Is the Fed manipulating inflation numbers?

Could J.Powell and the Fed manipulating inflation statistics to their advantage? Let's take a look at the criteria used by the US central bank to guide its monetary decision making. In order to begin tapering the asset purchase program, the Fed looks closely at two measures, namely the unemployment rate and inflation. While the unemployment rate is relatively simple to gage, inflation is much more subjective, as measures can be very volatile (CPI: March = 2.6% - April = 4.2%) and differ greatly from one indicator to another. It makes sense for the Fed not to use CPI to guide its decision making. So what does it use?

The Consumer Price Index (CPI) measures core inflation. In other words, it looks at variations in the price of a basket of goods and services including things like transportation, food, medical care, housing, clothing, entertainment, education, communication... Meanwhile, the Core Consumer Price Index (Core CPI) measures core inflation, and excludes food and energy costs. This may not sound right, because everyone has to buy food and uses energy, right ? So why exclude it? These categories actually experience large temporary price changes, both up and down. So these two groups can significantly skew inflation numbers from month to month.

As you can see, the US central bank cannot base its decisions on CPI statistics. So it uses the Core CPI, you might say, but not quite. The Core CPI remains a relatively volatile measure, and basing the monetary policy of an entire country on it alone is very risky. It will therefore try to smooth this measure and remove extreme variations to obtain the least volatile and most accurate measure possible. However, what it considers the most accurate measure is actually rather far from it. Let me explain. In order to get inflation numbers, while reducing intramonth volatility, it will make some adjustments. Let's see what they are.

In August, there were larger increases and decreases in the prices of goods and services that make up the CPI than in July. So the CPI in August rose slightly but the median price fell. This is elementary, but extreme variations tend to impact too strongly on average, which distorts median measures. In order to overcome this issue, which can considerably distort inflation figures over a month, the Fed uses the Dallas Fed gauge. This eliminates 24% of the largest decreases but also 31% of the largest increases in price changes of personal consumption expenditures. It smoothes out inflation and gives us a measure close to or equal to our long term goal of 2%. But this gauge skews data in favor of categories with the lowest inflation - and that is not the only problem.

An adjusted average further smoothes inflation measures. The Fed compares price changes not over 12 months as the CPI does, but over 18 months. Thus, in addition to smoothing inflation a little more, J. Powell compares current levels to those of February - March - April 2020 (18 months) when inflation levels were close to 2.4%. However, 12 months ago, the CPI was between 1.2 and 1.7%.

Finally, the central bank uses two measures that exclude durable goods, which alone contributed 1% to the latest inflation figure

Evolution of Consumer Price Indices over 12 months :

Source: Marketscreener

It is understandable that the Federal Reserve does not use changes in the CPI or even the 12-month Core CPI to guide its decision making. Indeed, short-term supply and demand are subject to too much variation. This is why it seeks to smooth inflation over time by comparing it over 18 months and excluding durable goods and certain goods that have moved sharply up or down.

However, the Dallas Fed's adjusted average, removing the highest 31% and lowest 24% of price changes in personal consumption expenditures, seems a bit too accommodative. Taking the Cleveland Fed's adjusted average, removing only the highest 16% and lowest 16%, you get 3.2% inflation over the last 12 months... which is higher than the Fed's 2% target...