The simple version

Two different measures come out of the same monthly report and only one of them existed this time. The change from the previous month was there as usual.

The comparison with a year earlier was not. No prices were taken in the month it would have been measured against, so there is no second figure to set it beside.

That is the whole mechanism. A change measured across twelve months needs two figures, and when funding stops for the agency that would record them, one of the two never gets made.

The numbers

• The October 2026 CPI report, release USDL-26-1685, was published by the Bureau of Labor Statistics on November 10, 2026 under an embargo until 8:30 a.m. Eastern.
• The report states that BLS did not collect survey data for October 2025 because of a lapse in appropriations, and that for most CPI series BLS could not publish October 2025 indexes.
• It follows that, in the agency's own words, BLS cannot publish 12-month percent changes for October 2026 for indexes where October 2025 data are unavailable.
• The tables show it rather than describe it. In the all items row, the October 2025 unadjusted index and the October 2025 to October 2026 percent change are both printed as a dash.
• A search of the full release text, all 101,551 characters of it, turned up no sentence giving a 12-month change. The phrase appears only as a table heading. That search is ours, not the agency's.
• What the report does give for all items, seasonally adjusted: a 0.4 percent rise in October, after 0.6 percent in September.
• All items less food and energy rose 0.3 percent in October, as it did in September and August.
• The unadjusted all items index reads 337.511 for October 2026 against 336.940 for September 2026, on the 1982-84 equals 100 base, an unadjusted monthly change of 0.2 percent.
• Category detail for October, seasonally adjusted: shelter up 0.3 percent, which the report says accounted for over thirty percent of the monthly all items increase; food up 0.5 percent; energy up 0.7 percent after a 5.2 percent increase in September.
• The report names used cars and trucks, and communication, among the small number of major indexes that decreased in October.
• BLS publishes a fact sheet on handling missing data, which the release points to directly.

Why a monthly change survives and a yearly one does not

The month-over-month figure compares this report with the one before it. Both of those readings exist, so the arithmetic has everything it needs.

The longer figure reaches twelve months back instead, and that is where it breaks down. One side of the comparison was never recorded, and an agency cannot divide by a reading it does not have.

There is a choice buried in that, and the agency made the conservative one. It could have estimated the absent month and published a comparison resting on that estimate. Instead it printed a dash and said why.

The actual math on what is missing

The unadjusted figures show what the arithmetic looks like when both sides exist. October 2026 reads 337.511 against September 2026 at 336.940, a difference of 0.571, which against the earlier reading comes to about 0.17 percent and prints as 0.2.

Now try the same sum across a year. The October 2026 reading is 337.511 and the October 2025 reading is a dash. There is nothing to take away and nothing to divide.

That is why the gap is not about rounding and not about precision. It is an absent input, and no amount of care with the figures that do exist will produce the one that does not.

What this means

If you went looking for the yearly rate and could not find it, nothing is wrong with your reading. For this base month it was not published, and a figure quoted as the yearly rate came from somewhere other than the report.

The gap also moves with the calendar. It affects the months whose year-earlier readings are missing, so it works its way out of the comparisons as those base months fall away.

What this is NOT

This article explains why a 12-month price change could not be published for a given base month, using the Bureau of Labor Statistics October 2026 CPI report, release USDL-26-1685, as its only source, read on November 10, 2026. It is not a forecast of any future reading, not a claim about what the missing yearly figure would have been, and not an argument that the published monthly figures are unreliable: the report states plainly which series are affected and prints a dash where a value cannot be computed. It takes no position on the 2025 lapse in appropriations, on who caused it, or on whether it should have happened, and it uses the agency's own term for the event throughout The numbers. Ours is the arithmetic working 337.511 against 336.940 to about 0.17 percent, and the search of the release text that found no published 12-month sentence; the 0.2 percent unadjusted figure is the report's own. One release was read and no other agency's data was examined. This is not a political endorsement or criticism of anyone.

Sources

• US Bureau of Labor Statistics, Consumer Price Index news release for October 2026, release USDL-26-1685, read November 10, 2026, for the missing-data statement, the dashes in the all items row, the seasonally adjusted monthly changes, the unadjusted index levels and the category detail.
• US Bureau of Labor Statistics, CPI fact sheet on approximating missing data, cited by the release itself as the reference for how absent months are handled: https://www.bls.gov/cpi/factsheets/approximating-missing-data.htm