By Michael McCracken, Economist; Aaron Amburgey, Research Associate
Once a month, policymakers, economists and investors tune in to learn the latest value of consumer price index (CPI) inflation—commonly referred to as headline inflation. There are some prices—such as the price of gas and common grocery items—for which changes attract similar, albeit less attention. Besides these select measures, it is very infrequent that you hear about the most recent change in the specific price of something like “footwear” or “public transportation.” In this blog post, we take a closer look at the prices of a variety of goods and services, with a focus on series with low volatility, or “sticky prices.”
To determine which prices are sticky, we also need to determine which prices are flexible. To conduct our analysis, we took 45 individual prices series from the Bureau of Labor Statistics (BLS).We choose to use the same prices contained in the Federal Reserve Bank of Atlanta’s Sticky-Price and Flexible-Price CPI series. We then measured the volatility of each price series by taking the variance of its annualized three-month change from 1985 to 2019. Finally, we grouped the 45 series into four quartiles.
The table below reports these groupings. The series in the first quartile (Q1) have the lowest volatility, while the series in fourth quartile (Q4) have the highest. Furthermore, the series in each group are listed in ascending order by volatility, with “food away from home” representing the stickiest series and “motor fuel” representing the most flexible.
To extend our work, we also calculated the volatility quartiles for 2020-21 (i.e., the COVID-19 pandemic). Series that are highlighted in blue (red) were in a higher (lower) quartile for this alternate measure; items without highlights didn’t move out of their quartile during this period.
|Food away from home||OER, Northeast||Motor vehicle insurance||Men’s and boys’ apparel|
|Rent of primary residence||Personal care products||Motor vehicle fees||Public transportation|
|Motor vehicle maintenance and repair||Motor vehicle parts and equipment||Education||Infants’ and toddlers’ apparel|
|OER, South||Communication||Meats, poultry, fish and eggs||Fresh fruits and vegetables|
|Personal care services||Medical care commodities||Miscellaneous personal goods||Tobacco and smoking products|
|OER, Midwest||Tenants’ and household insurance||Processed fruits and vegetables||Car and truck rental|
|Recreation||Alcoholic beverages||Leased cars and trucks||Gas (piped) and electricity|
|Water, sewer and trash collection services||Cereals and bakery products||Nonalcoholic beverages||Lodging away from home|
|OER, West||Other food at home||Dairy and related products||Women’s and girls’ apparel|
|Medical care services||New vehicles||Used cars and trucks||Fuel oil and other fuels|
|Miscellaneous personal services||Household furnishings and operations||Footwear||Motor fuel|
|Jewelry and watches|
|NOTES: The series are grouped into quartiles based on volatility from 1985 to 2019. Within each group, the series are listed in ascending order by volatility. Series that are highlighted in blue (red) were in a higher (lower) quartile during the COVID-19 pandemic. OER stands for owners’ equivalent rent of primary residence.|
|SOURCES: Haver Analytics and authors’ calculations.|
Focusing first on the original groupings, Q1—the least volatile—includes various measures of rent; medical, utility and personal services; food away from home; recreation; and motor vehicle maintenance and repair. Meanwhile Q4—the most volatile—contains several clothing and energy series along with public transportation, fresh fruits and vegetables, and tobacco and smoking products.
Amid the COVID-19 pandemic, the volatility of several series changed, however. Notably, medical and personal services, and several grocery item series became more volatile while leased cars and trucks, tobacco, and alcohol became less volatile. The increased volatility in grocery items is unsurprising given the doomsday shopping patterns that took place during the early stages of the pandemic.
Sticky prices are unexciting in that they are less prone to sudden changes, but this quality is also what makes them useful. Because they are so stable, changes in these prices are more likely to reflect important inflation trends. To put it another way, sticky prices are forward looking. Researchers at the Federal Reserve Bank of Cleveland formed a measure of sticky CPI and showed that it could improve forecasts of headline CPI.
We form a measure of sticky CPI by taking the unweighted average of the year-over-year changes in the Q1 series. The figure below plots this measure of sticky CPI (i.e., Q1 CPI) against headline inflation.
NOTES: Sticky CPI is measured by taking the unweighted average of the year-over-year changes among CPI items in the lowest volatility quartile. The gray shading represents recessions.
SOURCE: Haver Analytics and authors’ calculations.
While Q1 inflation tends to follow similar trends to headline inflation, there are some notable diversions. For example, in April 1986 Q1 inflation reached a sample high of 5.8% while headline inflation was only 1.6%. Low levels of headline inflation during this period were a result of the 1980s oil glut, which saw oil prices halved in 1986. This drop in oil prices was not related to any broader price trend, hence why Q1 inflation remained high given other inflationary pressures during this period.
More recently, Q1 inflation has been trending down, decreasing by 0.8 percentage points from 3.4% in July 2020 to 2.6% in June 2021. In direct opposition, headline inflation faced a sharp 4.4 percentage point increase from 1% to 5.4% during the same period. This stark difference is in part due to base effects related to dramatically low inflation levels at the start of the pandemic for which the Q1 series were mostly unaffected. Nonetheless, the much more stable Q1 series may indicate that recent swings in headline inflation are only temporary.