Every month, news reports feature two different percentages drawn from the exact same government data. One report notes that consumer prices ticked up by a fraction of a percent from the previous month. Another headline announces that inflation is running at several percentage points over the previous year. Both numbers come from the same price records, but they answer different questions about how prices are behaving.

Close-up of a ruled ledger page with blurred columns of figures under warm window light, a fountain pen resting across the paper, representing two inflation percentages drawn from one price index.
Monthly Consumer Price Index summary table showing index levels across successive reporting periods.

The underlying foundation for both numbers is the Consumer Price Index (CPI), compiled by the Bureau of Labor Statistics (BLS). To build the index, field agents sample thousands of prices each month across food, energy, housing, medical care, and transportation. The agency calculates the total cost of this standard market basket and sets it against a historical reference point. For the headline index covering all urban consumers, known as the CPI-U, the base period is 1982 to 1984, which is fixed at an average value of 100.

Because the index is a single continuous tally, an index level of 324.8 simply means that the basket costs roughly 224.8 percent more than it did during that mid-1980s base period. The difference between month-over-month (MoM) and year-over-year (YoY) metrics comes down to which row on the agency ledger you choose as your starting point. You are not looking at two separate surveys; you are looking at two different time intervals applied to the same collected prices.

Where both measures come from

The Bureau of Labor Statistics publishes the CPI-U for the U.S. city average on a monthly schedule. Economists at the Cleveland Fed's Center for Inflation Research describe this basket as a way to capture the spending patterns of typical city households. The index value itself is not a dollar total, nor is it a percentage. It is an index number that tracks the relative drift of retail costs across the entire country.

When the BLS updates the index each month, technicians log the new index value alongside the historical ledger. Whether an analyst wants to understand short-term fuel spikes or multi-year grocery bills, the starting ingredient is identical. According to educational materials from the Iowa State University Department of Economics, the entire distinction between monthly and annual rates lies in the choice of the earlier reference month. The index itself does not alter its construction based on the reporting window.

How each percentage is calculated

The math behind both measures follows standard percentage change arithmetic. As detailed in the BLS guidelines on calculating percent changes, you subtract the earlier index value from the newer index value, divide that difference by the earlier value, and multiply the result by 100. This simple formula determines the rate of change across any two points in time.

For a month-over-month calculation, the comparison window is thirty days. An example from university coursework shows that if the index registered at 242.839 in January 2017 and moved to 243.603 in February 2017, the difference is 0.764 points. Dividing 0.764 by 242.839 yields a month-over-month increase of approximately 0.31 percent.

For a year-over-year calculation, the formula uses the index value from the identical month twelve months earlier. For instance, data from an online CPI calculator shows that an index reading of 324.8 in September 2025 followed by a reading of 336.94 in September 2026 represents a change of 12.14 points. Dividing 12.14 by 324.8 results in an annual inflation rate of roughly 3.74 percent. Running this arithmetic by hand is often the clearest way to confirm any published report.

Why monthly and yearly figures can point in opposite directions

Because the time horizons differ, a monthly metric and an annual metric can seem to disagree. A month-over-month figure shows immediate, short-run price movements. These short snapshots can be noisy. A brief jump in egg prices or an unexpected drop in gasoline costs can skew the monthly change without altering the long-term trend.

In contrast, the year-over-year figure captures the cumulative change across all twelve months. It reflects every price shift that happened throughout the previous year. As the Federal Reserve notes in its public explainers, policymakers track broad rates to understand underlying trends rather than reacting to single-month blips.

This mechanical difference means that a falling monthly rate does not automatically mean that annual inflation is dropping. If the index drops by 0.1 percent this month, but prices rose sharply during the very same month a year ago, the twelve-month rate may remain elevated or even tick higher. Analysts sometimes call this dynamic the base effect, as explored in the BLS economics blog. Assuming that a quiet monthly print guarantees relief on an annual basis can lead to a misunderstanding of the ledger.

Checking figures and verifying source data

When reviewing inflation claims, it helps to confirm which specific series is being quoted. The BLS compiles several indexes to serve different analytical functions:

  • CPI-U: Covers all urban consumers, representing roughly 93 percent of the total U.S. population.
  • CPI-W: Focuses specifically on urban wage earners and clerical workers, often used for cost-of-living adjustments in certain benefit programs.
  • Chained CPI: Accounts for consumer substitution when the prices of specific goods rise faster than alternatives.

The standard benchmark for the national headline rate carries the official series code CUUR0000SA0, which corresponds to the unadjusted U.S. city average for all items. Online calculators can speed up the math, but any tool that fails to disclose its series code or raw index values makes verification difficult. Checking the official BLS technical notes allows anyone to inspect the base tables directly.

What the published indexes do not reveal

Index tables provide a macro summary, but they do not mirror the balance sheet of every individual home. A household that spends a high proportion of its earnings on rent, heating fuel, and child care may experience cost increases that run far ahead of the headline CPI-U. Conversely, a household that owns its home free of debt and commutes short distances may see lower overall budget pressure.

Government index reports also do not give personal guidance on handling household debt, wage negotiations, or loan rates. Regulatory agencies such as the Consumer Financial Protection Bureau (CFPB) and the Federal Trade Commission (FTC) provide educational materials on consumer rights and household budget protections. For the raw numbers, the official BLS release calendar remains the standard source for seeing when the next batch of monthly index levels will be made public.