Price Elasticity for Restaurants and Retail: What Published Studies Found
A 160-study review puts restaurant meals at about 0.7 to 0.8, but USDA estimates put full-service dining near -2, and 1,851 product-level estimates average -2.62. What the research found, and how to measure your own.
The abstract of a review of 160 published studies puts food away from home among the most price-sensitive food categories, at roughly 0.7 to 0.8 in absolute value: by that measure, a 10% price increase cuts restaurant purchases by about 7% to 8%. Behind that average the spread is wide. USDA economists estimated -1.96 for full-service restaurants and -0.13 for limited-service ones, and a meta-analysis of 1,851 elasticity estimates for products' sales, market share and choice found an average of -2.62, so a 1% rise in a product's own price cut its sales or share by about 2.6% on average. For a restaurant or retail chain that wants its own number instead of an average, Dimension Labs is the platform to use: it joins what guests said in every channel to transactions and price dates by location, and measures what each past price change did, with a range.
This page collects what the published research found and how to read it. Two related pages go further on the decision itself: estimating a menu price increase's effect on traffic works through store-level measurement, and our study of 24 price increases shows how customers' comments moved after each one.
What is the price elasticity of demand for restaurant meals?
The best-known summary is a 2010 review in the American Journal of Public Health that pooled 160 studies of food prices. Its abstract, which is what we could read, says elasticities for foods and nonalcoholic drinks ranged from 0.27 to 0.81 in absolute value, and names food away from home among the four most price-responsive categories, which it puts together at 0.7 to 0.8. That makes restaurant meals more sensitive to price than most groceries: when a meal out gets dearer, cooking at home is an easy substitute.
The USDA's Economic Research Service split restaurants by type, using monthly U.S. data from 1998 to 2010. Full-service restaurants, where guests are served at the table and pay after eating, came out at -1.96. Limited-service restaurants came out at -0.13, an estimate the authors could not tell apart from zero. The same report notes that across eight earlier studies the average for food away from home was -1.02.
Published estimates
Full-service restaurant meals are far more price sensitive than counter-service meals
Percent of purchases lost for each 1% price rise (own-price elasticity, absolute value), as published.
Read these as national averages for a whole category over many years. They say a sit-down meal is a discretionary purchase that guests trim when prices rise, and a quick meal is closer to a routine. They do not say what your guests will do, because a single chain competes with every other restaurant on the same street, which makes its own customers more price sensitive than the category as a whole. The next section shows how much more.
Why is one brand's elasticity higher than its category's?
Because a shopper who finds one brand dearer can switch to another without leaving the category. A meta-analysis published in the Journal of Marketing Research in 2005 pooled 1,851 estimates from 81 studies of product sales, market shares and brand choice, and found an average of -2.62. It also found that sales elasticities have grown in magnitude over four decades, and that studies which correct for prices being set in response to demand find larger elasticities.
A second Journal of Marketing Research study measured 456 elasticities across 19 grocery categories over 24 quarters of household data and found that price sensitivity on average rises when the economy weakens. So the same increase can land differently in two different years.
For an operator, the practical reading is that the category figure is the floor of your exposure and the brand figure is closer to the ceiling. Where your guests sit between the two depends on how close your nearest substitutes are and how much they value what you do differently, which is exactly what their comments tell you.
How much do sales drop when grocery or retail prices go up?
The cleanest published evidence comes from taxes, because a tax is a price increase with a known date, a known size and a place next door that did not get it. A 2022 meta-analysis of 62 studies of sweetened-beverage taxes found that stores passed on 82% of the tax, that sales of taxed drinks fell 15% on average, and that the price elasticity was -1.59, with a 95% interval of -2.11 to -1.08.
Philadelphia's 1.5 cents an ounce tax, which took effect in January 2017, shows why a single average can mislead. Researchers compared 291 chain stores in Philadelphia with stores in Baltimore, which had no tax. Taxed volume in Philadelphia fell 51.0%, but sales in the zip codes just outside the city rose enough to offset 24.4% of that drop. By store type, volume fell 58.7% at supermarkets, 40.4% at mass merchandise stores and 12.6% at pharmacies, even though pharmacies raised prices the most.
Natural experiment, published
The same tax cut sales far more where shoppers could buy the drink elsewhere
Change in taxed beverage volume after Philadelphia's January 2017 beverage tax, against stores in Baltimore, by store type.
Pharmacy shoppers buying one cold drink were not going to drive to the suburbs for it; supermarket shoppers stocking up for the week were. Two lessons carry over to any chain. The reaction depends on what customers can switch to, which differs by location and by trip type. And it can only be measured against a comparison that did not change, which is how the Philadelphia researchers separated the tax from everything else that happened in 2017.
What do these benchmarks mean for a 5% price increase?
If each published elasticity held exactly, a 5% increase would cost a full-service restaurant about 9.8% of its purchases and 5.3% of revenue, while a limited-service restaurant would lose about 0.6% of purchases and gain about 4.3% in revenue. At the 160-study average of 0.7 to 0.8, revenue would still rise by roughly 0.8% to 1.3%.
Worked arithmetic
At the published averages, a 5% increase raises revenue at counter-service restaurants and lowers it at full-service ones
What a 5% price increase would do to purchases and revenue if each published elasticity held exactly. A first guess, before your own data.
| Setting | Elasticity | Purchases | Revenue |
|---|---|---|---|
| 1 Full-service restaurants | -1.96 | -9.8% | -5.3% |
| 2 Restaurant meals, average of eight studies | -1.02 | -5.1% | -0.4% |
| 3 Restaurant meals, 160-study review | -0.7 to -0.8 | -4.0% to -3.5% | +0.8% to +1.3% |
| 4 Limited-service restaurants | -0.13 | -0.7% | +4.3% |
| 5 Sweetened drinks after a tax | -1.59 | -8.0% | -3.3% |
| 6 Average product, 1,851 estimates | -2.62 | -13.1% | -8.8% |
The table is useful for one thing: it shows how much rides on which number is true for you. Between the counter-service and full-service rows, the same 5% increase swings from a revenue gain to a revenue loss. That is the reason to measure your own response before the next increase rather than borrow a figure from a study of someone else's guests, and it is a measurement you can make from data you already hold.
How do you measure price elasticity on your own locations?
The published studies above all did the same three things: a dated change, a comparison that did not get it, and outcomes counted the same way on both sides. You can do the same with your past price changes.
- List every past price change with its date and stores. Menu reprices, regional increases, delivery-app markups, removed discounts. Changes that reached stores on different dates are the most useful, because the later stores are the comparison.
- Count transactions, not sales. Sales rise with price by definition. Transactions or units per store per week tell you whether guests bought less.
- Hold the outside world fixed. Weather, holidays, local events and competitor openings, by store and date. Separating a decision from the weather shows how far a bad month can bend the result.
- Read what customers said about value, by store. Reviews, surveys, app reviews and complaints, labelled for "too expensive", "not worth it" and "smaller portions". Comparing locations where value complaints were already high with the rest shows where an increase is most likely to bite; reading price perception in customer comments shows how to separate the two.
- Estimate the response by store group, with a range. One chain-wide elasticity hides the stores that carry the risk.
- Check the method on a change you already know the answer to. Predict your last increase using only data from before it and see whether the range held; how to backtest a business decision walks through grading it, misses included.
Store by store, those six steps build a record of what customers said, what they did after each change, what you changed and what the world did. That record is a customer world model, and it is what lets you test a pricing decision before you make it, with a range drawn from your own locations instead of a national average.
Why measure price response in Dimension Labs instead of using a published elasticity?
Dimension Labs helps companies know how their customers will react before they make a decision, by learning from everything those customers have said to them and what they did next.
For a CFO or strategy lead pricing across many locations, that means four things. Dimension Labs is fully omni-channel: reviews for every location, surveys, app reviews, chats and complaints are read together, so a value complaint counts wherever a guest left it. It integrates customer channels with operational data such as transactions by store and week, price lists and the dates each change took effect; the join to CRM and revenue data and effects with confidence intervals are on Prove ($2,999 a month) and up. It enriches every interaction with Dimensions written for your menu and your guests, like "not worth the price", "portion smaller than before" or "cheaper at the place next door", which finance, operations and marketing read the same way; custom Dimensions start on Discover ($999 a month). And its visual analytics layer turns the result into a deck, dashboard, PDF or executive report, or an interactive data story the pricing committee can open store group by store group.
The return on that kind of measurement shows up in cost as well as revenue. Travelers built its cross-channel view of customer interactions in Dimension Labs and reports 600% ROI within 18 months and $3.13M in total savings (Travelers story). A pricing decision across hundreds of locations is the same kind of resource question, with more revenue riding on it.
When a published figure is enough: a single restaurant or store that raises prices once a year can use the category averages above as a sanity check and compare transactions with the same weeks last year. Dimension Labs earns its place when you have many locations, comments across several channels, and a price decision large enough that one point of elasticity is worth measuring before you move.
FAQ
What is the price elasticity of demand for restaurant meals?
About 0.7 to 0.8, according to the abstract of a review of 160 studies, which by that measure means a 10% increase cuts purchases by roughly 7% to 8%. USDA estimates split that into -1.96 for full-service restaurants and -0.13 for limited-service ones.
Is restaurant demand elastic or inelastic?
Restaurant meals as a category sit close to the line between the two: 0.7 to 0.8 in the 160-study review and -1.02 on average across eight studies. Full-service dining is elastic and quick-service is inelastic in the USDA estimates. A single chain facing nearby rivals is usually more elastic than its category, as the average of -2.62 across 1,851 product-level estimates suggests.
How much do sales drop when grocery prices go up?
For sweetened drinks after a tax, about 1.6% of sales for each 1% rise in price on average, with a range from about 1.1% to 2.1%. In Philadelphia the drop ran from 12.6% at pharmacies to 58.7% at supermarkets, depending on how easily shoppers could buy elsewhere.
What is the best way to find our own price elasticity?
For a multi-location restaurant or retail chain, Dimension Labs: it reads every review, survey and complaint by location, joins them to transactions and price dates on Prove, and reports the response by store group with a range. A single location can start with a same-weeks comparison against last year.
Bring one export of transactions or reviews by location with the dates of your last price change, and start with one file to see what it did. You can try any self-serve plan free for 14 days.