The Bureau of Economic Analysis (BEA) today released its report on Personal Income and Outlays with data for February 2024, which includes February numbers for the Personal Consumption Expenditure (PCE) price index. While the Bureau of Labor Statistic’s (BLS) Consumer Price Index (CPI) gets most of the attention, and is released in a more timely manner, it is the PCE price index, or rather the inflation rate calculated from the PCE price index, that is the Federal Reserve System’s target for monetary policy.
The CPI with data through February 2024, and specifically the CPI for All Urban Consumers (CPI-U), was released on March 12, and by that measure the year-over-year inflation rate, measured over the twelve months from February 2023 to February 2024, was 3.2%, a slight uptick from the 3.1% recorded for January 2024. The PCE price index released today measures inflation over the same period at 2.5%, also a slight uptick from the 2.4% recorded for January 2024.
The Federal Reserve System targets the inflation rate to be 2%, measured by the twelve-month change in the PCE price index. This latest reading is consistent with the general gradual decline in the inflation rate since the 7.1% reading in June 2022. For that matter, the CPI inflation measure has also declined from its recent high of 9.0% in June 2022. At the same time, both measures showed the smallest of increases in these February measures.
As we have noted in prior posts, the CPI and PCE price index are both intended to measure the level of prices paid by consumers, but they differ in several ways. The PCE price index is a broader index that includes prices for some items not directly paid by consumers but paid on behalf of consumers. A prime example is medical expenses - the CPI includes out of pocket expenditures by households for medical services. The PCE price index includes medical services for consumers that are paid by employer-provided health insurance and by Medicare and Medicaid. Given this, the weights attached to the various items included in the two indices differ, in part because the PCE price index includes more items that must be assigned weights. Even among similar items, the weights used in the two indices can differ, as they are calculated from different surveys, different sources. Finally, there is a technical difference in how the two indices are constructed, as the CPI uses a “Laspeyres” formula that is easier to construct in a timely manner, while the PCE price index is based on a “Fisher-Ideal” formula. Economists in general prefer the Fisher-Ideal formula on theoretical grounds, but the Laspeyres formula can be implemented more quickly and easily.
How much do these differences matter? The answer depends on the issue being addressed. For monetary policy, the PCE price index is the announced target. We might ask, if the PCE price index was hitting the Federal Reserve System’s announced target, what would we expect to be the CPI measure of inflation? If we look at historical data, the CPI measure of inflation tends to run about 0.4% higher than the PCE measure. From January 2014 to February 2024, the CPI inflation measure indicates 2.8% inflation. Over that same period, the PCE price index measure indicates 2.4% inflation. We might conclude that, were the Federal Reserve System to achieve its 2% target for PCE inflation, the CPI measure would run, on average over time, at roughly 2.4%.
There are more wrinkles, of course. At times the Federal Reserve System talks about the PCE inflation measure less food and energy items, sometime labeled as “core PCE” inflation. The idea behind looking at inflation less food and energy is that energy prices and to some extent food prices are commodities with the potential for large price swings that lead to periods where the more inclusive inflation measures are temporarily high or low, arguably distorting the picture of the longer term underlying inflation rate. A further wrinkle is that there is a core CPI inflation measure as well.
The BEA releases data on the core PCE price index at the same time as the main or “headline” PCE price index, and the BLS similarly releases the CPI and the core CPI in the same data release. Core PCE inflation over the past twelve months ending February 2024 was 2.8%, rather higher than the headline reading of 2.5% mentioned above. This February measure is a slight decline from the 2.9% reading in January 2024. As for the CPI, core CPI inflation was 3.8% as compared to the headline figure of 3.2%, and this core CPI reading was also a slight downtick from the 3.9% core CPI measure in January 2024. Basically, these core measures indicate that food and energy prices have not been increasing as much as other prices over the last twelve months, and have thus served to reduce the measure of headline inflation. In other words, leaving out food and energy prices actually leads to higher measured inflation.
Why do we focus on core inflation measures? The basic idea is that inflation is volatile and hard to forecast, and part of this is due to the volatility of food and energy prices. While rents and doctor fees may increase at somewhat infrequent intervals, energy prices such as gasoline may change daily or weekly, and to some extent this is also true for food prices. If a forecaster is looking to find out what is going to happen to the overall level of prices going forward, then it might help to ignore or downweight observations on food and energy that frequently change. In terms of recent data, core inflation measures are stubbornly above the headline measures, which should also give pundits pause in claiming the inflation problem is nearly over.
In any case, the prices consumers are actually paying are measured by the headline price index, whether it be the CPI or the PCE price index. Core measures are arguably useful for forecasting future inflation (although that is a matter debate) but by leaving out food and energy they do not measure the overall prices consumers are paying.