Real news.A stronger Maryland.

MARYLAND HERALD


Maryland Edition

LATEST
Economy

How Inflation Is Measured: Inside the Consumer Price Index

Inflation in the United States is measured mainly by the Consumer Price Index, which tracks the prices of a representative basket of goods and services. Here is how it is built and what it can and cannot tell you.

Fresh produce in wooden market crate

Inflation in the United States is measured mainly by the Consumer Price Index, or CPI, which the Bureau of Labor Statistics (BLS) publishes every month. The CPI follows the prices of a representative basket of goods and services that urban households buy, and inflation is the percentage change in that index over time.

This guide explains what goes into the basket, how prices are collected, how the math works and why the national figure may not match what you feel at the checkout.

What does the Consumer Price Index actually measure?

According to the BLS, the CPI tracks the average change over time in the prices consumers pay for a basket of goods and services. It is meant to reflect inflation as households experience it in daily spending.

The basket covers more than 200 categories, grouped into eight major groups: food and beverages, housing, apparel, transportation, medical care, recreation, education and communication, and other goods and services. It includes user fees such as water and sewer charges and tolls, plus sales and excise taxes tied to purchases.

It leaves out income and Social Security taxes, and it leaves out investment items such as stocks, bonds, real estate and life insurance.

How does the Bureau of Labor Statistics collect the prices?

The BLS prices about 80,000 items each month, according to its published methodology. Sampling happens in stages: first geographic areas, then retail outlets, then specific items and, for rent, housing units. The outlet sample draws on Consumer Expenditure data about where households shop, and as of 2023 about 16 percent of price quotes, excluding rent, came from online outlets.

Data collectors gather prices by visiting stores, browsing websites, using apps and, in some cases, calling by phone. Food at home, energy and certain other items are priced every month. Most other items are priced monthly in New York, Los Angeles and Chicago and every other month elsewhere. Some items now come from outside datasets: the BLS says gasoline uses a secondary dataset, and new vehicles use J.D. Power transaction data.

Where do the weights come from?

Not every price matters equally. A 10 percent jump in the price of gasoline moves the index more than a 10 percent jump in the price of a niche product, because households spend far more on gasoline. The weights come from the Consumer Expenditure Surveys, which include interviews with more than 20,000 consumer units each quarter and about 12,000 consumer units keeping two-week diaries for frequently bought items, according to the BLS.

There is a built-in lag. The BLS notes that 2023 CPI data relied on 2021 survey data, a gap of about two years. Aggregation weights are updated each January.

Shelter is the biggest single piece

Housing dominates the basket. The BLS reports that, as of December 2025, shelter made up 35.625 percent of the CPI market basket. Within that:

Component (December 2025) Share of the basket
Shelter 35.625 percent
Owners’ equivalent rent of residences 26.204 percent
Rent of primary residence 7.840 percent
Lodging away from home 1.289 percent

Owned homes are not priced directly. The BLS treats owned housing as a capital good and instead measures what owners would pay to rent their own homes, unfurnished and without utilities. Owners are asked that question in the Consumer Expenditure Survey, while the CPI Housing Survey prices rental units.

How is the inflation rate calculated from the index?

The BLS explains that most CPI series use a base of 1982-84 equal to 100. An index reading of 110 means prices are 10 percent higher than the 1982-84 average. The inflation rate is the percent change between two index readings:

percent change = (later index ÷ earlier index − 1) × 100

A worked example

Suppose, purely for illustration, that the index stood at 300.0 a year ago and 309.0 today. Divide 309.0 by 300.0 to get 1.03, subtract 1, and multiply by 100. The 12-month inflation rate is 3.0 percent. The same formula works over any span: one month, three months or a year.

Rent shows how the BLS handles a slower-moving price. Each rental unit is priced every six months. The agency divides current average rent by the average from six months earlier, then takes the sixth root to get a one-month change. If a hypothetical rent ratio over six months were 1.05 (a 5 percent rise), the monthly change would be about 0.8 percent, since 1.05 raised to the power of one-sixth is about 1.0082.

What happens when products change or disappear?

Store shelves change constantly, so the BLS has procedures to keep a better or smaller product from being mistaken for a price change. When an item vanishes or changes in quality or quantity, a collector selects a replacement or records the change. Specialists at the BLS national office review the data and adjust for differences such as package size or features.

According to the BLS, when a replacement differs in quality, an analyst may compare the items directly, estimate the value of the difference (sometimes with statistical models of features) or impute the price change from similar items.

Since January 1999, most of the basic indexes have used a geometric mean formula, which allows for a modest amount of consumer substitution within an item category. For example, shoppers may switch between similar products when one gets pricier. The BLS also publishes a separate index, the Chained CPI for All Urban Consumers (C-CPI-U), that allows substitution across categories as well. It is revised after first release and finalized 10 to 12 months later, according to the BLS. The CPI-U and CPI-W are final when first published.

What is core inflation, and why do people watch it?

Core CPI is the index for all items less food and energy. The BLS says it is widely used as a gauge of underlying inflation because food and energy prices are relatively volatile. A hurricane, a refinery outage or a spike in crude oil can lift the headline number without saying much about broad price trends.

That is why analysts often read headline and core together, a point the Herald explored in its look at why gasoline was driving inflation while core measures deserved attention. For a recent reading of the monthly data in context, see the Herald’s report on the September CPI gasoline surge and falling real wages.

Which version of the CPI applies to whom?

The BLS publishes two main population versions:

  • CPI-U covers all urban households, more than 90 percent of the U.S. population. It excludes rural nonmetropolitan residents, farm households, military installations, religious communities and institutionalized people.
  • CPI-W covers about 30 percent of the population: households where wage earners in eligible occupations, such as clerical, sales, service, laborer and construction work, provide at least half of income. It excludes professionals and salaried workers, part-time workers, the self-employed, the unemployed and retirees.

The distinction matters because the CPI-W is what the Social Security cost-of-living adjustment is based on, using the third-quarter average. The BLS says more than 108 million people receive payments tied to the CPI by law, including more than 67 million Social Security beneficiaries and more than 41 million SNAP recipients. Tax brackets, some rents and some wage agreements are also adjusted using it.

Why doesn’t the CPI match a household’s own cost of living?

The BLS lists several common misconceptions.

  • It is not your personal inflation rate. The CPI reflects an average household. Someone who spends more than average on medical care, rent or gasoline may see a different rate.
  • It is not a full cost-of-living index. It does not capture factors such as safety, schools, the environment or income taxes.
  • It cannot compare cost levels between places. Area indexes show how prices changed over time within each area, not whether one area is more expensive than another. A Baltimore reading and a San Francisco reading cannot be used to say which city costs more.
  • Tips are generally excluded, though mandatory service charges and fees are included.
  • Tariffs are not counted directly, though they can affect prices indirectly.
  • Local indexes are bumpier. Smaller samples make them more volatile, so the BLS generally recommends national or regional indexes for escalation clauses.
  • Sampling error is not a calculation mistake. It reflects the limits of estimating from a sample, and the BLS publishes variance estimates each year.

How should readers use the figure?

When a new inflation number is released, a few checks help. Look at the 12-month change for the long view and the monthly change for the latest direction. Compare the headline rate with core. Check whether the reading is seasonally adjusted, since seasonally adjusted series can be revised for up to five years, according to the BLS.

For a household, the most useful comparison is often between inflation and income. Maryland workers earning the state minimum wage, for example, can look at how scheduled changes compare with price trends in the Herald’s guide to Maryland’s minimum wage. Central bank decisions, too, lean heavily on inflation readings, as covered in the Herald’s report on the Federal Reserve holding rates.

Reading the number with its limits in mind

The CPI is a carefully built average, not a verdict on any one household’s budget. It is based on spending surveys, tens of thousands of monthly price checks, explicit weights and published rules for handling product changes. Understanding those pieces makes it easier to see why the headline number, the core number and your own grocery and rent bills can tell different stories at the same time.

This article is general information about economic statistics and is not financial advice.

Sources: U.S. Bureau of Labor Statistics