Every month, Bureau of Labor Statistics staff record the shelf price of specific pairs of shoes in specific stores across the country, and the change in those prices becomes the three shoe indexes (men, women, children) that this site reads. No one at BLS averages the sticker prices of all shoes sold. The index follows the same items over time, and that single design choice explains most of what the numbers can and cannot tell you.
Where the prices come from
According to the BLS Handbook of Methods, the CPI survey collects about 100,000 prices a month for goods and services. Prices are gathered in 75 urban areas, which BLS groups into 32 index areas. About two-thirds of collection happens by personal visit to physical stores. The rest is done by phone or through the store’s website and apps, and BLS says about 8% of all CPI quotes now come from outlet websites.
Shoes are not priced everywhere every month. The handbook says food, energy and a few other items are priced monthly, and every item is priced monthly in New York, Los Angeles and Chicago. Elsewhere, the remaining goods and services are priced every other month, with each item assigned to either even or odd months.
How a particular pair gets chosen
The store list is not picked by a BLS employee who likes a certain mall. The handbook says the outlet sample for most items is built from the Consumer Expenditure Survey, a separate household survey of where people actually spend money.
Inside a store, the choice of the exact item is a chance process weighted by sales. The collector narrows the store’s stock in stages, picking one group at each step, and BLS states that the probability of selection at each stage is proportional to the sales of the items in each group. The collector can use the store’s sales figures, ranks, dollar volume or shelf space to set those odds. The result is a unique item, such as one style and material of men’s dress shoe, that the collector then prices again on later visits.
This is why the CPI can show a 2% rise in men’s shoes even when you saw no change in the pair you wear. Your pair is unlikely to be in the sample, and the index is an average across many sampled pairs in many places.
When the sampled pair disappears
A sampled pair does not stay on the shelf forever. The handbook describes two pieces of machinery for the moment it goes away.
Substitution. If an item is no longer sold, or its quality or quantity has changed, the collector picks the most similar item in the same store. A commodity analyst at BLS then reviews the new item and its price. When the two items differ, the new price may be quality adjusted so that only the pure price change enters the index. The CPI FAQ puts the goal plainly: commodity specialists try to keep changes in quality from affecting the measured price change.
Seasonal pricing. BLS classifies most apparel items as seasonal. For these, the handbook says the number of quotes is doubled, and each original quote is paired with a second quote in the same item group and store that is priced 6 months later, one designated fall/winter and one spring/summer. When a seasonal item is out of stock, its price is imputed (estimated from the price change of similar items), and when it returns its new price is compared with that imputed price. A permanent replacement is chosen only in months when a full selection of seasonal merchandise is on the shelves.
For a reader, the practical point is that the footwear index is built to measure price change at constant quality. It does not track what the average shopper spent on shoes. If people moved from leather dress shoes to cheaper sneakers, spending per pair could fall while the index stayed flat. Our method notes explain how the calculator turns this constant-quality index into a dollar figure.
What “1982-84=100” means
The men’s shoe index stood at 161.769 in September 2026. That number is not a price in dollars. The CPI FAQ says most CPI series use a 1982-84=100 reference base, and that an index of 110 means prices are 10% higher than in the reference period. By that rule, a comparable men’s pair was priced about 61.8% higher in September 2026 than in 1982 to 1984.
You can check the base in the data. When we average the 36 monthly values from January 1982 to December 1984 in our copy of the BLS file, men’s shoes come to 99.99 and women’s shoes to 100.00.
Two consequences follow:
- You can compare a series with itself over time. Men’s shoes at 76.675 in 1978 and 151.295 in 2025 means a rise of 97.3% for comparable pairs (both are annual averages from the BLS data).
- You cannot compare levels across series. Women’s shoes at 142.099 in September 2026 does not mean women’s shoes are cheaper than men’s at 161.769. Each index is 100 in its own base period, so the gap shows only that men’s shoe prices have risen more since 1982 to 1984. The CPI FAQ makes the same point about places, saying the CPI cannot measure differences in price levels between one area and another because it only measures change over time in each area.
The men’s shoe page and the women’s shoe page show each series on its own scale for this reason.
The month that does not exist, October 2025
The federal shutdown in the fall of 2025 left a hole in every footwear series. According to the BLS shutdown notice, no CPI prices were gathered between October 1, 2025, and November 12, 2025. The companion FAQ adds that BLS did not issue an October 2025 CPI news release. In BLS tables the missing month appears as a dash, with a footnote that the data are unavailable because of the 2025 lapse in appropriations.
To produce November 2025, BLS used carry-forward imputation for goods and services. Survey prices were carried from September 2025 into October, and the one-month change for November was measured against those imputed October prices. So the price change of shoes across the gap was assigned to November.
In our copy of the data, the men’s index went from 154.829 in September 2025 to 150.874 in November 2025, a drop of 2.6% over two months. Women’s shoes went from 136.700 to 133.322, down 2.5%. Because October is missing, there is no way to say how much of that drop happened in which month.
The gap has three effects you will see on this site:
- The 2025 annual average uses 11 months. BLS says that for series missing October, the 2025 average is the mean of the 11 published values. We do the same, which is why the 2025 price page is based on 11 months.
- No 12-month change for October 2026. The FAQ states that BLS will not publish 12-month percent changes for October 2026 for indexes with no October 2025 value. When the October 2026 numbers come out, the year-over-year change for shoes will be blank in BLS tables and on our pages.
- Seasonally adjusted series were patched. For seasonally adjusted data, BLS approximated the missing October value as the geometric mean of the September and November values. The unadjusted index, which our calculator uses, keeps the gap.
Reading the shoe index with this in mind
The footwear CPI is a careful measure of how much the price of a comparable pair has changed, built from sampled stores, sampled items, and rules for replacing items that vanish. It is a poor guide to the price tag on any one shoe and no guide at all to which category is cheaper in dollars. Used for what it measures, it lets you compare a pair bought in 1990 with one bought this year on equal terms, which is what the tool on the home page does.