Inflation base effects: why the annual rate can fall while prices rise

1. The old month matters too

A headline says inflation has cooled. Before interpreting that as a change in current conditions, ask what happened in the month being used as the comparison. Year-over-year inflation is calculated as the current price index divided by the index twelve months earlier, minus one, expressed as a percentage. Both observations matter.

Statistics Canada calls the influence of price movements twelve months earlier a base-year effect. When a large earlier increase drops out of the twelve-month comparison, it puts downward pressure on the annual rate. An earlier decrease dropping out can have the opposite effect. Neither guarantees the final result: new price movements also matter.

Think of the annual window as moving forward one month at a time. You gain a new month and lose an old one. The headline tells you the combined result, without isolating how much of the change came from each end.

Further reading: Statistics Canada: how base-year effects influence CPI

2. Follow four index observations

You can see the mechanism without forecasting an economy. Use the same hypothetical index for every observation, with no change in its definition. The key is to compare each current month with its matching month one year earlier, then separately calculate the latest monthly change.

In this example, the annual rate falls by about 0.84 percentage points while the index rises between the two current months. The earlier jump from 100 to 102 leaves the annual window. Calling this 'prices fell' would confuse the change in a growth rate with the change in the index itself.

To reproduce the calculation, keep the unrounded values until the final step. Index points and percentage changes are different units; subtracting two index readings alone does not calculate inflation.

3. Read the monthly figure with its adjustment label

The monthly change helps describe the newest movement, but first check whether it is seasonally adjusted. BLS explains that seasonal adjustment removes recurring influences, such as weather and holiday patterns, to make short-term changes easier to interpret. Adjusted and unadjusted series answer different questions.

Do not calculate a monthly change from one adjusted observation and one unadjusted observation. Copy the series name and adjustment label with the numbers. If you are reproducing a published calculation, use its stated series and period throughout.

One month also makes a fragile trend. Look at several comparable monthly readings before writing that pressure is consistently easing or strengthening. A lower annual rate and a stronger latest month can coexist; report both observations instead of forcing them into one slogan.

Further reading: BLS: CPI comparisons and seasonal adjustment

4. Check which prices changed

A total can conceal different movements among its components. Read the release's category discussion alongside the overall rate. For US CPI, BLS also publishes an index excluding food and energy, often described as core because these categories can be volatile.

Keep that exact definition attached to the number. Do not substitute a core rate for an all-items rate midway through a comparison. For your research note, choose one component mentioned in the release and record whether its annual and monthly movements support the same interpretation. Leave the explanation open when they do not.

Further reading: BLS: all-items and component price indexes

5. Separate the statistic from your household bill

A national inflation measure describes a weighted basket, not your personal shopping list. The ECB explains that items carry different weights according to household spending and that individual households buy different things. Someone facing a large increase in a major expense may experience a different change from the aggregate.

The comparison horizon matters as well. This year's annual rate does not erase increases accumulated over earlier years. A positive rate means the measured basket remains above its year-earlier level, even if that rate is lower than before. When discussing affordability, say whether you mean the latest monthly change, a twelve-month comparison or the price level since a much earlier date.

Further reading: ECB: inflation, household baskets and comparison periods

6. Turn the headline into a checkable note

Start with the original statistical release. Write down the reference month, exact index, annual rate and monthly rate with its adjustment label. Then inspect the matching period a year earlier and any base-effect explanation supplied by the agency. Keep the release link so you can revisit the evidence.

A useful conclusion is specific: 'The annual rate decreased, while the latest monthly index increased; the release identifies an earlier price jump leaving the comparison.' Use that wording only when those observations are verified. A base effect is a mathematical explanation to investigate, not a catch-all reason for every change.

Use SynthSignal's official-release guide to find the underlying documents and its news-check worksheet to keep your evidence together. Record any stock or crypto reaction separately. The inflation calculation alone cannot establish why an asset moved or what it will do next.