Alternative measures of core inflation and the cost of disinflation

The sacrifice ratio is a crucial concept in macroeconomics that measures the cost, in terms of lost real GDP, of reducing trend inflation by one percentage point. The sacrifice ratio is particularly important for monetary policymakers who wish to estimate the real cost attaining a low and stable rate of inflation. Ball’s (1994) seminal work on sacrifice ratios introduced a new episodic way in which the sacrifice ratio can be measured. Thereafter the literature has spent a great deal of time examining the determinants of the sacrifice ratio, such as trade openness and central bank independence. One of the most robust findings of the literature is that “cold-turkey” disinflation minimizes output losses when trying to reduce inflation, where disinflation is less costly if conducted quickly. However this entire literature is based on the use of headline inflation to both identify disinflation episodes and measure the sacrifice ratio. Mazumder (2014) [forthcoming in the Journal of Macroeconomics] demonstrates that switching to core inflation—specifically the consumer price index less food and energy—produces completely different results in terms of the timing and duration of episodes, when examining OECD economics. Moreover, the arguments in favor of cold-turkey disinflation disappear, particularly when using measures of the sacrifice ratio that allow for hysteresis. In this paper we examine what happens when using several alternative measures of U.S. core inflation to identify episodes and to estimate both sacrifice and benefice ratios. Finally, when examining determinants of sacrifice and benefice ratios, we introduce a new measure of central bank credibility that focuses on the monetary authority’s ability to match actual inflation with expected inflation.