An inflation report lands, the headline number moves by a tenth of a point, and within minutes there is a chorus of commentary. Some of it is careful. Much of it is people reading one figure as though it were a verdict. Price statistics are genuinely useful, but they are also among the easiest economic releases to misread, because the number that gets quoted is a summary of a summary: an average of thousands of prices, compared with another average, adjusted in ways the headline rarely explains. The good news is that most misreadings come from a handful of recurring mistakes, and each one is fixable with a few minutes of attention.
Headline and core answer different questions
The headline figure includes everything in the basket. Core inflation typically strips out food and energy, or whichever volatile components the statistical agency designates, on the reasoning that these prices swing for reasons that say little about underlying pressure. The exact definition varies by country and by agency, so it is worth checking rather than assuming.
The classic error is crowning one of them as the real number. “Core is what actually matters,” say some. “Headline is what people actually pay,” say others. Both figures have a job. Headline is closer to lived experience: it is what shows up in the supermarket and at the fuel pump, and it feeds into wage negotiations and indexation. Core is the better guide to persistence, because it is not dominated by a single commodity moving in one direction.
During an energy shock the two diverge in a predictable way. Headline jumps first; core follows slowly, if at all, as transport and input costs feed through. Read both, then ask a causal question: what is moving, and is it the kind of thing that reverses or the kind that sticks?
Base effects: the ghost of last year
Year-on-year comparisons are arithmetic against a fixed point. If prices jumped in a particular month last year, that month becomes a high base, and this year's annual rate will look lower even if the monthly pace has not changed at all. The reverse is equally true: a weak month last year flatters this year's figure.
Picture an index where prices rose sharply in one month last year after a tax change, and this year prices rise only slightly in the same month. The annual rate falls noticeably. Nothing about price pressure has improved; the comparison point moved. Politicians and commentators discover this effect roughly once a year, usually by accident.
The practical fix is to look at the monthly path from twelve to eighteen months back, which most releases publish, or to use shorter annualised windows such as the three-month and six-month rates. A falling annual rate alongside a steady monthly pace is not a contradiction. It is a base effect.
A monthly move is not an annual rate
Multiply a monthly change by twelve and you get a rough gauge of the current pace, not a forecast. A 0.5% monthly rise is often quoted as “six percent inflation”, which is a reasonable mental shortcut and a poor prediction, because few monthly rates repeat twelve times in a row.
Single months are also noisy. Sales timing, airfares, clothing seasonality, a change in value-added tax, a strike, a harvest, a new tariff: any of these can distort one print. Seasonal adjustment removes the predictable part of that noise, not all of it.
- Average the last three to six monthly changes rather than reacting to one.
- Compare that average with the monthly pace consistent with the target — for a 2% target, roughly 0.15% a month, compounded.
- Check whether the monthly figure you are reading is seasonally adjusted while the headline annual figure is not. Mixing the two is a common and entirely avoidable error.
Inflation falling is not the same as prices falling
Two words get conflated constantly. Disinflation means prices are rising more slowly. Deflation means they are falling. A country can report months of disinflation while every price on the shelf continues to climb.
The second half of this mistake is about levels. Indexes rarely travel back down. If the price level rises substantially over a few years and then inflation settles at zero, prices stay where they are — permanently higher than before. This is why the argument “inflation is down, so why are people still unhappy?” misses the point. Households budget against the level, not the rate, and the level is what changed.
When you read a report, track the index as well as the percentage change. The rate tells you how fast the ground is moving. The level tells you where you are standing.
Your basket is not the average basket
A consumer price index is a weighted average across all households in a country. Yours is different. If you rent rather than own, have young children, commute by car, or live in a region with different energy costs, the official average may describe you poorly. Weights also matter within the index: a large-weight item such as housing moving slightly can matter far more than a small item moving dramatically.
You can approximate your own inflation without much effort. Write down the main shares of your monthly spending — housing, food, transport, energy, health, leisure — and check how each component in the release moved over the past year. The result will not be statistically rigorous. It will still explain why the official number sometimes feels wrong, in either direction.
Read the release, not the reaction
Most of the information sits below the headline, and a little methodology goes a long way.
- Flash versus final. Many agencies publish a preliminary estimate that is later revised. Revisions are usually small and occasionally not.
- Reference period. A report published this month often describes last month. Check the date on the data, not the date on the news story.
- Which index. National CPI, harmonised indices and consumption deflators differ in coverage, and central banks may target a different measure from the one making headlines.
- What the release says about causes. Statistical offices usually list the components that contributed most. That paragraph is often the most useful part of the document.
A five-minute routine before you react
Next time a number lands, run this in order. First, note both headline and core, and check what “core” excludes in that country. Second, look at the three-month and six-month annualised rates to see past the base effect. Third, read the monthly pattern rather than the single print. Fourth, check whether the index level is still rising, regardless of the rate. Fifth, glance at the contribution table to see which categories did the work.
That is enough to keep you out of most of the traps, and it takes less time than reading the commentary you would otherwise be tempted by. One caution: if you are making a decision with money attached — a mortgage, a contract with indexation, a pension assumption — treat the report as an input and not a conclusion. A qualified adviser who knows your situation is worth more than any single data release.
Inflation figures are a photograph of a moving object, taken with a slightly shaky hand. Read them as evidence, not as an announcement.
Photo: Magda Ehlers / Pexels