How to Read CPI: Timeframes, Shelter, and Real-Wage Checks

To read the monthly CPI report, focus on percentage changes rather than the raw index numbers, know which version of the index you’re looking at, and pair the seasonally adjusted month-over-month figure with the unadjusted year-over-year figure to see both the trend and the direction. Everything else in the release is detail that supports those two moves. Once you know how to read the Consumer Price Index report this way, the wall of tables turns into a practical tool for tracking inflation, evaluating a raise, or checking a contract escalation.

What the Index Numbers Mean

Every CPI figure is anchored to a base period. The Bureau of Labor Statistics set the average price level across 1982, 1983, and 1984 equal to 100. So an “All Items” index value of 315.6 means the overall consumer basket costs about 215.6 percent more than it did in the early 1980s.¹

The raw number by itself is not what you want. What matters is the change between two index values, and that change has to be expressed as a percentage, not as a point difference. If one item rises from 112.5 to 121.5 (a 9-point jump) and another rises from 225.0 to 243.0 (an 18-point jump), both increased by exactly 8 percent. Points depend on where the index started; percentages do not. When you look at the report, the percentages are the story.

Which CPI You’re Actually Reading

The monthly release produces thousands of individual indexes, but three measures dominate the discussion.

Headline CPI (All Items) covers every category of consumer spending, from groceries to gasoline to medical care. Because it includes everything, it swings hard on commodity prices and weather events.²

Core CPI (All Items Less Food and Energy) strips out food and energy because those categories are volatile and often driven by supply shocks rather than broad price trends. The BLS has not “removed” food and energy from the official inflation measure; it publishes both series so analysts can see underlying momentum more clearly.³

Supercore (Core Services Less Housing) is the label the financial press uses for core services stripped of shelter. Fed Chair Jerome Powell has highlighted this category as important for reading where inflation is headed, because housing costs move slowly in the CPI and can mask what’s happening in insurance, transportation services, and medical care.⁴

Reading all three together tells you what kind of price change is driving the report. When headline and core diverge sharply, food or energy is usually the reason. When core and supercore diverge, housing is the swing factor.

Which Population the Index Covers

The BLS publishes CPI data for different population groups, and each one drives different policies.

CPI-U (All Urban Consumers) is the broadest measure, covering about 88 percent of the U.S. population, including professionals, the self-employed, the unemployed, and retirees. This is the version most commonly reported in the news.⁵

CPI-W (Urban Wage Earners and Clerical Workers) is narrower, covering about 28 percent of the population, specifically households where more than half of income comes from clerical or wage jobs. By law, this is the index that sets Social Security Cost-of-Living Adjustments.⁶

C-CPI-U (Chained CPI) accounts for consumers substituting cheaper alternatives when prices rise. If beef gets expensive and buyers switch to chicken, the standard CPI-U doesn’t fully capture that shift, but the Chained CPI does. Since the 2017 Tax Cuts and Jobs Act, the IRS uses the Chained CPI to adjust federal income tax brackets.⁷⁸

A boundary worth knowing: none of these indexes covers people living in rural areas, active military personnel, or those in institutions such as prisons.⁵ If you live on a farm or in a small rural town, the CPI may not reflect your cost experience well.

Where the Report Lives and What to Open First

The BLS publishes the CPI Summary on a set schedule, typically between the 10th and 14th of each month, always at 8:30 a.m. Eastern Time. The full release calendar is posted on the BLS site.⁹ Going to the source protects you from financial outlets that sometimes round numbers, mix up seasonally adjusted and unadjusted figures, or lead with whichever version makes a better headline.

The most useful piece of the release is Table 1, which shows the CPI-U by expenditure category for the U.S. city average.¹⁰ It has four column groups worth learning:

Relative Importance — the percentage weight each category carries in the overall index.
Unadjusted Indexes — the raw index values for the current and prior month.
Unadjusted Percent Change — year-over-year and month-over-month changes before seasonal adjustment. The year-over-year column is where the annual inflation rate for each category lives.
Seasonally Adjusted Percent Change — month-over-month changes after seasonal patterns are removed. This is what economists watch for short-term trends.

Reading the Two Timeframes Together

Two timeframes carry most of the meaning in the report, and they answer different questions.

The month-over-month change shows how much prices shifted from the prior month to the current one. A figure like +0.3 percent means the overall price level rose by three-tenths of a percent in a single month. A negative number means prices fell. Monthly changes are small, so they’re reported to one decimal place, and even a tenth of a point matters when annualized.

The year-over-year change compares the current month to the same month one year earlier. This is the headline you hear quoted — “inflation was 2.4 percent.” It smooths monthly noise and reflects what households have actually experienced over the past year.

The insight comes from reading them together. A high annual rate paired with low or declining monthly changes means inflation is cooling: prices are still above last year, but they’re no longer climbing as fast. A low annual rate paired with accelerating monthly changes is an early warning. Most people stop at the annual headline and miss the direction of the monthly trend.

When to Use Seasonally Adjusted vs. Unadjusted

Some price changes are predictable. Gasoline tends to spike in summer driving season. Clothing prices drop after the holidays. Heating costs rise in winter. These patterns repeat year after year and say nothing about underlying inflation.

Seasonally adjusted data removes those predictable swings so you can see whether prices are actually accelerating or just following the calendar.¹¹ For month-to-month changes, always use the seasonally adjusted column. Without it, a January drop in clothing prices might look deflationary when it’s really just post-holiday clearance.

For year-over-year comparisons, seasonal adjustment matters less because you’re comparing the same calendar month to itself, and both periods carry the same seasonal effects. That’s why Table 1 shows year-over-year changes using unadjusted data but month-over-month changes using seasonally adjusted figures.¹⁰

Why Shelter Tells You So Much

The CPI doesn’t treat every purchase equally. Each spending category is weighted by how much of a typical household’s budget it represents. As of December 2025, the major CPI-U weights break down roughly as follows:¹²

Shelter alone is more than a third of the entire index. A 5 percent jump in shelter costs moves the headline number far more than a 5 percent jump in clothing prices. When you’re trying to explain why the overall CPI moved the way it did, check shelter first.

One quirk worth carrying with you: the BLS does not use home purchase prices or mortgage interest rates in the CPI. For homeowners, it uses Owners’ Equivalent Rent, and the price change is calculated from rents collected on renter-occupied units.¹⁴ Because of the way that measurement works, housing inflation in the CPI lags the real estate market by several months. When market rents spike, the CPI will not fully reflect it for a while, and when rents cool, the CPI will keep showing elevated shelter inflation longer than you’d expect.

Regional and Metro Numbers

The national CPI gets the attention, but the BLS also publishes data for four census regions (Northeast, Midwest, South, and West) and for selected metropolitan areas.¹⁵ If you’re in Miami, the national rate may not match your experience at all, because housing costs, insurance rates, and food prices vary sharply by location.

The catch: regional and metro data is published less frequently than the national release and covers fewer categories. Smaller metro areas are reported bimonthly or semiannually. For a timely local salary or lease question, the national figure may be your only current option, but the regional breakdowns add useful context.

Turning the Report Into a Real-Wage Check

One of the most practical uses of the CPI is checking whether your income is keeping up with inflation. Divide your nominal wage by the CPI index value for that period, then multiply by 100. The result is your “real” wage in base-period dollars.

Say you earned $25 an hour in January 2025 when the CPI was 309.7, and you earn $26 an hour in January 2026 when the CPI is 317.1. Your nominal raise is 4 percent. But your real wage went from ($25 / 309.7) × 100 = $8.07 to ($26 / 317.1) × 100 = $8.20. In base-period dollars, your purchasing power rose about 1.6 percent, less than half of what the nominal raise suggested. If your raise had been only 2.4 percent, your real wage would have been essentially flat.

This is the single most useful calculation for anyone evaluating a raise, comparing job offers across years, or checking whether savings are actually growing.

How the Numbers Feed Federal Programs

Reading the report also tells you what’s about to change in federal programs and government-backed investments.

Social Security COLA

Social Security benefits are adjusted annually using the CPI-W, not the broader CPI-U. The formula compares the average CPI-W for the third quarter (July through September) of the current year to the third quarter of the previous year. If prices rose, benefits go up by that percentage the following December.⁶ The 2025 COLA was 2.8 percent.¹⁶ In years when the CPI-W did not rise over that window, beneficiaries received no increase.¹⁷

Federal Tax Brackets

The IRS adjusts income tax brackets, the standard deduction, and many credit thresholds each year using the Chained CPI. Because the Chained CPI accounts for substitution to cheaper goods, it generally grows more slowly than the standard CPI-U.⁷ Tax brackets inch up a bit less each year than they would under the older method, and over a decade or more that slower adjustment pushes more income into higher brackets.⁸

TIPS and I Bonds

Two federal investment products track the CPI-U directly. Treasury Inflation-Protected Securities adjust their principal value based on changes in the non-seasonally adjusted CPI-U. When inflation rises, principal grows; when deflation occurs, principal shrinks.¹⁸ Series I Savings Bonds work similarly: their inflation-rate component resets every May and November using the non-seasonally adjusted CPI-U for all items.¹⁹ If you hold either, the CPI report tells you directly whether your inflation protection is rising or falling.

CPI in Private Contracts

Thousands of private contracts also use the CPI as an escalation mechanism. Commercial leases often raise rent annually by the CPI. Alimony agreements, long-term service contracts, and union wage deals frequently work the same way.²⁰

If you’re a party to one of these agreements, the details in the clause matter: which CPI version (CPI-U or CPI-W), which geographic area (national or a specific metro), seasonally adjusted or not, and which base month triggers the adjustment. A contract pegged to national CPI-U All Items produces a different number than one pegged to a regional CPI-W for All Items Less Food and Energy. Reading the escalation clause alongside the BLS release is the only way to verify the adjustment is calculated correctly.

1
U.S. Bureau of Labor Statistics. Consumer Price Index Frequently Asked Questions
2
U.S. Bureau of Labor Statistics. Consumer Price Index Home
3
U.S. Bureau of Labor Statistics. Common Misconceptions about the Consumer Price Index – Section: Has the BLS Removed Food or Energy Prices in Its Official Measure of Inflation
4
Federal Reserve Bank of St. Louis. Measuring Inflation: Headline, Core and Supercore Services
5
U.S. Bureau of Labor Statistics. Why Does BLS Provide Both the CPI-W and CPI-U
6
Social Security Administration. Cost-of-Living Adjustment (COLA) Information
7
U.S. Bureau of Labor Statistics. Frequently Asked Questions about the Chained Consumer Price Index for All Urban Consumers (C-CPI-U)
8
Office of the Law Revision Counsel. 26 U.S. Code 1 – Tax Imposed
9
U.S. Bureau of Labor Statistics. Schedule of Releases for the Consumer Price Index
10
U.S. Bureau of Labor Statistics. Consumer Price Index Economic News Release – Table 1
11
U.S. Bureau of Labor Statistics. Seasonal Adjustment in the CPI
12
U.S. Bureau of Labor Statistics. Relative Importance of Components in the Consumer Price Indexes
13
U.S. Bureau of Labor Statistics. Measuring Price Change in the CPI: Rent and Rental Equivalence – Section: Relative Importance
14
U.S. Bureau of Labor Statistics. Measuring Price Change in the CPI: Rent and Rental Equivalence
15
U.S. Bureau of Labor Statistics. 12-Month Percentage Change, Consumer Price Index, by Region and Division, All Items
16
Social Security Administration. Latest Cost-of-Living Adjustment
17
Social Security Administration. Cost-Of-Living Adjustments
18
TreasuryDirect. TIPS/CPI Data
19
TreasuryDirect. I Bonds Interest Rates
20
U.S. Bureau of Labor Statistics. Writing an Escalation Contract Using the Consumer Price Index