Dispatch · July 16, 2026 · 15 min read · By Don Goldstein

The menu looks calm. Five fronts are pushing.

+3.4% against 42%. Restaurant menu prices rose 3.4% over the year to June 2026, the tamest inflation in years — yet 42% of operators still closed last year without a profit. The gap between those two numbers is five cost fronts, moving on five different clocks under one calm average.

Why restaurant costs feel worse than the 3.4% number

I keep the order guide on a clipboard by the walk-in, and in June the number that stopped me wasn’t the beef. It was a case of to-go lids, up a few cents each, quiet, easy to sign for without looking. The beef I had priced in months ago. The lids I hadn’t, and a few cents on every order that leaves the building is real money by Sunday. If you run a floor, you know that feeling in your stomach when a line you weren’t watching moves. You already price your proteins and watch your food cost every week; the lines that hurt in 2026 are the ones nobody told you to watch.

Here are two numbers that are both true, and the space between them is the whole dispatch. Restaurant menu prices rose 3.4% over the year to June 2026 — the tamest menu inflation in years. And 42% of operators closed last year without a profit. The average is calm. The floor is not.

The average is the tamest it has been in years. The spread is the widest I have paid.

I’m not here to tell you costs are exploding, because the data and my own invoices disagree with that story. Food away from home — the government’s name for a restaurant meal — rose 3.4% over the year to June 2026, with full-service at 3.7% and fast food at 3.1%. That is a blended figure, one line drawn across every restaurant in the country. It gives you the temperature of the room and nothing about the fire in the corner.

The counterweight sits right next to it. In the National Restaurant Association’s 2026 read, 42% of operators said their restaurant wasn’t profitable last year, and about 60% said traffic got softer even as their sales dollars grew. Checks up, guests down. More than seven in ten diners say they’d eat out more if they had more money to spend. Right now the revenue is carried by higher prices, not more covers, and that is a fragile way to carry it.

Underneath the average, the menu splits into fronts moving on their own clocks. This is the chart that should reset the panic in either direction.

One year of price moves: the menu average vs the lines under it

Restaurant menus (all)

+3.4%

Full-service menus

+3.7%

Grocery (2026 forecast)

+2.8%

Retail beef (year-on-year)

+12.9%

Coffee (April peak, YoY)

+18.5%

Beef and coffee are retail shelf prices, not your delivered wholesale cost — use them for direction, not for your plate math. Coffee has since cooled to a +2.9% beverage reading by June.

The menu average is the tamest in years; the two loudest lines beneath it move roughly four to five times faster.
Sources: BLS Consumer Price Index and USDA ERS, June/2026 outlook

U.S. Bureau of Labor Statistics — Consumer Price Index, food away from home +3.4% over the 12 months ending June 2026 (full-service +3.7%, limited-service +3.1%); retail beef and coffee indexes for the year-on-year and April figures. bls.gov/cpi. USDA Economic Research Service — 2026 Food Price Outlook: food-away-from-home forecast +3.6%, food-at-home (grocery) +2.8%, all-food +3.2%. ers.usda.gov/food-price-outlook. National Restaurant Association — 2026 State of the Restaurant Industry: 42% of operators not profitable last year; ~60% reported softer traffic; more than 9 in 10 cite food, labor, insurance, energy and swipe fees as significant challenges; $1.55 trillion projected 2026 sales. restaurant.org. Retrieved 2026-07-16.

An average is a compromise. It smooths five very different pressures into one polite figure, and the smoothing is exactly what hides the pain. USDA’s outlook for the year has restaurant prices rising 3.6% against grocery’s 2.8% — eating out getting more expensive faster than eating in, which your guest feels at the table every time. More than nine in ten operators name food, labor, insurance, energy and card fees as significant challenges, all at once. The industry is still projected to ring up a record $1.55 trillion in sales this year. The rest of this piece is the corners the average paints over: five fronts, one at a time, and which one is actually on your menu.

Front one: why beef is expensive in 2026, and why it isn’t the tariffs

Beef is the domestic cattle cycle — not tariffs, not weather — and it is the one genuinely structural front in the whole storm. If you run beef, this is the line doing the most damage to your plate cost this year, and it is the exception that proves the rule about the spread: a real, multi-year squeeze hiding under a calm menu average.

The number that matters isn’t a price, it’s a headcount. The US cattle herd on January 1, 2026 was 86.2 million head — the smallest since 1951, a 75-year low. Beef cows came in at 27.6 million, down another 1%. Years of drought and high feed costs pushed ranchers to sell down their breeding stock, and you cannot rebuild a herd in a quarter. Retail beef ran 12.9% higher than a year earlier in May 2026, though it slipped 1.3% month-on-month — still expensive, no longer accelerating. The level is high; the momentum is cooling.

Now the honest part, because even the structural front comes with a forecast, and forecasts wobble. USDA’s 2026 call for beef and veal is about +7.5%, inside a wide band of 3.1% to 12.2% — and that call was revised down from roughly 12.1% a month earlier. Read the herd count as the signal and the forecast percent as the noise around it. The signal says tight; the noise says nobody knows the exact number.

  1. 1

    The herd is the smallest since 1951

    86.2 million head on January 1, 2026; beef cows down another 1% after years of drought and liquidation.

  2. 2

    Ranchers hold back heifers

    Rebuilding means keeping breeding stock off the market — the first move in any recovery.

  3. 3

    Supply tightens before it loosens

    Fewer cattle reach slaughter now, so the price stays firm through the rebuild.

  4. 4

    New production is a 2028 story

    A heifer retained today calves and reaches beef production in late 2028 at the earliest.

  5. 5

    The peak lands in 2027, not now

    Domestic supply, not the Pacific and not a tariff. “Wait it out” is the wrong plan through 2027.

Beef is a domestic supply story on a multi-year clock: the herd is the smallest since 1951, and new production doesn’t arrive until 2028 at the earliest.
Sources: USDA NASS Cattle Inventory and ERS Food Price Outlook, 2026

USDA National Agricultural Statistics Service — Cattle Inventory (January 1, 2026): all cattle and calves 86.2 million head, the smallest since 1951; beef cows 27.6 million, down 1%. nass.usda.gov. USDA Economic Research Service — 2026 beef & veal price forecast about +7.5% (prediction interval 3.1–12.2%), revised down from ~12.1% the prior month; reported retail beef ran +12.9% year-on-year and −1.3% month-over-month in May 2026 (BLS retail data). ers.usda.gov/food-price-outlook. The herd-rebuild timeline (retained heifer → production in late 2028+, price peak likely 2027) reflects standard cattle-cycle biology as described by USDA and land-grant extension analysts. Retrieved 2026-07-16.

The operator move on beef is the opposite of patience. Public wholesale quotes never arrive as your delivered price, so watch your own food-cost percentage on the beef plates line by line. If beef is a top-three cost on a plate, reprice it deliberately, on this year’s reality, not seasonally on a hunch that it’ll pass. It won’t pass in 2026.

Front two: why coffee ran hot, then cooled

Not everything is on fire, and the honest read has to hold the relief as firmly as the pain. Two fronts — coffee and insurance — are measured relief right now, and pretending otherwise would be its own kind of dishonesty.

Coffee ran genuinely hot early in the year. April 2026 had coffee up 18.5% over twelve months, with roasted at 17.3% and instant at 22.8%. Then it turned. By June, coffee fell 2.0% for the month and the broader non-alcoholic beverage index was up 2.9% year-on-year — still positive, no longer climbing. The honest phrase is “still expensive, no longer accelerating,” the same shape as beef but bending the friendlier way.

Coffee ran hot, cooled, then twitched (2026)

  1. Apr Coffee CPI +18.5% year-on-year — the peak
  2. Jun Coffee −2.0% for the month — cooling
  3. late Jun Arabica futures near $2.70/lb — the trough
  4. mid-Jul Futures back to ~$3.30–3.50/lb — the twitch
Coffee is the clearest easing front and the least finished: down at the retail counter, back up on the futures board.
Sources: BLS coffee CPI, ICE arabica futures, USDA FAS on the Brazil crop

U.S. Bureau of Labor Statistics — coffee CPI +18.5% year-on-year in April 2026 (roasted +17.3%, instant +22.8%); coffee −2.0% month-over-month and non-alcoholic beverages +2.9% year-on-year in June 2026. bls.gov/cpi. ICE arabica futures — roughly $2.70/lb in late June 2026, rebounding to about $3.30–3.50/lb in mid-July on Brazil crop uncertainty (per market data). USDA Foreign Agricultural Service — the elevated 2026 base traces to a Brazil arabica shortfall (2025/26 down about 13%) plus the July 2025 tariff disruption; a record Brazil 2026/27 crop (about 71.9 million bags, +14%) is forecast to ease prices. fas.usda.gov. Futures figures are forecasts of sentiment, not a realized cost. Retrieved 2026-07-16.

The cause underneath is worth stating plainly, because it’s a correlation, not a single villain. Coffee is high because of a real Brazil arabica shortfall — the 2025/26 crop fell about 13% on bad weather — layered on top of a July 2025 tariff that cut Brazilian shipments to the US by more than half. Both are easing: a record Brazil crop is forecast for 2026/27, and the tariff picture changed (more on that next). USDA’s 2026 forecast for non-alcoholic beverages is still +5.7%, above the menu blend, so “easing” is not the same as “cheap.” If coffee is a signature pour, hold your price — don’t lock a permanent number on an input this jumpy.

Insurance is the second, quieter relief. After years of steep renewals, commercial property premiums are forecast to rise only about 3–4% in 2026 as that market softens. The exception is liquor liability, up as much as 20% in some markets, so a bar-heavy concept doesn’t get to bank the same relief a bakery does. More than eight in ten operators still name energy and card fees as strain. Relief in 2026 is real, and it is uneven — you can bank the coffee cool-down, but not evenly across every concept.

Front three: how tariffs actually hit restaurant costs in 2026

Tariffs did more damage in the headline than on most of my invoices, and the honest version splits perception from the line item cleanly. In the NRA’s 2026 survey, 68% of operators said tariffs contributed to higher food and beverage costs. That is a reported perception, not a measured amount — it tells you how the year felt, not that tariffs added a specific percentage to any plate. Never let a vendor turn that 68% into “tariffs added X% to your cost.”

The measured reality moves in two directions at once. The average effective US tariff rate was about 8.9% in April 2026, down from a 16.5% peak after the Supreme Court struck the broadest tariffs, but still the highest since the 1970s. On the up-line: a Section 232 order kept a 50% tariff on steel, aluminum and copper, with 25% on derivative goods, effective June 8, 2026 — and that is the to-go-lid line from the top of this piece, plus cans, kegs, and a chunk of kitchen equipment. On the down-line: the 40% tariff on Brazilian coffee, beef, cocoa and fruit came off in November 2025. One arrow moved packaging up; another moved coffee down, in the same year.

Tariffs did more damage in the headline than on most of my invoices. One line went up, another came down, same year.

Here is why every point still stings even when the arrow is small: restaurant food costs are already up about 34% against pre-pandemic levels, so you’re taking each new tariff on top of a base that never came back down. The move isn’t to fear the word “tariff” — it’s to ask your rep which SKU, in which direction. If it’s a can, a lid, or a piece of equipment, that’s a real tariff line you can plan around. If it’s coffee, the tariff went the other way.

Sources: NRA survey, Tax Foundation, Federal Register, White House

National Restaurant Association — 2026 State of the Restaurant Industry: 68% of operators said tariffs contributed to higher food and beverage costs; food costs up about 34% vs pre-pandemic. restaurant.org. Tax Foundation — average effective US tariff rate about 8.9% as of April 2026, down from a 16.5% peak, the highest since the 1970s. taxfoundation.org. Federal Register — Section 232 proclamation: 50% on steel, aluminum and copper articles, 25% on derivatives, effective June 8, 2026. federalregister.gov. The White House — executive order removing the 40% duty on Brazilian coffee, beef, cocoa and fruit, effective November 13, 2025. whitehouse.gov. The 68% figure is a survey perception, not a measured cost magnitude. Retrieved 2026-07-16.

Front four: the labor and staffing squeeze underneath every plate

Labor isn’t a front beside the others — it’s the floor under all of them, and I’ll handle every number here as reported and cited, with no claim about policy or cause. Wages set the base cost of making anything on your menu, so when they move, they move with your food cost inside your prime cost.

The backdrop first. Private-sector wages rose 3.5% over the year to June 2026, to an average of $37.64 an hour; front-line food-service work sits far below that, with a median near $14.92. And the job market is cooling, not booming — leisure and hospitality shed about 61,000 jobs in June 2026. A softer hiring market would normally take pressure off wages.

Except the applicant pool is thinning at the same time, and that is the part specific to this year. Foreign-born workers are about 22% of the restaurant workforce and a much larger share of the kitchen — 43 to 46% of chefs, roughly one in three of all chefs and cooks, per industry-association tallies of Census data. Over March to July 2025, an estimated 137,000 immigrant restaurant workers left the workforce, part of about 1.7 million foreign-born workers — a 5% drop — leaving the US labor force in that window. A cooling job market and a thinning kitchen bench, landing on the same schedule, is why the labor line stays firm even as hiring slows. Those figures are reported staffing counts, cited below; the point here is only what they do to your cost of a shift.

Sources: BLS wages and employment; industry and Census workforce tallies

U.S. Bureau of Labor Statistics — total private average hourly earnings +3.5% over the year to $37.64 in June 2026; median wage for food and beverage serving workers about $14.92; leisure and hospitality employment down about 61,000 in June 2026. bls.gov. National / New Mexico Restaurant Association & Census analyses — foreign-born workers about 22% of the restaurant workforce, 43–46% of chefs, roughly one in three chefs and cooks. Documented / Davis Vanguard analyses — an estimated 137,000 immigrant restaurant workers lost from the workforce March–July 2025; about 1.7 million foreign-born workers (−5%) left the US labor force in that window. Figures are reported as published. Retrieved 2026-07-16.

Front five: why credit-card swipe fees are a top-three restaurant expense

The quietest front is the one that grew the most while everyone watched food and labor: card swipe fees. US merchants paid a record $198.25 billion in swipe fees in 2025, up from $187.2 billion in 2024, with the credit-card portion alone at $157.8 billion. The NRA ranks card processing as the third-largest operating expense for a majority of restaurants, behind only food and labor, and it has climbed more than 80% since the pandemic. Two in three operators say their fees rose in the past two years.

The reason it only ever ratchets is structural. The average Visa and Mastercard credit interchange rate — the biggest slice of a swipe fee — reached 2.36% of a purchase in 2025, up from 2.02% in 2010. It behaves like a commission on every sale you make, food and tip and tax alike, and it compounds as your prices rise: raise the check to cover beef, and the swipe fee on that check goes up too.

Credit interchange per $100 of card sales

2010

$2.02

2025

$2.36

After 2026 settlement

~$2.26

That is the average credit interchange rate — 2.36% of a sale in 2025, up from 2.02% in 2010. US merchants paid a record $198.25 billion in swipe fees in 2025; the 2026 settlement trims about 10 basis points for five years — real, but a trim against fifteen years of climbing.

Per $100 of card sales, credit interchange (the biggest slice of a swipe fee) climbed from $2.02 to $2.36 over fifteen years; the 2026 settlement trims it about a dime.
Sources: Nilson Report, Merchants Payments Coalition, court filings

Nilson Report / Merchants Payments Coalition — US merchants paid a record $198.25 billion in card swipe fees in 2025 (up from $187.2 billion in 2024); credit-card portion $157.8 billion; average Visa/Mastercard credit interchange 2.36% in 2025 vs 2.02% in 2010. merchantspaymentscoalition.com. National Restaurant Association — card processing is the third-largest operating expense for most restaurants, up more than 80% since the pandemic; 66% of operators say fees rose in the past two years. restaurant.org. Court filings (reported) — a $38 billion Visa/Mastercard settlement won preliminary approval in June 2026, cutting average interchange about 10 basis points for five years across 12 million-plus merchants; an earlier ~$30 billion deal was rejected in 2024. Retrieved 2026-07-16.

Relief is coming, and it is partial. A $38 billion Visa and Mastercard settlement won preliminary approval in June 2026; it would trim the average interchange rate about 10 basis points for five years across more than 12 million merchants. Ten basis points against a 2.36% base is a trim, not a rescue — the earlier, larger deal was thrown out in 2024 as too small to matter. Meanwhile roughly a third of merchants now add a surcharge, up from one in five two years ago. Read this line every month the way you read food cost, and model the surcharge-versus-absorb question deliberately, because it trades a fee against your guest’s goodwill.

How restaurants are actually responding to higher costs

The evidence on what operators do is clear, and it carries a warning inside it. Of the 82% who paid more for food in 2025, the National Restaurant Association found the most common move wasn’t the menu price — it was the phone call to the supplier. 55% renegotiated with existing suppliers, 47% cut costs elsewhere, 43% adjusted portion sizes, 37% tightened waste tracking, and 31% substituted lower-cost ingredients. The operators who moved first mostly moved on cost, not on price.

How operators with higher food costs responded (share of those operators)

Renegotiated suppliers

55%

Cut costs elsewhere

47%

Adjusted portion sizes

43%

Tracked waste more

37%

Substituted cheaper items

31%

Separately, 71% plan a menu-price raise in 2026, up from 57% a year earlier — but a price hike into softer traffic is the trap the average sets.

Renegotiating suppliers beat raising the menu: the operators who moved first moved on cost, not price.
Sources: NRA 2026 State of the Industry; Popmenu operator survey

National Restaurant Association — 2026 State of the Restaurant Industry: 82% of operators had higher food costs in 2025; of those, 55% renegotiated suppliers, 47% cut costs elsewhere, 43% adjusted portions, 37% increased waste tracking, 31% substituted lower-cost items. restaurant.org. Popmenu — survey of 328 US operators (January 2026): 71% plan to raise menu prices in 2026 (up from 57% a year earlier); 31% considering variable/dynamic pricing. Retrieved 2026-07-16.

Here is the warning. The field’s default forward move is the price hike — 71% plan one this year, up from 57% — and the field’s central problem is that traffic already softened for about 60% of operators. Raising the check while the guest count falls is exactly the trap the tame average sets. Portion, mix, waste and supplier terms move margin without spending the contribution margin you get from a full room. Benchmark what you pay against the market before you touch the menu — Muntin’s free Vendor Benchmark is built for exactly that conversation with a rep.

How to read your own invoices instead of the headline

The average is not your menu, and it is certainly not your P&L. So the close is a method, not a mood. Pull your top-five plate costs, and mark each front’s real direction against them: beef structural and rising through 2027, coffee cooling but jumpy, packaging and equipment up on Section 232, card fees up-then-trimmed, labor firm. Weight those by your menu mix, not the government’s basket, and the storm resolves into a short list of things that actually touch your plates.

  1. 1Is beef a top-three cost on any plate?

    The one structural, multi-year front — a 75-year-low herd, not a passing spike.

    Yes Treat it as durable. The cycle peaks in 2027 and relief is a 2028 story. Reprice deliberately, not seasonally — “wait it out” loses money through 2027.

  2. 2Is coffee a signature pour?

    Easing at retail, but not settled — the futures board is still jumpy.

    Hold Retail cooled (−2.0% for the month) but futures spiked back to $3.30–3.50. Don’t lock a permanent price on an input this volatile.

  3. 3Do cards carry most of your sales?

    The quiet third-largest expense, and it only ratchets.

    Structural The 2026 settlement trims about 10 basis points and no more. Read the line monthly and model surcharge vs absorb against your guest’s goodwill.

  4. 4Do you import coffee, cocoa, or fruit?

    The one place the tariff arrow pointed down in the last year.

    Relief The Brazil 40% duty came off in November 2025 — a cost cut, not an increase. Re-check your landed price and make sure the reduction reached your invoice.

Sort each front onto your own board — structural squeeze, easing spike, or outright relief — before you touch a single price.
Sources: composite of the figures cited throughout this dispatch

This diagnostic composes the sourced figures above: USDA ERS/NASS (beef, herd), BLS (coffee, wages), the White House (Brazil 40% duty removed November 13, 2025), and the Visa/Mastercard settlement (preliminary approval June 2026). Each front’s direction is measured or reported; the “for your menu” verdict is an operator judgment to apply against your own invoices, not a sourced prediction. Retrieved 2026-07-16.

The two-pile sort

Put each front in one of two piles. Durable signal: beef, the swipe fee, the 42% profit gap, restaurants inflating faster than grocery — plan around these. Spikes easing: coffee, the Brazil tariff removal, the $38 billion card settlement, softening property premiums — bank these, but don’t over-reprice on them. Anything a headline calls a “crisis” belongs in one pile or the other, never both.

What isn’t on the CPI table is your own paperwork, and that is the only weather report that pays my rent. Read the weekly Cost Index against each ingredient’s own baseline window; price the plate in front of you against today’s invoice with plate-cost; run the margin math before you set a new number; and watch the cost pulse week over week so a line like the to-go lids never sneaks up on you again. The storm was never one wind. It was five, at five speeds, and the only forecast that ever changed a price on my menu was the one printed on my own invoices.

Frequently asked questions

How much have restaurant prices gone up in 2026? Restaurant menu prices — what the government calls food away from home — rose 3.4% over the 12 months ending June 2026, with full-service at 3.7% and fast food at 3.1%. That’s the tamest menu inflation in years, and USDA’s outlook has restaurants rising about 3.6% for the year against grocery’s 2.8%. The catch: the calm average hides much faster-moving lines underneath, like retail beef at 12.9% year-on-year.

Why are so many restaurants not profitable in 2026? In the National Restaurant Association’s 2026 read, 42% of operators said their restaurant wasn’t profitable last year, and about 60% reported softer customer traffic even as sales dollars grew — revenue carried by higher checks, not more guests. With more than nine in ten operators naming food, labor, insurance, energy and swipe fees as significant challenges at once, the squeeze is on the whole P&L, not one line.

Why is beef so expensive for restaurants in 2026? Beef is a domestic cattle-cycle story, not tariffs or weather. The US cattle herd on January 1, 2026 was 86.2 million head, the smallest since 1951, after years of drought and herd liquidation. Retail beef ran 12.9% higher year-on-year in May 2026, and USDA forecasts 2026 beef and veal up about 7.5% (within a wide 3.1–12.2% band). Because rebuilding a herd takes years, the price peak most likely lands in 2027, not this year.

Are tariffs raising restaurant food costs in 2026? In part, and unevenly. 68% of operators said tariffs contributed to higher food and beverage costs — but that’s a reported perception, not a measured amount. The measured picture moves both ways: a Section 232 order kept 50% tariffs on steel, aluminum and copper effective June 8, 2026, raising cans, lids and equipment, while the 40% tariff on Brazilian coffee, beef and fruit came off in November 2025. Ask which SKU and which direction before accepting a tariff surcharge.

Did coffee prices go down in 2026? They cooled. Coffee ran hot early in the year — up 18.5% year-on-year in April 2026 — then turned, falling 2.0% month-on-month in June, with the broader non-alcoholic beverage index up 2.9% year-on-year. Arabica futures dropped to about $2.70 a pound in late June before rebounding to roughly $3.30–3.50 in mid-July. So retail coffee is easing but not settled, and USDA still forecasts non-alcoholic beverages up 5.7% for the year.

Why are credit card swipe fees such a big restaurant expense in 2026? US merchants paid a record $198.25 billion in card swipe fees in 2025, and the NRA ranks card processing as the third-largest operating expense for a majority of restaurants, behind food and labor, up more than 80% since the pandemic. The average credit interchange rate reached 2.36% of a purchase in 2025, up from 2.02% in 2010, and it compounds as prices rise. A $38 billion Visa and Mastercard settlement won preliminary approval in June 2026 but trims only about 10 basis points for five years — real relief, not a reversal.

What can restaurants do about rising costs in 2026? Of operators with higher food costs, the most common move wasn’t the menu price — it was the supplier call: 55% renegotiated with existing suppliers, 47% cut costs elsewhere, 43% adjusted portion sizes, 37% tightened waste tracking, and 31% substituted lower-cost items. 71% plan to raise prices this year, but raising checks into softer traffic is the trap the tame average sets. Sort each front onto your own menu, benchmark what you pay before you reprice, and read your own invoices against the Cost Index and plate-cost, not the headline.

Every number here carries its source in the drawers above — the BLS Consumer Price Index, USDA ERS and NASS, the National Restaurant Association’s 2026 State of the Industry, the Nilson Report and court filings on swipe fees, and the Tax Foundation and Federal Register on tariffs. This is a read of public data as of July 16, 2026, not a forecast of your costs. Check my math.