4. Some of the recent weakness in core CPI and (to a lesser degree) core PCE inflation reflects residual seasonality in areas such as airfares and hotel accommodation that depressed prices sequentially in the summer and will
boost it in the fall. These are not the only distortions. The peculiar treatment of
health insurance in the CPI—which is based on health insurers’ profit margins—has been depressing core CPI inflation since last fall but will boost it starting in October. Conversely, the peculiar measurement of financial services inflation in the PCE index—which is partly based on stock prices—has been boosting core PCE inflation in recent months but will probably turn more neutral soon. Fortunately, these distortions largely occur in the “tails” of the price distribution, so statistical measures of underlying inflation that systematically exclude outliers can cut through them. And both on a CPI and PCE basis, trimmed-mean inflation—our favorite statistical measure of underlying inflation—sends a highly encouraging signal, with a sharp slowdown to the 2-2½% range in recent months. Thus, underlying inflation may already be near the Fed’s target, despite the ups and downs of commodity prices, traditional ex food and energy inflation, and the core services ex housing CPI or PCE measures (which by construction assign larger weights to these poorly measured components and are therefore even more distorted).