A smaller increase still leaves a bigger bill
A receipt for the same quantities and quality of groceries can remain expensive while a headline correctly reports lower inflation. The price level is the basket’s cost; inflation is its percentage change over a stated period. A smaller positive percentage still makes the next basket more expensive.
Headlines about ‘easing’ or ‘cooling’ usually describe that slower increase. They do not establish that earlier increases have reversed or that income has recovered its lost purchasing power. Those are separate comparisons.
Follow one basket through three years
Take a fictional basket that initially costs $100. Keep its contents fixed so that we are comparing the same purchase throughout. These numbers illustrate the arithmetic; they are not a record of any country’s inflation.
| Point in the example | Increase over the preceding year | Basket price |
|---|---|---|
| Starting point | — | $100.00 |
| End of year one | 10% | $110.00 |
| End of year two | 5% | $115.50 |
| End of year three | 2% | $117.81 |
Each percentage applies to the latest price. The final step adds 2% of $115.50, or $2.31. Adding the three rates together would miss this compounding: the total increase is 17.81%, rather than 17%.
The annual rate has fallen sharply, yet the basket costs $17.81 more than it did at the start. Nothing in that calculation requires a mistake in the headline or an imaginary increase on the receipt.
Disinflation, zero inflation and deflation
The St. Louis Fed calls a slowing inflation rate disinflation. In our example, the move from 10% to 5% is disinflation; so is the move from 5% to 2%. Prices keep rising through both stages.
Zero inflation over a particular period would leave the overall price level unchanged between its endpoints. It would preserve the expensive starting point, rather than restore an earlier one.
Deflation means a decline in the general price level. A discount on one product is not enough to establish it. An economy can have cheaper fuel and more expensive housing at the same time, with the combined index still rising.
The comparison year can change the story
An annual rate compares the current price level with the level twelve months earlier. That moving reference point can make the reported rate change even when a particular price has stopped moving.
Suppose a fictional loaf rises from $2 to $2.20, then stays there. During comparisons with the old $2 price, it shows a 10% increase. Once both ends of the annual comparison are $2.20, its year-on-year increase is zero. Nobody has reduced the shelf price.
The ECB explains this mechanism as a base effect: an unusually large monthly movement eventually drops out of the twelve-month comparison. Its 2021 discussion concerns euro-area energy prices and taxes. The loaf is our own simplified illustration of the comparison, not a finding from that historical episode.
This is why the annual rate alone cannot settle what prices did last month. The time period and reference point belong beside the number.
An average basket is not your shopping basket
The U.S. Bureau of Labor Statistics constructs the Consumer Price Index from a representative collection of goods and services. Expenditure shares determine how much different categories contribute. Housing, transport and groceries do not each receive an equal vote simply because they appear on a list.
BLS also warns that the average does not necessarily match a particular household’s experience. A renter facing a large increase may gain little from falling prices for goods they rarely buy. Someone who drives daily has a different exposure to fuel prices from someone who walks to work.
That does not make the index meaningless. It makes it an answer to a population-level measurement question. Your own budget asks a narrower question: what happened to the prices of the things you actually pay for?
Why the target does not restore old prices
The Federal Reserve’s stated longer-run inflation goal is 2%, measured using the personal consumption expenditures price index. PCE and CPI are different measures; the target should not be silently relabelled as a CPI target.
A positive target aims for a continuing, modest increase in the price level. It is not a commitment to return groceries to a particular year’s prices. The Fed explains that monetary policy influences overall demand through financial conditions, including borrowing costs. That affects the pressure on future price growth; it does not issue instructions to individual shops to undo earlier increases.
Why not pursue general price declines instead? Philip Jefferson’s 2023 speech describes one concern: if deflation comes with flat or falling nominal wages, fixed debt payments can become harder to meet. A mortgage bill does not shrink automatically with the borrower’s income. This is a conditional risk, not a claim that every cheaper product damages the economy.
Affordability needs an income comparison too
BLS distinguishes nominal income, the dollar amount you receive, from real income, its purchasing power after adjusting for prices. That distinction explains how affordability can improve without a return to old prices.
In a hypothetical year with 5% income growth and 10% price growth, purchasing power falls: 1.05 divided by 1.10 is about 0.955. A later year with 5% income growth and 2% price growth raises purchasing power by about 2.94%, because 1.05 divided by 1.02 is about 1.0294. That later improvement does not fully repair the earlier loss.
Income growth is also uneven and never guaranteed. Lower inflation can prevent a budget from deteriorating as quickly while leaving its owner under real strain. To assess a cheerful headline, keep three questions separate: are prices rising more slowly, has the price level fallen, and is income buying more? They can have three different answers.
