Bagged Salad's Real Price: What a Gram of Greens Costs After Waste

Two federal series price salad greens at the national level, and they do different jobs.

The first is the Bureau of Labor Statistics Average Price program, which publishes a monthly dollars-per-pound figure for whole heads. As of the August 2026 release, iceberg lettuce averaged $1.471 per pound and romaine averaged $3.286 per pound across US cities. Those are as-purchased prices: head, core and all. Romaine's trailing twelve months ran $3.195 to $4.021 across ten published months. Iceberg is noisier and published less consistently — the series has only two observations in 2020, none at all in 2021 or 2022, and eight in 2023 — so its twelve-month mean of $1.774 rests on nine months rather than twelve.

The second is the Economic Research Service's Purchase to Plate National Average Prices file, which prices foods as people actually buy them. Its 2017/18 vintage is built from retail scanner sales linked to national dietary survey food codes, and it prices mixed salad greens at $1.345 per 100 edible grams — which is to say, the bagged and clamshell formats, weighted by how much of each Americans bought. The same file gives raw spinach at $0.990, kale at $0.681, arugula at $2.040, and a romaine-and-leaf row at $0.784.

Those PP-NAP figures are eight years old, so they need deflating. Running the lettuce row through the Consumer Price Index lettuce series and the rest through the other-fresh-vegetables series puts all of it in August 2026 dollars. The lettuce index has risen 37.4 percent since its 2017–18 average; other fresh vegetables, 20.6 percent.

Here is what the two series say side by side, before a single adjustment.

Two things are worth flagging before anyone reads a verdict into that table. The PP-NAP rows are already expressed per edible gram, which means the convenience formats have had their trim netted out at the source. And the romaine-and-leaf row bundles romaine hearts with endive, chicory, escarole and green leaf, which is why it lands 40 percent above the BLS whole-romaine figure on an edible-weight basis. Different populations of product, not a contradiction.

The reason a shelf tag cannot answer the question is the same reason store-level price rankings cannot: comparing retailers on price per ounce measures what leaves the store, not what gets eaten.

The intuitive objection to a head of lettuce is that you are buying inedible weight. The federal nutrient database puts a number on exactly that, and the number is smaller than most people assume — for the two greens this article is about.

USDA's Standard Reference Legacy release records a refuse percentage and a refuse description for every whole food. Iceberg lettuce carries 5 percent refuse, described simply as "Core," and cos or romaine carries 6 percent, also "Core." A romaine head averages 626 grams, so the core you discard is roughly 38 grams. At August 2026 prices that is about 27 cents of a $4.53 head.

Where refuse gets expensive is the rest of the lettuce shelf, and this is the part the "buy the head" advice tends to skip.

A bunch of green leaf lettuce loses more than a third of its purchase weight before it reaches a bowl. Bunch arugula loses two-fifths. Those are the cases where the bag's zero-trim advantage is real money rather than a rounding error — and they are also the cases where the bagged version of the same green is the one most shoppers are actually buying.

Applying refuse to the two BLS head prices gives the first honest comparison in this article. Iceberg moves from $1.471 to $1.549 per edible pound, or 34 cents per 100 edible grams. Romaine moves from $3.286 to $3.496, or 77 cents per 100 edible grams. Against the $1.62 that bagged mixed greens cost per edible 100 grams, the head is still ahead by a factor of about two — and against iceberg, by a factor of nearly five.

That is the whole trim adjustment. It changes the romaine figure by 6.4 percent. Hold that number; the next one dwarfs it.

Greens are the fastest-spoiling thing most households buy, and federal data tracks the consequence in two separate places.

The first is the Economic Research Service's Loss-Adjusted Food Availability series, which decomposes the gap between what is produced and what is eaten. For fresh head lettuce and for fresh romaine and leaf lettuce, the consumer-level "other" loss — cooking loss and uneaten food — is assumed at 24 percent, on top of nonedible shares of 16 and 20.7 percent respectively. Spinach is assumed at 9 percent uneaten, kale at 38 percent. Stacked across the whole chain, the series puts total loss at 45.6 percent for head lettuce, 55.3 percent for romaine and leaf, 52.8 percent for spinach, and 75.6 percent for kale.

Read that carefully, because it contains a reconciliation worth having. ERS assigns romaine-and-leaf lettuce a 20.7 percent nonedible share while the nutrient database assigns romaine itself 6 percent. Both are right. The ERS category is a blend, and the leaf lettuces in it carry 20 to 36 percent refuse. The blend is doing what a blend does.

One caveat belongs on the record: these consumer-level loss percentages are fixed assumptions rather than annual measurements. They have not moved since at least 2018, and ERS itself describes the loss-adjusted series as preliminary because the underlying assumptions require further work. They are the best national figures available and they are not observations of your refrigerator.

The second place is the cost side. The Environmental Protection Agency's 2025 accounting of household food waste puts the total at about $728 per person per year, roughly $14 a week and around 11 percent of consumer food spending. Fresh vegetables rank fifth by dollar value at $59.49 per person and second by sheer weight at 31.0 pounds. The agency is careful to note that the heaviest-wasted categories are not automatically the biggest savings opportunities — but greens sit high on both lists, which is unusual.

For anyone tracking how much of a grocery budget is structurally unrecoverable, the state-by-state view of what a healthy grocery basket actually costs is the companion number to this one.

The spoilage side becomes tractable once you stop treating "goes bad fast" as a vibe and start treating it as a published window.

USDA's FoodKeeper dataset is the reference. Its current release is version 128, dated September 6, 2018, and the date matters enough to print. Within it, whole lettuce — iceberg and romaine — is given one to two weeks refrigerated from date of purchase, while bagged greens are given three to five days unopened and exactly two days after opening. Loose leaf lettuce and spinach get three to seven days. Kale gets three to five. Arugula gets three to seven.

Two days. That is the entire operating window of an opened bag of salad, and it is the single most consequential number in this article.

It is also easy to miss, because it is not printed on the chart most people find first. The cold-storage chart on FoodSafety.gov lists no entry at all for salad greens, lettuce, bagged greens or pre-cut produce — its only salad line is for egg, chicken, ham, tuna and macaroni salads at three to four days. The produce windows live in the FoodKeeper database, one layer down.

Set the two windows against each other and the asymmetry is stark. A whole romaine head has seven days at the conservative end. An opened bag has two. That is a 3.5-fold difference in the time available to eat the thing, and it applies to the format that costs twice as much per edible gram to begin with.

Shoppers weighing the two formats on grounds other than price have a separate set of trade-offs to consider, and the handling questions that surface after a multistate lettuce recall sit outside this calculation entirely.

Here is the whole model in one place. Effective price per 100 edible grams equals the sticker price, divided by one minus the trim share, divided by the share actually eaten. BLS rows get the trim divisor because they are priced as-purchased. PP-NAP rows do not, because that file already reports price per 100 edible grams — netting trim out twice would be double-counting.

Three eaten-share scenarios run across the table:

• All eaten — the ceiling nobody hits, included so the other two have something to be measured against.
• LAFA — the federal consumer-level uneaten assumption for that specific commodity: 24 percent for lettuce, 9 percent for spinach, 38 percent for kale.
• Window + 2 — the package outlives its FoodKeeper window by two days and whatever remains at the end of the window is discarded. The eaten share is the window divided by the window plus two: 78 percent for a head, 50 percent for an opened bag.

Read down the romaine row and the proportions come into focus. The as-purchased sticker is 72.4 cents per 100 grams. Trimming the core moves it to 77.1 cents — a 4.6-cent adjustment. Missing the storage window by two days moves it to 99.1 cents — a 22.0-cent adjustment. The spoilage penalty is 4.8 times the trim penalty, on the format where trim is supposedly the problem.

For bagged greens there is no trim penalty at all and the spoilage penalty is the entire story: $1.62 becomes $3.25 on a two-day overshoot, and $4.06 if the bag is picked at for five days. Converted to something a person can feel, an 85-gram reference serving costs 29 cents from an iceberg head, 66 cents from a romaine head, and $1.38 from a bag of mixed greens with nothing wasted.

None of this is unique to greens. The same structure — a premium format that sells convenience and charges for it — shows up wherever the per-gram cost of a ready-made version can be set against the ingredients.

The comparison most shoppers think they are making — bag versus head — has a stable answer across every scenario above. The comparison they are actually making, standing in front of a case that offers the same greens in three sizes, does not.

The usable rule is a required discount: the markdown a package needs to carry, per gram, before its extra spoilage is paid for. It needs no price data at all, which means it works on any shelf tag, in any city, in any week.

Take a household eating one 85-gram reference serving per person per day, and the two-day opened window from FoodKeeper.

For one person, the 10-ounce bag has to be priced 40 percent below the 5-ounce bag per gram to break even, and the 16-ounce clamshell has to be 63 percent below. Unit pricing inverts: the package with the lower shelf price per ounce is the more expensive one per gram eaten. For two people the same clamshell needs only 25 percent, and the 10-ounce bag needs nothing at all — it finishes inside the window.

The head-versus-bag version of the same rule is blunter. To tie a whole romaine head with both formats fully eaten, bagged mixed greens would need to be 52.5 percent cheaper than they are. With the head at its window-plus-two eaten share and the bag at its own, the gap widens to 69.5 percent. Against iceberg it runs from 79.0 to 89.5 percent. Those are not discounts that exist.

What the bag buys, then, is not a price. It is the removal of a preparation step and the elimination of trim — genuinely worth something to a household that would otherwise let a head sit uncut. Roughly twice as many packages of bagged salad greens are bought as heads of lettuce, a shift that tracked with Americans eating more greens overall, which is a real argument for the format that a per-gram figure cannot capture. It is just not a price argument.

Readers who like this shape of calculation will recognize it from the batch-cook versus pre-packaged comparison, where the same required-discount logic decides the answer.

Several limits are worth stating plainly rather than burying.

The PP-NAP file is eight years old and will stay that way for a while — ERS expects the 2021–23 vintage no earlier than spring 2027. Deflating by a category index is a reasonable bridge, not a measurement, and it assumes bagged greens tracked their CPI category rather than diverging from it. Method-2 rows in that file are imputed and were excluded, which is why iceberg is priced from BLS here rather than from PP-NAP.

PP-NAP prices are national scanner-weighted averages across every store format and every promotion, not a shelf quote. BLS average prices are national too. Neither is what your store charged this week, and the gap between a national average and a regional shelf can easily exceed the differences this article is measuring for any single shopper.

The FoodKeeper release is from 2018. It remains the current version, and its windows are the conservative published ones rather than what a cold, well-managed refrigerator might deliver. The eaten-share scenarios are scenarios, not forecasts — the "window plus two" case is a construct chosen because it is symmetric across formats, not because households overshoot by exactly two days.

And the biggest limit is the honest one: the eaten share is the variable, and it is the one the data cannot observe for you. Two households buying the identical clamshell can land 30 cents apart or $2.70 apart per 100 grams depending on nothing but how fast the drawer empties. That is the finding, restated. The format decision is worth a factor of two. The consumption-rate decision is worth a factor of three on top of it.

Anyone running this arithmetic on a whole basket rather than one aisle will find the cost-to-the-cent approach applied to a full week of meals uses the same machinery.