Inflation Targeting vs. Price-Path Targeting: Looking for Improvements

The past 15 years have been dramatic ones for central banks around the world. Increases in independence and transparency have been coupled with a shift in focus. Price stability is now the paramount objective for the vast majority of modern central bankers. Combined, these changes in central bank structure and policy framework have yielded substantial benefits. Low and stable inflation has brought with it high and stable growth. Taking recent successes as a starting point, this paper looks at the possibility for further improvements. Could countries benefit by shifting from inflation targeting to price-path targeting? The answer depends on the structure of each country's economy. Specifically, how slowly output growth returns to its sustainable growth rate following a movement away. The more persistent output deviations from potential, the more likely it is that a country can benefit from a shift to price-path targeting. But price-path and inflation targeting are not the only choices. Combinations are also possible. Hybrid rules that combine the two extremes are in fact optimal in most cases. While inflation targeting implies that the (log) price level will be a random walk, and price-path targeting means that prices will quickly revert to their target and be as close to white noise as possible, hybrid rules imply a specific level of persistence. Following the work of Cecchetti and Kim (2005), our discussion focuses on the hybrids, deriving the optimal rule for a broad set of countries. Building on this earlier work, we go on to show to study the horizon over which price-path or inflation targets should be evaluated. Building on intuition first provided in King (1999), we examine the equivalence between a rule with a high weight on inflation targeting that is evaluated only infrequently, and one with a high weight on the price-path that is evaluated often. For each country, we are able to derive a horizon for target evaluation that would result in optimal policy. The general result is that the less persistence output deviations from potential, the shorter the horizon for optimal target evaluation.download