Monday, January 30, 2023

Core Inflation Woes — Brian Romanchuk

Alex Williams recently wrote “What Is ‘Core PCE Services Ex-Housing’ Anyway?,” which dissects the measure that the Fed is using to get a handle on “underlying” inflation. The most interesting bit (for me) is that about 1/4 of this measure is an imputed price index, based on wages. This means that this component will track wages (giving a convenient analytical relationship) by definition.

The logic of following this measure is that the Fed convinced itself that the core (ex-food and and energy) personal consumption expenditure is the best measure of “underlying” inflation, but it turns out that the housing part of that has construction issues (too smoothed to pick up current events), and so they wanted to strip that out of the measure.

This measure is obviously problematic. Its use reflects intellectual herding: following the methodologies of previous academics, without stopping to ask basic “why are we doing this?” questions. This is a typical feature of areas of academic failure. Since it is impossible to come up with useful results, the “publish or perish” imperative means that what gets published are marginal changes to an existing literature that has serious defects. (Source for that assertion: what I saw in my career in academia.)...
Bond Economics
Core Inflation Woes
Brian Romanchuk

No comments: