This is an excerpt from our most recent Economic Outlook report. To access the full PDF, please click here
I. Introduction
The USA has 2 primary measures of prices paid by consumers, for goods and services.
These 2 consumer price index groups get constructed by separate Federal macroeconomic bureaus -- and behave differently -- over time.
To begin to understand their differences…
A major upfront difference is: the Personal Consumption Expenditure (PCE) index group measures goods and services bought by all U.S. households and nonprofits. These “nonprofits” include the Federal, State, and Local governments.
Consumer Price Inflation (the CPI) taps into just urban households.
This upfront reason is why, month-to-month, Fed Chair Powell and voting members of the FOMC rely more upon the price data flowing in from PCE indices; than data flowing in from CPI indices.
- The PCE gets prepared by the U.S. Bureau of Economic Analysis (BEA). The PCE uses the macroeconomic accounting data from the gross domestic product (GDP) reporting system, and from suppliers.
From the BEA chart portal, here is y/y percent change in the broad PCE from Q1-2020 to Q2-2022 (below):
Image Source: Zacks Investment Research
You can see.
The BEA’s PCE rate, drawn directly from the GDP accounts, begins to rise NOTABLY in Q2-2021, well above the Fed’s +2.0% y/y statutory threshold to well above +6.0%. This rising PCE trend actually starts in Q4-2020.
At the very least, Q1 or Q2 of 2021 is when a data-driven Fed, using the PCE measure of consumer price inflation should have acted, by raising the Fed Funds rate.
A truly proactive FOMC would have moved 6 months in advance of that. Perhaps in September of 2020. They first rate hike in this cycle happened March-April 2022.
Appropriately, that 1st hike was 18 months late. Conservatively, it was 12 months late.
- The Consumer Price Index (CPI) gets produced by the U.S. Bureau of Labor Statistics (BLS). The CPI uses data from household surveys. The CPI only accounts for all urban households.
- I show you the Bureau of Labor Statistics CPI for All Urban Consumers, not seasonally adjusted, from 2012 to 2022 (below).
Image Source: Zacks Investment Research
In Feb. 2021, this BLS’s broad CPI data was running at a +1.7% year-on-year rate. In March 2021, this was +2.6%. In April 2021, it was +4.2%.
Once again, we see.
The Q2-2021 moment in macroeconomic time was when U.S. COVID vaccination campaigns had reached a critical threshold.
U.S. mobility and demand were exploding. Particularly for urban households.
II. Beginning to Understand the Differences Between PCE and CPI
The CPI probably gets more press, in that it is used to adjust Social Security payments and is also the reference rate for some financial contracts, such as:
- Treasury Inflation Protected Securities (TIPS), and
- Inflation Swaps
The Federal Reserve, however, states its goal for inflation in terms of the PCE.
The two measures, though following broadly similar trends, are certainly not identical.
The CPI and PCE each come in two flavors, a so-called “headline” or “broad” measure and a “core” measure, which strips out the more volatile food and energy components.
Over the short term, the core measure may give a more accurate reading of where inflation is headed. But people do buy food, fill up their gas tanks, and heat their homes, so headline inflation more accurately represents people’s actual expenses.
Like the headline measures, core CPI tends to show higher inflation than core PCE.
In general, the CPI tends to report somewhat higher inflation.
- From 2000 to 2014, prices measured by the core CPI rose by +39%, while those measured by the core PCE Rose by +31%, leading to differing average annual inflation rates of +2.4% and +1.9%.
- From 2000 to 2014, broad CPI averaged annual increases of +3.9%, and broad PCE has averaged +3.4% the same half a percentage point difference as between the core numbers.
- In this century, then, CPI inflation has run about half a percentage point higher than PCE inflation. When calculated from 1960 the difference is almost the same, +3.9% for the broad CPI and +3.3% for the broad PCE.
Why does PCE end up being lower than CPI?
We need to do a deeper dive to explain that phenomenon.
III. Categorizing the Differences Between PCE and CPI
The differences between the CPI and PCE measures of inflation can be summarized into four categories or effects.
(1) Scope effects
This conceptual difference means some items and expenditures in the PCE index are outside the scope of the CPI.
- The PCE index measures the change in goods and services consumed by all households, and nonprofit institutions serving households.
- The CPI measures the change in the out-of-pocket expenditures of all urban households.
For example, medical care services in the PCE index include services purchased out-of-pocket by consumers, and those services paid for, on behalf of consumers. This then includes medical care services paid for by employers through employer-provided health insurance, as well as medical care services paid for by governments through programs such as Medicare and Medicaid.
In contrast, the expenditure weights for medical care services in the CPI are derived only from out-of-pocket expenses paid for by consumers.
These differences can also be isolated and measured, and can be referred to as “scope effects.”
(2) Weight effects
Relative weights assigned to each of the CPI and PCE categories of items are based on different data sources.
- The relative weights used in the CPI are based primarily on the Consumer Expenditure Survey, a household survey conducted for the BLS by the Census Bureau.
- The relative weights used in the PCE index are derived from business surveys—for example:
- The Census Bureau’s annual and monthly retail trade surveys
- The Service Annual Survey, and
- The Quarterly Services Survey
In order to estimate a “weight effect,” or the effect of using different weight sources on CPI and PCE index changes, CPI relative weights for comparable item categories can be used to estimate the PCE fixed-weight price index.
(3) Formula effects
The CPI and the PCE index are constructed from different index-number formulas.
- The CPI index is an average based on a Laspeyres formula, whereas
- The PCE index is based on a Fisher-Ideal formula.
A Fisher-Ideal index is considered a “superlative” index. It reflects consumer substitution among detailed items as relative prices change.
In practice, superlative indexes are difficult to implement in real time because such indexes require expenditure data for the period that is current, and such data are not available. For example, data on household consumer expenditures that are used to estimate the CPI are not available for the current period.
For the Consumer Price Index for Urban Consumers (CPI-U), a Laspeyres index provides an alternative to the Fisher-Ideal index.
To estimate a “formula effect,” or the differences in the rates of growth between the CPI and PCE caused by the differences in formula, the detailed price and quantity data used to estimate the PCE index can be reaggregated with the use of a Laspeyres price-index formula.
(4) Other effects
A variety of remaining differences consisting of seasonal-adjustment differences, price differences, and residual differences.
These must be taken into account for a complete understanding of the differences between the CPI and the PCE index.
- For example, the PCE index for airline fares is based on passenger revenues and the number of miles traveled by passengers.
- The CPI, however, is based on prices charged for air travel for sampled routes.
IV. The Fed’s Thinking
January 2012 was when the Fed announced it would use the PCE as its primary measure of inflation, preferring it for three primary reasons:
(1) Scope is better: PCE includes more comprehensive coverage of goods and services.
(2) Weighting is better: Expenditure weights in the PCE can change as people substitute away from some goods and services toward others.
(3) Data revisions are better: PCE data can be revised more extensively than the CPI, which can only be adjusted for seasonal factors, and only for the previous five years.
Thus, if the price of bread goes up, people buy less bread, and the PCE uses a new basket of goods that accounts for people buying less bread. The CPI, however, is less fluid in response to changing consumer preferences.
In summary, the CPI represents a basket of goods and services that a consumer would buy without making substitution changes when prices change.
The PCE encompasses a broader range of goods and services than the CPI, from a broader range of buyers. It tries to track what is actually purchased, and represents how consumers change their buying patterns when relative prices change.
This leads to smoother price changes in the PCE and typically lower levels of reported inflation, at least as experienced by consumers.
V. Conclusion: Other Fed Computations of Consumer Price Inflation
While the PCE is considered better than the CPI for Fed purposes, that has not stopped the Federal Reserve from computing still more consumer price indexes.
Next, I show you three others.
(1) The Trimmed Mean PCE Inflation Rate produced by the Federal Reserve Bank of Dallas is an alternative measure of core inflation in the price index for personal consumption expenditures (PCE).
This data series is calculated by the Dallas Federal Reserve Bank, using data from the U.S. Bureau of Economic Analysis (BEA).
Image Source: Zacks Investment Research
Calculating the trimmed mean PCE inflation rate for a given month involves looking at the price changes for each of the individual components of personal consumption expenditures.
The individual price changes are sorted in ascending order from “fell the most” to “rose the most,” and a certain fraction of the most extreme observations at both ends of the spectrum are thrown out or trimmed. The inflation rate is then calculated as a weighted average of the remaining components.
The trimmed mean inflation rate is a proxy for true core PCE inflation rate. The resulting inflation measure has been shown to outperform the more conventional “excluding food and energy” measure as a gauge of core inflation.
(2) Median Consumer Price Index (CPI) is a measure of core inflation calculated the Federal Reserve Bank of Cleveland and the Ohio State University.
Median CPI was created as a different way to get a 'Core CPI' measure, or a better measure of underlying inflation trends.
Image Source: Zacks Investment Research
To calculate the Median CPI, the Cleveland Fed analyzes the median price change of the goods and services published by the BLS.
The median price change is the price change that’s right in the middle of the long list of all of the price changes. This series excludes 49.5% of the CPI components with the highest and lowest one-month price changes from each tail of the price-change distribution resulting in a Median CPI Inflation Estimate.
According to research from the Cleveland Fed, the Median CPI provides a better signal of the inflation trend than either the all-items CPI or the CPI excluding food and energy. According to newer research done at the Cleveland Fed, the Median CPI is even better at PCE inflation in the near and longer term than the core PCE.
(3) Sticky Price Consumer Price Index (CPI) is calculated from a subset of goods and services included in the CPI that change price relatively infrequently.
Because these goods and services change price relatively infrequently, they are thought to incorporate expectations about future inflation to a greater degree than prices that change on a more frequent basis.
One possible explanation for sticky prices could be the costs firms incur when changing price.
Image Source: Zacks Investment Research
Image: Bigstock
A Deep Dive Into Consumer Price Inflation
This is an excerpt from our most recent Economic Outlook report. To access the full PDF, please click here
I. Introduction
The USA has 2 primary measures of prices paid by consumers, for goods and services.
These 2 consumer price index groups get constructed by separate Federal macroeconomic bureaus -- and behave differently -- over time.
To begin to understand their differences…
A major upfront difference is: the Personal Consumption Expenditure (PCE) index group measures goods and services bought by all U.S. households and nonprofits. These “nonprofits” include the Federal, State, and Local governments.
Consumer Price Inflation (the CPI) taps into just urban households.
This upfront reason is why, month-to-month, Fed Chair Powell and voting members of the FOMC rely more upon the price data flowing in from PCE indices; than data flowing in from CPI indices.
From the BEA chart portal, here is y/y percent change in the broad PCE from Q1-2020 to Q2-2022 (below):
Image Source: Zacks Investment Research
You can see.
The BEA’s PCE rate, drawn directly from the GDP accounts, begins to rise NOTABLY in Q2-2021, well above the Fed’s +2.0% y/y statutory threshold to well above +6.0%. This rising PCE trend actually starts in Q4-2020.
At the very least, Q1 or Q2 of 2021 is when a data-driven Fed, using the PCE measure of consumer price inflation should have acted, by raising the Fed Funds rate.
A truly proactive FOMC would have moved 6 months in advance of that. Perhaps in September of 2020. They first rate hike in this cycle happened March-April 2022.
Appropriately, that 1st hike was 18 months late. Conservatively, it was 12 months late.
Image Source: Zacks Investment Research
In Feb. 2021, this BLS’s broad CPI data was running at a +1.7% year-on-year rate. In March 2021, this was +2.6%. In April 2021, it was +4.2%.
Once again, we see.
The Q2-2021 moment in macroeconomic time was when U.S. COVID vaccination campaigns had reached a critical threshold.
U.S. mobility and demand were exploding. Particularly for urban households.
II. Beginning to Understand the Differences Between PCE and CPI
The CPI probably gets more press, in that it is used to adjust Social Security payments and is also the reference rate for some financial contracts, such as:
The Federal Reserve, however, states its goal for inflation in terms of the PCE.
The two measures, though following broadly similar trends, are certainly not identical.
The CPI and PCE each come in two flavors, a so-called “headline” or “broad” measure and a “core” measure, which strips out the more volatile food and energy components.
Over the short term, the core measure may give a more accurate reading of where inflation is headed. But people do buy food, fill up their gas tanks, and heat their homes, so headline inflation more accurately represents people’s actual expenses.
Like the headline measures, core CPI tends to show higher inflation than core PCE.
In general, the CPI tends to report somewhat higher inflation.
Why does PCE end up being lower than CPI?
We need to do a deeper dive to explain that phenomenon.
III. Categorizing the Differences Between PCE and CPI
The differences between the CPI and PCE measures of inflation can be summarized into four categories or effects.
(1) Scope effects
This conceptual difference means some items and expenditures in the PCE index are outside the scope of the CPI.
For example, medical care services in the PCE index include services purchased out-of-pocket by consumers, and those services paid for, on behalf of consumers. This then includes medical care services paid for by employers through employer-provided health insurance, as well as medical care services paid for by governments through programs such as Medicare and Medicaid.
In contrast, the expenditure weights for medical care services in the CPI are derived only from out-of-pocket expenses paid for by consumers.
These differences can also be isolated and measured, and can be referred to as “scope effects.”
(2) Weight effects
Relative weights assigned to each of the CPI and PCE categories of items are based on different data sources.
In order to estimate a “weight effect,” or the effect of using different weight sources on CPI and PCE index changes, CPI relative weights for comparable item categories can be used to estimate the PCE fixed-weight price index.
(3) Formula effects
The CPI and the PCE index are constructed from different index-number formulas.
A Fisher-Ideal index is considered a “superlative” index. It reflects consumer substitution among detailed items as relative prices change.
In practice, superlative indexes are difficult to implement in real time because such indexes require expenditure data for the period that is current, and such data are not available. For example, data on household consumer expenditures that are used to estimate the CPI are not available for the current period.
For the Consumer Price Index for Urban Consumers (CPI-U), a Laspeyres index provides an alternative to the Fisher-Ideal index.
To estimate a “formula effect,” or the differences in the rates of growth between the CPI and PCE caused by the differences in formula, the detailed price and quantity data used to estimate the PCE index can be reaggregated with the use of a Laspeyres price-index formula.
(4) Other effects
A variety of remaining differences consisting of seasonal-adjustment differences, price differences, and residual differences.
These must be taken into account for a complete understanding of the differences between the CPI and the PCE index.
IV. The Fed’s Thinking
January 2012 was when the Fed announced it would use the PCE as its primary measure of inflation, preferring it for three primary reasons:
(1) Scope is better: PCE includes more comprehensive coverage of goods and services.
(2) Weighting is better: Expenditure weights in the PCE can change as people substitute away from some goods and services toward others.
(3) Data revisions are better: PCE data can be revised more extensively than the CPI, which can only be adjusted for seasonal factors, and only for the previous five years.
Thus, if the price of bread goes up, people buy less bread, and the PCE uses a new basket of goods that accounts for people buying less bread. The CPI, however, is less fluid in response to changing consumer preferences.
In summary, the CPI represents a basket of goods and services that a consumer would buy without making substitution changes when prices change.
The PCE encompasses a broader range of goods and services than the CPI, from a broader range of buyers. It tries to track what is actually purchased, and represents how consumers change their buying patterns when relative prices change.
This leads to smoother price changes in the PCE and typically lower levels of reported inflation, at least as experienced by consumers.
V. Conclusion: Other Fed Computations of Consumer Price Inflation
While the PCE is considered better than the CPI for Fed purposes, that has not stopped the Federal Reserve from computing still more consumer price indexes.
Next, I show you three others.
(1) The Trimmed Mean PCE Inflation Rate produced by the Federal Reserve Bank of Dallas is an alternative measure of core inflation in the price index for personal consumption expenditures (PCE).
This data series is calculated by the Dallas Federal Reserve Bank, using data from the U.S. Bureau of Economic Analysis (BEA).
Image Source: Zacks Investment Research
Calculating the trimmed mean PCE inflation rate for a given month involves looking at the price changes for each of the individual components of personal consumption expenditures.
The individual price changes are sorted in ascending order from “fell the most” to “rose the most,” and a certain fraction of the most extreme observations at both ends of the spectrum are thrown out or trimmed. The inflation rate is then calculated as a weighted average of the remaining components.
The trimmed mean inflation rate is a proxy for true core PCE inflation rate. The resulting inflation measure has been shown to outperform the more conventional “excluding food and energy” measure as a gauge of core inflation.
(2) Median Consumer Price Index (CPI) is a measure of core inflation calculated the Federal Reserve Bank of Cleveland and the Ohio State University.
Median CPI was created as a different way to get a 'Core CPI' measure, or a better measure of underlying inflation trends.
Image Source: Zacks Investment Research
To calculate the Median CPI, the Cleveland Fed analyzes the median price change of the goods and services published by the BLS.
The median price change is the price change that’s right in the middle of the long list of all of the price changes. This series excludes 49.5% of the CPI components with the highest and lowest one-month price changes from each tail of the price-change distribution resulting in a Median CPI Inflation Estimate.
According to research from the Cleveland Fed, the Median CPI provides a better signal of the inflation trend than either the all-items CPI or the CPI excluding food and energy. According to newer research done at the Cleveland Fed, the Median CPI is even better at PCE inflation in the near and longer term than the core PCE.
(3) Sticky Price Consumer Price Index (CPI) is calculated from a subset of goods and services included in the CPI that change price relatively infrequently.
Because these goods and services change price relatively infrequently, they are thought to incorporate expectations about future inflation to a greater degree than prices that change on a more frequent basis.
One possible explanation for sticky prices could be the costs firms incur when changing price.
Image Source: Zacks Investment Research