated with standard flexible aggregate and non-traded inflation targeting regimes (that is, with some preference for output stability and interest rate smoothing). In the case of an exchange rate shock, aggregate inflation would be allowed to move away from target in the short term, thereby lessening the need to generate volatility in non-traded inflation, output and interest rates. In the case of a supply shock (that is, a shock to non-traded inflation), the return to target could be more gradual, thereby avoiding extreme movements in interest rates and the exchange rate. The same is true of a demand shock.

A goal of price stability over the medium term may appear to undo some of the benefits of having a more narrowly defined inflation objective for the central bank. However, there is likely to be a trade-off here; if the objective is too restrictive and creates excessive volatility in either financial or product markets, then the credibility of the regime will no doubt suffer.