Walk into a loss prevention office and say the word shrinkage and nobody will think about package sizes. In retail the word already means something. It means product that left the building without being paid for.
Stolen. Broken. Miscounted. Written off.
The industry takes this seriously in a way that is hard to overstate. American retailers lost about 1.68 percent of revenue to shrink in the most recent full survey. That is the highest rate in more than a decade. At a 1.6 percent rate a couple of years earlier that came to $112 billion. There were thirty-two consecutive years of a national survey devoted to counting it. There are loss prevention departments, locked cases, guards, and cameras.
An entire profession exists to measure product disappearing off a shelf.
Now consider the other way product disappears.
The one nobody counts
A mega roll of toilet paper carried 264 sheets in 2022. It carries about 224 now. The sheet itself narrowed, from the 4.5 inches that used to be standard to 3.9 on most brands.
Nothing was stolen. Nobody broke in. The paper was removed at the factory, deliberately, by the company that sells it.
There is no thirty-two year survey of that. There is no department. The shopper does not notice, which is the entire point. And the retailer is not the victim, so has no reason to count it.
Why a company shrinks instead of charging
The obvious question is why bother. A firm that needs more money per package could simply raise the price. Shrinking it costs engineering time, new tooling, new artwork. And the risk of being caught at it.
The Government Accountability Office looked at this and found the answer, and it is a number.
Across six studies, a 1 percent rise in price costs a brand about 1.95 percent of its sales. A 1 percent cut in size costs it 0.76 percent.
Shoppers punish a price rise roughly two and a half times harder. Same loss, delivered as less product, and they mostly let it go.
That is not a story about deception for its own sake. It is a measured demand curve, and it is telling the firm something specific.
What the gap actually says
Read those two numbers through the Toilet Paper Theory and they stop being a marketing fact.
The theory says that for a necessity, the quantity a household can use is roughly fixed. You use what you use. And a household at its ceiling has no room above it.
Now put yourself inside the company. Raise the price and the customer sees a number they cannot afford. So they go to the store brand. That is the 1.95. Cut the size and the customer buys the same thing at the same sticker. The sticker is what they were checking.
But notice the second half, which is the part that matters here. They do not buy less toilet paper. They cannot. They use what they use, so a smaller roll simply means coming back sooner.
So shrinkage is a confession. Not of dishonesty, necessarily. Of a market where the seller has concluded two things. That demand is capped, and that the buyer is pinned against a price they will not cross.
Where it happens is the finding
Here is the part that turns this from a complaint into evidence.
Downsizing is not spread evenly across the store. The GAO ranked which categories it contributed most to, and the list is striking.
| Category | Downsizing contribution | Can a household simply buy less? |
|---|---|---|
| Household paper | 3.0% | No. Usage is set by the household, not the budget. |
| Snacks | 2.6% | Partly, but bought in fixed habits and lunchboxes. |
| Candy and gum | 2.3% | Partly. Small tickets, bought on impulse. |
| Ice cream | 1.8% | Partly. One tub is one tub. |
| Breakfast cereal | 1.6% | No. Someone eats breakfast every day. |
| Coffee | 1.4% | No. A daily habit, and famously inelastic. |
| Pain relievers | 0.5% of sales | Barely downsized at all. Bought when needed, not on a schedule. |
Look at what is at the top. Paper, cereal, coffee. Things a household consumes on a schedule it does not control.
Look at what is at the bottom. Pain relievers, where only about half a percent of downsized items showed up in sales.
The pattern is not random and it is not about which firms are greediest. It tracks whether the household can respond by using less.
Which means the shrinkage map is a map of capped demand. Where a household can cut back, firms raise the price and accept the lost units. Where it cannot, they shrink the package and keep them.
The complication, which is large
Now the number that argues against everything above, and it belongs here rather than in a footnote.
Across the whole economy, downsizing is almost nothing.
The GAO put its contribution at 0.06 percentage points. Consumer prices rose 34.5 percent between December 2019 and December 2024. That is about one hundredth of a point a year. As an explanation for the cost of living, shrinkage is a rounding error.
Anyone telling you that hidden downsizing explains your grocery bill is wrong, and the federal measurement says so.
Two things follow, and they point opposite ways.
The first is that this is not an inflation story and should stop being sold as one. The second is that a signal can be small in aggregate and still be informative. A fever of one degree is small. It still tells you something. Downsizing is concentrated in a handful of categories, and which ones is the interesting part.
One correction worth making
There is a claim that circulates about shrinkflation and it is false. The claim is that official inflation misses it.
It does not. The Bureau of Labor Statistics collects effective unit prices. Its data collectors are trained to flag size changes. Its commodity specialists then apply a quantity adjustment, so that less product at the same sticker registers as a price increase. That is why the GAO could put a number on it at all.
The adjustment lags by months, and the shopper standing in the aisle sees none of it. But the index is not fooled. A framework that argues from measurement has to say so, even when the sloppier version would suit it better.
Why this is the purest case in the theory
The Toilet Paper Theory began with a simple claim. A shortage can be counterfeit. The goods exist and the household simply cannot reach them. From the empty shelf, the fake and the real look identical.
Shrinkage is that claim running in reverse, and it is cleaner.
There is no empty shelf at all. The shelf is full. The package is the same height, the same color, in the same place, at the same price. Everything a shopper checks is unchanged.
A counterfeit shortage is a real appearance with nothing behind it. Shrinkage is a real reduction with no appearance at all. They are the same trick from opposite ends. The second is harder to see. The first at least leaves a gap on a shelf where something used to be.
The ceiling gets paid either way. This is just the quietest way to collect.
What would prove this wrong
Three things, all checkable.
Say downsizing starts showing up mainly in discretionary categories rather than staples. Then the capped demand reading is wrong, and something else is driving it. Watch where the GAO's next category ranking lands.
If sheet counts and package weights start rising again while shelf prices hold, the pressure has come off. The ceiling has slackened. That would be a genuine easing, and it would show up here before it showed up in a headline.
And say the elasticity gap closes, with shoppers punishing size cuts about as hard as price rises. Then the commercial case collapses, and firms go back to pricing. Labeling rules of the kind France and South Korea have adopted are the most likely way that happens. The GAO lists them among its policy options.
Two cautions. The elasticity figures average six studies of different products and periods. Treat 2.6 as the shape of the effect, not a constant. And the sheet counts quoted here come from brand comparisons, not a government series. They illustrate the category. They do not measure it. The GAO figures are the ones to lean on.