Fundamentals

Consumer Price Index (CPI): What It Measures and How to Interpret It

By Daniel Sardá · Published on September 24, 2026

6 min read1,177 words

The CPI summarizes how prices change for a representative basket. Understanding its base period, weights, and limits helps prevent misleading readings.

The Consumer Price Index, known as the CPI, summarizes how the prices of goods and services consumed by a reference population change on average. Its practical purpose is to answer a concrete question: how much does a representative basket cost today compared with another period?

The word “representative” is essential. The CPI does not track the exact purchases of one household, nor is there a universal basket valid for every country. Each statistical system defines its coverage, collects prices, and assigns weights according to the spending patterns it seeks to represent.

Key idea: The CPI is a measuring instrument. Reading it well requires knowing which population, expenditures, and period it represents.

What the CPI measures

The CPI measures the average change in consumer prices within a defined scope. It may include food, housing, transport, health, education, communications, and other spending groups, although the specific selection and classification vary by national methodology.

By itself, it does not measure the causes of those changes. Nor does it show how much incomes have risen or whether everyone can maintain the same standard of living. Its role is narrower: to turn many price observations into an aggregate indicator that can be compared over time.

This definition is consistent with the approach in the ILO Consumer Price Index Manual, which addresses the coverage, weights, sampling, and adjustments needed to construct an index.

How it is built: basket, prices, and weights

Construction begins with a reference basket. It groups goods and services that reflect the consumption of the population under study. It should not be confused with a basic basket, which is usually intended to define an essential set of needs.

The statistical office then observes prices in selected establishments, locations, or sales channels. Those observations are organized into categories and subindices until they produce the overall index.

But not all prices have the same influence. Each category receives a weight, usually related to its share of spending by the households being represented. If housing takes up a larger share of spending than an item bought only occasionally, movements in housing prices will carry more weight in the aggregate result.

Imagine a hypothetical basket with two categories: food and transport. If food has a weight of 70% and transport 30%, an increase in food prices will affect the overall index more than an identical percentage increase in transport prices. In practice, baskets contain many more categories and use more complex statistical methods.

Information sources, update frequency, and exact formulas are not identical everywhere. Eurostat’s methodological manual, for example, documents one particular harmonized system. It helps explain the logic of weights, but its rules should not automatically be attributed to every national CPI.

Base period, level, and rate of change

The CPI level is commonly expressed in points relative to a reference period. That figure alone says little. If the index reads 126, it does not mean inflation is 126%: to know the change, it must be compared with another period.

The percentage change between an initial and a final value can be calculated as follows:

`(final CPI / initial CPI - 1) × 100`

Suppose the CPI rises from 120 to 126. The calculation is `(126 / 120 - 1) × 100`, which gives 5%. That means the aggregate price level measured by that index increased by 5% between the two periods.

The choice of periods changes the question. A monthly change compares with the preceding month; a year-over-year change compares with the same month of the preceding year. That is why a figure without its comparison interval is incomplete.

Useful distinction: Points show the level of the index; the percentage between two levels shows its change.

The CPI, inflation, and the cost of living are not the same

The CPI is an indicator. Inflation is a generalized rise in prices over time, and it can be analyzed through the CPI or other measures. A change in the CPI shows consumer-price inflation within the index’s own coverage, not a complete description of every price in the economy.

Nor is it exactly equivalent to the cost of living. Maintaining a given standard of living depends on more than the prices of a fixed or updated basket: the ability to substitute products, available public services, household conditions, and other changes in the environment also matter.

The distinction is also clear when compared with the Producer Price Index, or PPI. While the CPI observes consumer prices paid within its defined scope, the PPI follows prices at earlier stages of production. Both can provide relevant information, but they answer different questions.

What it is used for

The CPI makes it possible to track consumer-price developments and compare periods. It also provides a reference for studying the purchasing power of monetary amounts and for informing public and private decisions.

Depending on a country’s rules or the terms of a given agreement, its changes may be used as a reference for adjusting payments, wages, rents, pensions, or contracts. This use is neither automatic nor universal: it depends on the applicable legal and contractual framework.

It is also one input for economic analysis and public policy. But the indicator does not prescribe a response. Deciding what to do about inflation requires examining its causes, effects, and the costs of different alternatives.

Limits worth keeping in mind

The first limit is the average. Two households may face very different changes if their spending is concentrated in different categories. If a household’s personal expenses rise more than the CPI, that alone does not prove the index is wrong; it may show that its experience differs from the average pattern represented.

The second limit is methodological. Habits change, some products disappear, and others improve or deteriorate. Statistical offices must decide how to replace items, update weights, and separate a quality change from a pure price change. These are normal measurement problems, not automatic evidence of manipulation.

The third limit is coverage. An index may exclude certain households, territories, purchasing channels, or types of spending. To judge whether it is suitable for a particular question, readers should consult its methodological notes, not only the headline figure. The BLS explanation of the CPI shows, for its specific scope, why an average need not match each consumer’s experience.

Caution: A CPI figure becomes more meaningful when it is published together with its comparison period, coverage, and methodology.

A reliable measure needs reliable institutions

The CPI influences expectations, economic assessments, and, in some cases, monetary commitments. Its credibility therefore requires public rules, accessible documentation, technical continuity, and independence from political or private interests.

Transparency does not eliminate the difficulties of measuring a changing economy, but it makes it possible to discuss statistical decisions with evidence and compare results over time. A responsible reading of the CPI does the same: it considers the change, asks what the measure represents, and recognizes what an average cannot say about every household.