The bill rises by $8, then by another $3
Start with purchases costing $100, including food, transport and other goods and services. Buy exactly the same amounts at the same quality on the same date each year. These categories and prices are invented. They are not a country’s official price statistics or a suggested household budget.
Scroll wide tables sideways. Keyboard: focus the table, then use the arrow keys.
| Category | Starting point | Year 1 | Year 2 |
|---|---|---|---|
| Food | $40 | $44 | $46 |
| Transport | $30 | $33 | $33 |
| Other spending | $30 | $31 | $32 |
| Total | $100 | $108 | $111 |
| Change from a year earlier | Not calculated | 8% | 2.78% |
In the first year you pay $8 more than $100. That is eight parts out of a hundred, or 8%. In the second, you pay $3 more than $108. Divide the extra cost by that year’s starting bill: $3 ÷ $108 × 100 ≈ 2.78%. The percentage increase is smaller, while the actual bill is three dollars bigger.
Compared with the original $100, the basket now costs 11% more. Adding 8% and the rounded 2.78% gives 10.78%, which is wrong because the second percentage uses a different starting amount. Using the exact ratios, 1.08 × (111 ÷ 108) = 1.11.
The food portion has risen from $40 to $46, a 15% increase over the two years. Even within this tiny example, the grocery bill and the total basket tell different stories.
A slower increase still leaves you paying more
The $111 you must pay is the price level of these purchases. How much that price changes over a stated time is the inflation rate. Here the price level goes from $108 to $111 while the yearly inflation rate goes from 8% to about 2.78%. The two numbers describe different parts of the same change.
Prices still rising by a smaller percentage is called disinflation. A fall in the general price level is called deflation. Our example has the first: the rate of increase falls, but the price does not. Asking “did it rise less, or did it actually fall?” is enough to tell these terms apart.
These terms describe an overall pattern. They do not require every price to move together. Transport in the table stops rising in Year 2; food and other spending do not. One discounted product would not establish that the economy was experiencing deflation.
Even 0% inflation would leave this basket at $111
Now imagine three possible outcomes over the following year. Each starts at $111. The 2% figure is just an assumption for this calculation, not a forecast or a statement about every central bank's target.
| Annual change | Calculation | Next year's cost |
|---|---|---|
| +2% | $111 × 1.02 | $113.22 |
| 0% | $111 × 1 | $111.00 |
| −2% | $111 × 0.98 | $108.78 |
At 0%, there is no further increase in the basket's total. At −2%, it becomes cheaper, but still costs more than the original $100. Returning all the way to $100 would require a change of ($100 ÷ $111 − 1) × 100 ≈ −9.91%.
The required fall is not 11%. The earlier 11% rise was measured against $100; the reversal is measured against $111. Neither slower inflation nor a particular negative rate automatically restores a price you remember.
Your grocery budget need not match the average
If food takes more of your budget, a rise in food prices affects you more. If transport takes more, its prices matter more. A spending weight is simply that category’s share of spending. Two households can face the same prices yet have different increases in the cost of their usual purchases because they buy different amounts. Official price measures cover many purchases, beyond groceries alone.
For a separate example, suppose food prices rise 10%, transport prices stay flat and other prices rise 2%. Two households each start with $100 of purchases but divide it differently. Keep each household's quantities and quality fixed.
| Starting spending | Household 1 | Household 2 |
|---|---|---|
| Food (+10%) | $50 / 50% | $20 / 20% |
| Transport (0%) | $20 / 20% | $50 / 50% |
| Other (+2%) | $30 / 30% | $30 / 30% |
| Starting total | $100 | $100 |
| Cost after the changes | $105.60 | $102.60 |
| Increase | 5.6% | 2.6% |
Household 1 pays $5 more for food and $0.60 more for other purchases: a $5.60 increase. Household 2's food increase is only $2, so its total rises by $2.60. The prices changed identically; food takes a larger share of the first household's budget.
A consumer price index, or CPI, combines the price changes of many goods and services people buy. This small, unchanged group of purchases does not reproduce an official CPI. Statistical agencies use much more information, update spending shares and allow for product quality changing. A better product costing more is not automatically treated as the same product simply becoming dearer.
Your own higher spending can also reflect buying more, choosing a different product or changing package sizes. Before treating a bigger bill as a price increase, check what you bought.
Check what the headline is comparing
A headline saying inflation has eased leaves two questions open: compared with when, and for which purchases? A monthly food-price change and a twelve-month change in the full consumer basket can move in different directions. Find the period and coverage in the report before comparing it with your own bill.
If you want to compare your receipts, match quantities and quality before comparing totals. A larger pack can cost more while being cheaper per unit. Keep that spending change separate from a price increase.
In our first basket, “annual inflation fell to 2.78%” and “the same purchases cost 11% more than two years ago” are both correct. The first describes the latest year's change. The second explains why the old $100 budget no longer covers the bill.
Different US inflation reports also measure different things. Our CPI, PCE and core inflation guide compares their coverage and exclusions, then follows one fictional index through monthly and annual changes. It also explains base effects: how the price level used for comparison can make the current rate look larger or smaller.
About the examples
The baskets, budgets, prices and periods are fictional. Quantities and quality stay fixed within each comparison, and household weights use spending before prices change. Calculate with full values, then round for display. These examples are not an official consumer price index or a forecast. See our editorial policy.