Official inflation says 87%. Your receipt says 220%.
Someone is wrong about inflation, and it isn't your receipt. Four structural reasons why the official number was never measuring your life to begin with.
There's a moment everyone has had at the checkout in the last few years. The total flashes up, and your brain does a little stutter-step. That can't be right. You scan the cart looking for the wagyu you didn't buy, the truffle oil that must have fallen in. Nope. Eggs, bread, coffee, dish soap. Just groceries.
Then you go home, turn on the news, and a very calm person tells you inflation is running at 3 percent. Under control. Cooling, even.
Both of these things cannot be true. And I'm going to argue something that sounds conspiratorial but is really just arithmetic: the official number is not measuring your life. It measures someone else's, an “average” person who does not exist, buying a basket of things you don't buy, with statistical adjustments you'd never agree to if anyone asked.
Let me show you the receipts. Literally.
The $190 hole
When a 1997 H-E-B grocery receipt went viral last year, 122 items for $155, the internet did the obvious thing and re-bought the whole haul at today's prices. The new total: just over $500.
The part that matters comes next. According to the Bureau of Labor Statistics' own inflation calculator, $155 in 1997 should equal about $312 today.
$312 is what inflation says happened. $500 is what actually happened. That's a $190 hole per cart, every cart, and one commenter nailed why it stings: “Adjusted for inflation $155 would be $312. The extra $192 needed is the problem.”
So where did the $190 go? Nowhere, really. It's hiding in four places the official number is structurally bad at seeing.
Hiding place 1: the shrinking package
Before you even get to price increases, there's the increase they don't print on the tag. The 500g pack becomes 450g. The six-pack becomes five. The chocolate bar quietly loses two squares. The shelf price barely moves, the index records almost nothing, and your actual cost per unit just jumped 9 to 16 percent.
Statisticians will tell you, correctly, that CPI tries to track price per unit. But shrinkflation works precisely because you don't compare unit prices; you compare the package to your memory of the package. A cereal box that drops from 18 oz to 15 oz at the same price is a 17 percent per-ounce increase that doesn't feel like inflation at all.
One analyst called shrinkflation “a tax on consumer attention,” and I can't improve on that. Consumer research estimates that shrinkflation and its uglier cousin skimpflation, same price and worse product, added roughly 2 to 4 percentage points to the inflation real households experienced from 2021 to 2024, on top of the official figure.
Hiding place 2: the statisticians assume you downgraded
There's a methodological choice buried in here that most people have never heard of, and it should make you angry. Official CPI builds in substitution: when beef gets expensive, the model assumes you rationally switch to chicken, so measured inflation gets adjusted downward, whether or not you actually switched, and whether or not a genuine substitute exists for your situation.
Think about what that means. If you kept buying the food your family actually eats, the index quietly assumed you didn't, and marked your inflation lower for it. The cheaper flat two boroughs away is not a substitute for the flat near your kid's school. The methodology isn't fraud; it's a simplifying assumption. But it is, by design, a downward distortion of the price increases faced by anyone with real constraints.
Same story with hedonic adjustments: if this year's product is judged higher-quality than last year's, part of its price increase simply doesn't count as inflation. Defensible for laptops. Insulting for a chicken breast.
Hiding place 3: the basket isn't your basket
CPI is a weighted average across a standardised basket for a hypothetical average urban consumer. Nobody is that consumer. If housing eats 45 percent of your budget while the official basket weights it at 25, and rents are rising fast, your real inflation runs well above the headline, and both numbers are technically true.
And the categories that ran hottest since 2020 are precisely the ones you can't opt out of.
Approximate price change since 2020, by category:
| Category | Change |
|---|---|
| Auto insurance | +35% |
| Housing | +25% |
| Food away from home | +25% |
| Electricity | +20% |
| Childcare | +15% |
| Electronics and clothing | flat or down |
This is why the Common Man CPI, an alternative index tracking only necessities like food, energy, clothing and shelter, showed worse inflation than official CPI in nearly every month between 2020 and 2024, peaking near 12 percent when the headline peaked around 9. An Everyday Price Index of frequently purchased items shows the same pattern over a longer arc: from 2001 onward it rose about 3.1 percent a year against CPI's 2.3.
When you strip out the discretionary stuff and measure what people are forced to buy, the gap between the news and your receipt shrinks dramatically. Funny how that works.
Housing deserves its own indictment. Because CPI imputes homeownership costs from rents, one analysis found the true cost of buying a home rose roughly 80 percent from January 2021, while the CPI's shelter component implied about 20. Anyone who tried to buy a house in that window does not need a footnote to confirm which number is closer to reality.
Hiding place 4: your brain (yes, some of the gap is you)
Honesty requires this section, so here it is. Part of the perception gap runs the other way. Humans weight price increases far more heavily than decreases, and when researchers built an index trimming out the steep price declines consumers mentally ignore, the gap between measured and perceived inflation nearly vanished.
We also over-index on things we buy weekly, food and fuel, and under-index on things we buy rarely, the well-documented frequency bias. And once inflation grabs your attention, the attention sticks: perceptions stay elevated long after the actual rate falls.
So no, CPI is not a conspiracy. It's a consistent, carefully defined statistical average, and its limitations come from the same standardisation that makes it useful.
The only inflation rate that matters is yours
The practical takeaway is more empowering than the doom-scroll version. Since no headline number describes your life, the only useful move is knowing your own numbers. People who track their real category spending routinely find their personal inflation rate sitting 2 to 5 points away from the official figure, in one direction or the other.
A renter with a new lease and a grocery-heavy budget lives in a different economy than a homeowner on a 2021 fixed rate, and both of them live in a different economy than the evening news.
Which raises an uncomfortable question: how well do you actually know what things cost right now? Not what they cost when you first started paying attention. Now. Most people's mental price list is years out of date, and the gap between remembered prices and real ones is exactly where shrinkflation and quiet repricing live.
Sources
- AOL: 1997 grocery receipt stuns TikTok
- Yahoo Creators: A viral grocery receipt from 1997 is breaking people's brains
- Money Progression: Personal inflation rate calculator
- Sadiq: Personal inflation rate vs headline CPI
- NNNG: Personal inflation calculator and category breakdown
- Citrine Capital Advisors: Personal inflation vs CPI
- Fox Business: A tale of two economies
- BLS Monthly Labor Review: A price index that matches perceptions of inflation
- Statistics Canada: Perceived versus measured inflation
- European Parliament: Inflation perceptions across household groups
- CEPR VoxEU: The perceived inflation wedge
- MultiCalculators: How CPI is calculated
More guides
- What groceries cost the year you were born, and what they cost in workEggs, bread, gas and coffee in 1980, 1990, 2000 and 2010, from BLS average prices, and how many minutes of work each took then and now.
- Groceries take 24% less work than in 1980, and 12% more than in 2020A fixed basket of 12 U.S. groceries and gas took 112 minutes of work in 1980, 76 in 2020 and 86 in 2025. Both halves of the grocery argument are right.
- Only three groceries got harder to buy since 1980Of 15 U.S. groceries BLS has priced since the 1980s, only ground beef, eggs and bacon now take more minutes of work. Bananas take 58% less.
- How to play Price Year, Open Range and Pricele 1000Three new daily price games: date a grocery receipt, give a price range, or score out of 1,000. The rules, the scoring and how to get better at each.