The Five-Cent Business: Where Your Restaurant Dollar Goes
A meal out costs far more than it did in 2019, and by the industry's own figures, higher menu prices have mostly kept pace with higher costs rather than fattening margins. Where each dollar on the check goes, why the old "nine in ten fail" line is wrong, and why towns with a short season do the math differently.
November 2, 2026
In one sentence
A typical independent restaurant keeps about a nickel of each dollar it takes in, before taxes, so its prices have to rise about as fast as its costs just to keep that nickel; since 2019 they have risen a little faster than the industry's estimate of its costs, while many restaurants serve fewer customers.
New articles and charts arrive in one short email on Monday mornings. Get the weekly digest
The analogy
A household that brings in $100 a week, spends $95 and saves $5. Then every bill goes up by a third: groceries, rent, the power bill, the car. To keep saving the same share, income has to rise by a third too. If it does, the household saves about $6.67 a week, which buys what $5 used to. It isn't richer. It's back where it started. That is roughly where the typical restaurant has been since 2019.
Most figures below are as of late October 2026; the restaurant-price and restaurant-job figures stay current.
Where the dollar goes
Before the pandemic, the National Restaurant Association, the industry's trade group, modeled a typical independent restaurant splitting each dollar of sales like this (National Restaurant Association):
- Food and beverages: 33 cents
- Labor: 33 cents
- Everything else: 29 cents. That is rent and other occupancy costs, utilities, supplies, repairs, administration and the fees on every card swipe.
- Left over, before income taxes: 5 cents
Where a restaurant dollar goes
On $1.5 million a year in sales, that nickel is $75,000 (National Restaurant Association). It is a model, not a census: fast-food counters, full-service dining rooms and bars split the dollar differently, and any one restaurant can sit far from the average.
What happened to each slice
By the association's count, average hourly earnings of restaurant workers have risen 41% since before the pandemic and wholesale food prices 35%, and utilities, occupancy, supplies and card fees have each risen by double digits. Altogether, the typical restaurant's costs rose about 36% from 2019 to 2026 (National Restaurant Association, July 2026).
Run the model restaurant through that. With 2019's sales and today's costs, it would lose about $432,600 a year, nearly 29% of its sales. To break even it needs sales 29% above 2019's, and to get its 5% margin back, 36% above. The association notes that menu prices rose 36% from February 2020 to May 2026, by the Bureau of Labor Statistics' count, about what it takes, as long as a restaurant serves as many customers as before. It also says customer traffic has softened across much of the industry, and fewer customers at higher prices means sales grow more slowly than prices (National Restaurant Association).
The same BLS index of restaurant prices (food away from home) rose 38% from February 2020 to October 2026. Over the last twelve months alone, it rose 3.8%. Measured from the 2019 average, restaurant prices are up 41%, against 32% for everything the consumer price index covers. That is a few points more than the association's 36% estimate of costs over the same years. Restaurant prices have outrun prices in general largely because restaurants' two biggest costs, wages and food, did too.
Standing still is the point. A restaurant whose prices and costs both rose 36%, serving as many customers as it did in 2019, takes in 36% more dollars and keeps 36% more dollars of profit, and those dollars buy about what the old ones did (real vs. nominal). Even that wasn't universal: 42% of operators told the association their restaurant wasn't profitable in 2025 (National Restaurant Association). And there is a ceiling. In the James Beard Foundation's survey of over 380 independent restaurant owners, chefs and operators, restaurants that raised menu prices more than 10% in 2025 were the most likely to report lower profits and expect fewer customers, and many operators said customers had reached the limit of what they would or could pay (James Beard Foundation).
Why don't menus come back down when the price of eggs does? Because food is only about a third of a restaurant's costs, and most of the rest is wages, rent and other bills that almost never fall. A cheaper case of eggs moves a small slice of one line. Prices that stop rising are not prices that go back (rate vs. level).
Why a nickel is fragile
A thin margin magnifies small changes. Two illustrations using the model's split (arithmetic, not data):
- If food rises from 33 to 35 cents of each dollar and the menu doesn't change, the nickel becomes three cents. Two cents of extra cost takes 40% of the profit.
- If a tenth of the customers stop coming, the food bill falls with them, but the rent doesn't, and the staff can only be cut so far. Suppose half of the labor bill can shrink in step with sales and the other half can't. The restaurant now takes in 90 cents for every dollar it used to, and spends about 90 cents. The whole profit is gone.
The same leverage runs the other way, which is why a busy month can carry a slow one. Online ordering and delivery haven't eased it for many: in the James Beard survey, more than four in ten operators who added them reported lower profits, and the foundation's Anne McBride pointed to platform fees and packaging costs (Axios).
The 90% that never was
"Nine out of ten restaurants fail in their first year" is one of the most repeated numbers in the business, and no one has found data behind it. An American Express ad repeated it on network television in 2003; when the company was asked for its data, it said in writing that it couldn't provide any. Researchers led by H. G. Parsa at Ohio State also used a simple simulation to show the figure couldn't fit an industry that keeps growing: at that rate, the number of restaurants in a market would shrink year after year (Parsa and others, 2005).
Measured rates are far lower:
- About a quarter. Following every licensed restaurant in Columbus, Ohio from 1996 to 1999, the same study found that 26% changed hands or closed in the first year and about 60% within three (Parsa and others). That counts a sale to a new owner as a failure, so it overstates closures.
- 17%. Using twenty years of non-public Bureau of Labor Statistics records on 81,000 full-service restaurants in the western states, a BLS researcher and a UC Berkeley statistician found that 17% of independent startups failed in their first year, against 19% for other service businesses. The typical restaurant lasted about 4.5 years, slightly longer than other service businesses; the smallest, with five or fewer employees, about 3.75 years, slightly shorter than service businesses that size (Luo and Stark, 2014).
One likely reason the myth survives is that restaurant closings happen in public. A closed restaurant is a dark window on a main street that everyone walks past; a closed accounting office isn't.
Opening one, briefly
Most of a new restaurant's money goes in before the first customer does: the kitchen, the ventilation hood, plumbing and grease traps, permits, furniture, and rent during construction. And a lease commits the owner for years, whether the doors are open or not.
That explains two things diners notice. One is why a struggling restaurant keeps going. As long as each week's sales cover the costs that stop when the doors close (food, most wages, the utility bill), staying open loses less than closing, because the rent and the loan payments are owed either way. Economists call it the shutdown decision. The other is why a closed restaurant so often reopens as another restaurant: the expensive kitchen stays behind, so the next tenant starts for less, and the turnover happens in plain sight.
Towns with a short season
All of this gets sharper where the year has a season. A Cape Cod business owner told the Federal Reserve Bank of Boston that restaurants among his clients may employ 100 people in season and 25 or 30 off it (Federal Reserve Bank of Boston). In desert resorts like Palm Springs the busy months run from fall through spring, peaking January to April, and summer is the slow season (Fodor's). Ski towns have their own version, with a winter peak and slow springs and falls. The rent, the insurance and the loan payments run all twelve months; most of the money comes in over a few.
That makes the shutdown decision a yearly one. Closing for the slow months saves food, wages and utilities. It doesn't save rent, and it costs something that never shows on the month's books: the staff. A Martha's Vineyard owner put it plainly: closing for the winter would be cheaper, but reopening would mean starting with a whole new crew (The Martha's Vineyard Times). The owners of a Provincetown restaurant said in 2020 that they kept it open all year, at a loss for several months, because the skilled chefs and managers they recruit need year-round work, and because closing for a month or two also costs money, in unused inventory and building upkeep (The Provincetown Independent). Around Palm Springs, where many restaurants have long closed for the hottest months, two restaurants said this year that they stay open through the summer, both citing their staff (NBC Palm Springs, NBC Palm Springs).
Short-season towns also feel a drop in visitors harder, because a lost month can't be made up in another. The number of Canadian visitors to California fell by just over 18% in 2025 (Los Angeles Times), a hit to the winter resorts that count on Canadian snowbirds.
How much of the dollar goes to labor depends partly on where the restaurant is.
Tipped wages. Federal law lets employers pay tipped workers as little as $2.13 an hour in cash, as long as tips bring them up to the $7.25 federal minimum. Seven states, California among them, require the full state minimum before tips (Montana allows a lower wage at its smallest businesses), and others set cash wages somewhere in between. California's minimum is $16.90 an hour in 2026 (U.S. Department of Labor) and rises to $17.40 on January 1, 2027 (California Department of Industrial Relations). The same plate of food carries a very different labor cost from one state to the next.
California's fast-food wage. Since April 2024, workers at limited-service restaurants in chains with at least 60 locations nationwide have been owed at least $20 an hour, with some exceptions (California Department of Industrial Relations). A state Fast Food Council may raise it each year by up to 3.5% or the rise in the CPI-W, whichever is smaller (Fisher Phillips). As of May 2026 the council had no chair and hadn't met in more than a year (CalMatters), and the state still lists the wage at $20 (California Department of Industrial Relations).
Service charges. California's hidden-fee law, passed in 2023 and in effect since July 2024, requires advertised prices to include mandatory charges, but a follow-up law, SB 1524, lets restaurants keep a separate service charge or surcharge if the menu shows it clearly and says what it's for (California Legislature, Fox Business). Restaurants have used such fees partly to narrow the pay gap between servers and kitchen staff, the employment law firm Littler notes (JD Supra).
How it reaches your table
- The check. Restaurant prices have risen faster than prices overall, largely because restaurants' biggest costs did. A meal out is one of the clearest places to see what wages and food cost now.
- The tip. A tip figured as a percentage rose with menu prices on its own. The same 20% of a bigger check is more money, though nobody changed the percentage.
- The fine print. A service charge is part of the price wherever it sits on the check. In California it has to be on the menu.
- The jobs. Restaurants, bars and other food-service places had 12.4 million jobs in October 2026 (BLS), about 1% more than in February 2020, just before the pandemic.
Where the analogy breaks
A household can't set its own pay. A restaurant sets its prices, but its customers decide whether to come back at those prices, which is the ceiling operators described to the James Beard Foundation. And a household's bills don't fall in a slow month. A restaurant's food bill does, its rent doesn't, and that mix is what makes a five-cent margin so sensitive to the number of people through the door.
What people argue about
Whether restaurant prices have outrun costs. Three senators wrote to McDonald's in October 2024 that its price increases may be outstripping inflation, citing its own reports: operating expenses up 16.5% from 2020 to 2023 while its net profit margin rose from about 26% to 32%. McDonald's USA's president had said that May that the chain's average menu prices were up about 40% since 2019, in line with about a 40% rise in its costs for wages, ingredients and paper, and the company replied that the letter showed a lack of understanding of its franchise business model (NBC News). Big chains' margins are public in their filings. For the typical independent restaurant, most of what's known about margins comes from the industry's own surveys, like the association's model above. A chain's filings and a trade group's model each tell part of the story.
What California's $20 wage did. Economists Michael Reich and Denis Sosinskiy, using payroll records, job postings and phone-location counts of workers in restaurants, found that pay at covered chains rose about 11%, employment didn't fall, and prices rose about 1.5%, about six cents on a $4 item (UC Berkeley working paper). Economists Jeffrey Clemens, Olivia Edwards, Jonathan Meer and Joshua Nguyen, using government employment records, found that two years in, employment at California's limited-service restaurants was about 3.5% lower than it would otherwise have been (NBER working paper). The two teams use different data and different comparisons.
See also
- Rate vs. level: why menu prices don't go back down.
- Real vs. nominal: why a bigger profit in dollars can be the same profit.
- Why everything costs 30% more than it did in 2019: the rest of the receipt.
- You run a small business: the same squeeze, outside the kitchen.
- Hourly pay: what the wage figures measure.
Sources
- National Restaurant Association, "Elevated costs continue to pressure restaurant profitability" (July 8, 2026): the model restaurant, cost increases since 2019, the 42% not profitable in 2025, traffic, and menu prices from February 2020 to May 2026. The association is the industry's trade group.
- FRED: CPI, food away from home (CUSR0000SEFV), CPI, all items (CPIAUCSL) and all employees, food services and drinking places (CES7072200001), from the Bureau of Labor Statistics.
- James Beard Foundation, 2026 Independent Restaurant Industry Report, with Deloitte, and its announcement (February 23, 2026); Axios on its online ordering and delivery findings.
- Parsa, Self, Njite and King, "Why Restaurants Fail" (Cornell Hotel and Restaurant Administration Quarterly, 2005).
- Luo and Stark, "Only the Bad Die Young: Restaurant Mortality in the Western US" (2014).
- Federal Reserve Bank of Boston, "The Seasonality Challenge"; The Martha's Vineyard Times (2011); The Provincetown Independent (2020); NBC Palm Springs (June 18, 2026) and NBC Palm Springs (September 8, 2026); Fodor's on the desert season; Los Angeles Times (February 18, 2026), citing Visit California, on Canadian visitors.
- U.S. Department of Labor, Minimum Wages for Tipped Employees (July 1, 2026); California Department of Industrial Relations (August 13, 2026).
- California's fast-food wage: California Department of Industrial Relations, fast food minimum wage FAQ and minimum wage page, Fisher Phillips, CalMatters (May 18, 2026), Reich and Sosinskiy, UC Berkeley Institute for Research on Labor and Employment working paper (March 31, 2026), and Clemens, Edwards, Meer and Nguyen, NBER Working Paper 34033.
- Service charges: SB 1524 (2024), with Fox Business and Littler, via JD Supra on how it changed SB 478.
- NBC News (October 22, 2024), on the senators' letter and McDonald's figures and reply.