RDP 2015-09: Inflation Targeting: A Victim of Its Own Success? 1. Introduction

Despite one of the largest global recessions in decades during the financial crisis, global inflation barely budged. In some respects, this could be seen as a triumph for inflation targeting – inflation remained close to target despite some of the largest economic shocks in living memory. In the eyes of some, however, the financial crisis has demonstrated the weaknesses of inflation targeting. It has been argued that, in the face of record levels of unemployment in many economies, central banks should weigh unemployment outcomes much more heavily in their objectives. There have also been arguments that central banks, in responding to imported inflation shocks while domestic demand is depressed, or focusing on low headline inflation while asset prices are accelerating, have focused on inappropriate or misleading inflation measures. This paper makes the argument that these two, seemingly contradictory, outcomes – inflation remaining close to target and inflation targeting being heavily criticised – are a reflection of the general success of inflation targeting. Like a vaccination program, once the disease is effectively conquered, people begin to question the value of vaccination. This means that the communication challenges for central banks are magnified, but it doesn't necessarily mean that the vaccination program itself, inflation targeting, needs to be fundamentally re-engineered.

To reach this conclusion, we first look at the behaviour of inflation in Australia over the past 25 years or so since inflation targeting was introduced. While we look at data from Australia, reflecting our familiarity with the Australian experience, the findings are illustrative of a broader experience that is common across most inflation-targeting central banks, and our subsequent discussion is not specific to any one country (see IMF (2013)). We document significant changes in the behaviour of inflation over that time period: long-term inflation expectations have become firmly anchored at target inflation rates; the simultaneous flattening of the Phillips curve has contributed to a substantial reduction in the variability of prices affected by domestic monetary policy; and imported inflation now accounts for a much larger share of the variability in consumer price inflation than in the past, while also having less ongoing influence on inflation.

These changes in the inflation process have made CPI inflation a less reliable guide to the appropriate stance of monetary policy. Changes in CPI inflation are now more likely to reflect idiosyncratic shocks than changes in domestic economic conditions. Furthermore, the flattening of the Phillips curve, whether caused by or coincident with the adoption of inflation targeting, has complicated the task of identifying deviations in output from potential and, thus, forecasting inflation. Inflationary pressures arising from imbalances between demand and supply are smaller and more difficult to separate from idiosyncratic variation in inflation.

We consider the implications of these changes in the inflation process for the conduct of inflation targeting over the next 25 years. We focus our discussion around the central bank objective of maintaining price stability rather than also exploring the other major responsibility of central banks – financial stability. This is not to say that financial stability is not important. Rather, it is a sufficiently large topic that it would be difficult to do it justice within the same paper. Notwithstanding this, we do touch on financial stability considerations to the extent that financial stability can affect price or output stability. Thus, reflecting our focus on the price stability mandate, we discuss two particularly prominent proposals for change: either adopting explicit dual unemployment-inflation mandates; or changing the target to a measure more closely related to domestic economic conditions than CPI inflation. Our discussion emphasises that a breakdown in the correspondence between output and inflation stabilisation, caused in part by the success of inflation targeting, motivates these proposals for change and can help understand the perceived ‘failings’ of inflation targeting during the recent crisis. We conclude by suggesting some particular areas of the practice of central banking that will need to change and improve if inflation targeting is to celebrate its 50th anniversary 25 years from now. We do not recommend wholesale change, but there may be some scope for enhancements.