Base effect
A change in the 12-month rate caused by an unusual price level a year earlier — not by anything happening to prices right now.
In plain words
The 12-month inflation rate compares prices now to prices exactly a year ago. If prices were unusually high or low a year ago, this month's rate will look artificially low or high — even if nothing has changed recently.
That distortion is the base effect. It's not a new price trend — it's a ghost of last year's unusual prices haunting this year's comparison.
The classic example: global energy prices spiked in mid-2022. A year later in mid-2023, even with energy prices flat, the 12-month rate dropped sharply — simply because it was now comparing stable prices to last year's spike. Disinflation looked dramatic. Much of it was a base effect.
See it in the data
The 2022 energy spike created a textbook base effect. When that spike dropped out of the 12-month window in 2023, the rate fell mechanically — even before any real disinflation.
Build your own base effect
Set a price spike in Year 1, then watch how it distorts the 12-month rate a year later — even with stable prices.
(prices actually unchanged)
Favourable vs unfavourable base effects
Base effects work in both directions. A low base from a year ago pushes the rate up; a high base pushes it down.
When last year's price level was unusually low (e.g. energy collapsed in 2020), this year's rate looks high even at normal prices. The base effect inflates the reading.
e.g. Energy CPI in early 2021 — comparing to pandemic lowsWhen last year's price level was unusually high (e.g. energy spike in 2022), this year's rate looks low even if prices haven't changed. The base effect suppresses the reading.
e.g. Energy CPI in mid-2023 — comparing to 2022 spikeHow long does it last?
π(t) = I(t) ÷ I(t−12) − 1 A spike in I(t−12) mechanically lowers π(t) — even if I(t) is unchanged I(t−12) ↑↑ (spike last year) Makes the denominator large → rate looks low: unfavourable base effect I(t−12) ↓↓ (collapse last year) Makes the denominator small → rate looks high: unfavourable base from below Common misreadings
"Inflation is falling — central bank policy is working."
Some of the fall may be a base effect — prices from a year ago were unusually high, mechanically reducing the rate. Analysts always separate base effects from genuine disinflation before crediting policy.
"A negative energy inflation rate means energy is getting cheaper."
A negative 12-month rate can simply mean energy prices this year are below the spike they hit a year ago — not that they're low in absolute terms. Prices may still be well above pre-spike levels.
"Base effects cancel out over time, so they don't matter."
They do resolve after 12 months — but they matter a great deal during those 12 months. Central banks, markets, and policymakers must distinguish between genuine disinflation and a mechanical base effect to avoid over- or under-reacting.
Frequently asked questions
What is a base effect in inflation?
A base effect is when the year-ago comparison point was unusually high or low, making this year's 12-month rate look artificially lower or higher than underlying price trends warrant. It's a measurement artefact, not a real change in inflation pressure.
How long does a base effect last?
Exactly 12 months. The 12-month rate compares to the same month a year ago. Once you pass the anniversary of the unusual month, the distortion disappears from the calculation. The unusual price level moves from the denominator (comparison point) to just being part of history.
How do analysts strip out base effects?
The main tools are: (1) looking at the annualized monthly or 3-month rate instead of the 12-month rate — these only reflect recent price changes; (2) examining the price level rather than the rate of change; (3) stripping out the volatile components (energy, food) to look at core inflation, which is less affected by commodity price swings.