The numbers you came here for
There are two honest ways to answer “how much should I tip,” and they are different questions. What people actually leave: card and digital tips at full-service restaurants averaged 19.4% in the first quarter of 2025, and 15.8% at quick-service restaurants, according to Toast’s Restaurant Trends Report as analyzed by LendingTree — and those are averages across transactions where a tip was left by card, so they exclude cash and anyone who left nothing. What guidance recommends: at least 18% of the pre-tax bill for sit-down waitstaff, per hospitality researcher Cortney Norris in Bankrate’s tipping guide — and notably different rates elsewhere, from 3-5% as a rideshare floor to 20% at a spa.
That second point is the one most tip charts get wrong. “18-20% for everything” is not what the guidance actually says: it is a band for effort-heavy table service that gets applied, incorrectly, to every other category. Below, each number is labelled with where it came from and what kind of number it is — a measured card average, a survey of what people say they do, or an expert recommendation. Those three are not interchangeable, and a chart that blends them into one column is how “20% everywhere” became folklore.
The reason a single percentage no longer settles it: the check arrives, you glance at the total, and the mental negotiation starts. Is 18% enough? Is 20% expected? Does this place even deserve a tip? In 2026, tipping norms vary by venue, service type, and context — and the confusion is quantifiable.
Bankrate’s 2025 Consumer Tipping Attitudes Survey found that 63% of Americans hold at least one negative view about tipping culture, up from 59% the previous year. A Pew Research Center study found that 72% of Americans say tipping is expected in more places today than it was five years ago. The landscape has changed. These charts reflect what people actually tip, drawn from card and digital tip averages across Toast’s restaurant payment network and cross-referenced with academic research.
Sources: Toast Q1 2025 Restaurant Trends Report (Jan. 1 – Mar. 31), as reported in the LendingTree Tipping Rates Study, 2025. Toast’s averages exclude cash tips and cover only transactions where a tip was left. Also: Bankrate Consumer Tipping Attitudes Survey, 2025; Pew Research Center, 2023
The 2026 tipping chart: sit-down restaurants
Full-service restaurant tipping is the one category where norms are most settled — and most studied. Cornell University’s Michael Lynn published a 2025 analysis in the International Journal of Hospitality Management that pooled average tip percentages from academic studies across several decades. His regression found tip percentages rose by about 1.5 percentage points per decade over the last half century — a trend that held under every robustness check he ran, including one restricted to 1980-2019. The norm has crept upward — though what Lynn measured is average tips paid, which is not the same thing as the norm people believe in.
Sit-down restaurants: every published number, in one place
Two things worth sitting with. First, there is no fine-dining row and no buffet row, because there is no fine-dining or buffet number. Toast reports two restaurant segments — full-service and quick-service — and Bankrate’s etiquette answer names services (“waiting tables”), not price tiers. Any chart that hands you a specific percentage for a white-tablecloth room or a steam table is filling a gap the research left open. Second, look at the last two rows against the first two: what people say they would leave sits well below what the card data says they actually leave. Pew found 57% would tip 15% or less for an average sit-down meal, and only a quarter said 20% or more — while the measured card average is 19.4%. Treat that gap as suggestive, not as a measured finding: the two numbers are built differently. Toast’s average covers only card and digital transactions on which a tip was actually left, so non-tippers and cash tips are outside it entirely, and Pew asked a hypothetical about an average meal of everyone. A survey median and a conditional card average were never going to line up, and some of the distance between them is method rather than behaviour.
Why 15% no longer reads as standard: when the 1975 study further down this page was run, the operative norm was that a tip should equal 15% of the bill. Lynn’s 2025 analysis puts the drift since then at about 1.5 points per decade, and LendingTree’s read of Toast card data is that the full-service average “has hovered between 19% and 20% since at least 2018,” landing at 19.4% in Q1 2025. Worth noting how little the published advice moved by comparison: Lynn’s paper points out that Emily Post’s etiquette books “advocated tipping 15 to 20 percent in restaurants as far back as 1959 and continue to do so today.” Behavior drifted inside the guidance, not past it.
Sources: Lynn, “How have U.S. restaurant tips changed over time?” International Journal of Hospitality Management (2025); LendingTree Tipping Rates Study, 2025; Bankrate Consumer Tipping Attitudes Survey, 2025; Pew Research Center, 2023
Fast-casual, counter service, and the iPad question
This is where tipping gets contentious. You order at a counter. You carry your own food. The iPad flips around and offers you a row of preset percentages. Demi Shenrui Deng (Auburn), Lu Lu (Temple), and Ruiying Cai (Washington State) ran two coffee-shop experiments in 2025 — 320 and 414 participants — published in the International Journal of Hospitality Management. Both found that the tipping request itself produced negative emotional reactions, driving down how deserved the tip felt and how satisfied people were with their own decision. Two things softened it: asking after service rather than before, and making the service effort visible. Note the limits of what that shows: these were scenario-based coffee-shop experiments measuring emotion, deservingness and decision satisfaction — not whether people are willing to tip, and not behaviour outside a coffee counter. What it does suggest is that when the ask happens matters, not just how much is asked. We cover that mechanism in the iPad tipping fatigue guide.
It is also the part of the chart where recommendation and measurement diverge most. Bankrate’s guide does give a counter number — 10-15%, or about $1 on a drip coffee — but it sits well below the restaurant band, and the survey data shows most people ignoring it entirely. Deng, Lu and Cai study how people react to being asked; they never say what to leave. Here is the recommendation alongside what is actually measured.
Counter service & quick-service: what is measured
Read those rows together and the interesting gap is participation, not rate. Among transactions that carried a card tip in Q1 2025, the counter average was 15.8% against 19.4% at full-service — lower, but not the token amount many people assume. Separately, Toast’s Q1 2024 figures show 49% of quick-service customers tipping by card against 75% at full-service. Different quarters, so read them as two facts rather than one trend: the counter gap looks larger in who tips than in how much they leave.
The frequency data points the same way. Bankrate found that 18% of Americans always tip coffee shop baristas, down from 20% the previous year and 23% in 2021, and that 12% always tip when picking up takeout. Pew puts the number who always or often tip when buying a coffee at 25%, and at a fast-casual restaurant at just 12%. Be careful how far you push that, though: baristas are the category that fell, while takeout actually ticked up from 11%, and most of Bankrate’s other categories held or rose. What the numbers support is a retreat at the coffee counter specifically — not a broad collapse in counter-service tipping. For a deeper look at the psychology behind those iPad prompts, see our counter service tipping guide.
Sources: Deng, Lu & Cai, “Rethinking tipping request,” International Journal of Hospitality Management (2025); Toast Q1 2025 and Q1 2024 Restaurant Trends Reports via LendingTree; Bankrate Consumer Tipping Attitudes Survey, 2025; Pew Research Center, 2023
Delivery tipping: DoorDash, Uber Eats, and everything else
Delivery is one of the places Americans are least sure of the rules. Bankrate opens its 2025 survey with exactly this question — “Do you tip the DoorDasher who picked up your food order and dropped it at your door?” — and answers it only obliquely: delivering food is one of the five services its etiquette guidance names explicitly, at 18 to 20 percent. Its fuller tipping guide is more specific, and more useful, citing Grubhub’s own advice: $5 or 20%, whichever is greater. That is a floor and a percentage, which matters on small orders.
What none of these sources describe is how the delivery platforms themselves handle tips — whether a driver sees your tip amount before accepting, how offers get routed, what a small tip does to your wait. Those mechanics are discussed constantly online and are not measured in any survey, POS dataset, or peer-reviewed paper behind this page. Treat what you read about them as claim, not finding.
Delivery
The gap between those middle two rows is the whole delivery debate in miniature. Just over half of Americans told Bankrate they always tip a delivery person; roughly three-quarters told Pew they do it always or often. Different surveys in different years, so do not read the gap as a precise headcount — but between “always” and “always or often” sits the group deciding order by order, and that is the group every argument about delivery tipping is actually about.
The small-order problem, and what nobody has measured: a percentage tip scales with the check, not with the trip. The same 18-20% that reads as generous on a family dinner is a fraction of that on a single sandwich — same drive, same mileage, same driver. You will see a “$5 minimum” quoted widely online. It is not in any survey we could source: Bankrate switches to flat dollar amounts only where percentages are ambiguous (“$10 per hour or $5 per bag carried or shuttle driven”), and delivery is not on that list. Treat the floor as your own judgment, not an industry standard. For more on how delivery tips work with group orders, see our Uber Eats tipping guide.
Sources: Bankrate Consumer Tipping Attitudes Survey, 2025; Pew Research Center, 2023
The group tipping problem
Tipping gets harder in groups. Not because people are cheap — because of a proposed psychological mechanism called diffusion of responsibility. In a 1975 study published in Personality and Social Psychology Bulletin, Freeman, Walker, Borden, and Latane measured the tips left by 396 groups of restaurant diners. One-third of the variability was explained by a single rule: tip should equal 15% of the bill. The interesting part was the leftover. Variation around that norm tracked group size as an inverse power function, which the paper writes as 18% / N.22.
Read that carefully, because it is easy to over-read. The expression describes the variation around the 15% norm — it is not a formula that predicts what a party of six leaves, and the paper does not present it as one. What the field took from the study is directional, and Elman’s 1976 reply in the same journal states the direction plainly: “the percentage amount of a restaurant tip decreases as the number of persons at a table increases.” Elman then questioned the explanation, arguing that diffusion of responsibility may not be the cause and suggesting “theoretical and methodological alternatives.” Treat the group-size effect as a real but contested finding rather than settled fact: two papers from the 1970s do not close a literature, and Azar’s later review of the field describes the evidence on group size as mixed. What is safe to say is that the association was measured, and that its explanation was disputed almost immediately.
What the 1975 study actually reported
396 groups of restaurant diners
One-third of tipping variability explained by: tip = 15% of bill
Variation around that norm: an inverse power function of group size,
written 18% / N.22
Association: observed in 1975. Cause: contested since 1976.
“The amount tipped by 396 groups of restaurant diners was a function of the number of people eating together as well as the size of the bill.
Freeman, Walker, Borden & Latane, Personality and Social Psychology Bulletin (1975)
Restaurants that add an automatic gratuity to large parties are addressing the same arithmetic, though no source here studied why the practice was adopted — the alignment is an observation, not a documented motive. Diners are not fond of it: Pew found that about seven-in-ten adults (72%) oppose businesses including automatic service charges or tips on customers’ bills, regardless of group size, including half who strongly oppose it. Only 10% favor the practice. If your group is splitting a bill and calculating tips, the math compounds: who tips on their portion? On the pre-tax total? On the post-tax total? Each person’s mental accounting produces a different number. For more on how group size affects your bill, see our large group splitting guide.
Sources: Freeman, Walker, Borden & Latane, “Diffusion of Responsibility and Restaurant Tipping: Cheaper by the Bunch,” Personality and Social Psychology Bulletin (1975); Elman, “Why Is Tipping ‘Cheaper by the Bunch’: Diffusion or Just Deserts?” Personality and Social Psychology Bulletin (1976); Pew Research Center, 2023
Beyond restaurants: services you might forget
Tipping extends well past dining, and this is where the “18-20% for everything” shorthand falls apart. Bankrate’s tipping guide gives genuinely different rates by category — some of them far below the restaurant band — drawing on etiquette expert Elaine Swann and hospitality researcher Cortney Norris. These are recommendations, not measurements: no card dataset backs them the way Toast backs the restaurant numbers.
Published guidance by service (Bankrate)
Recommendations are one thing; what people do is another. Bankrate also asked Americans which services they always tip, and the answer fractures by category far more sharply than any percentage chart suggests. Note these are shares among people who use each service.
Share who always tip, among users of each service (Bankrate)
Note the shape of that list. It is not a generosity ranking — it is a map of where the norm is settled. Seventy percent always tip a server; nine percent always tip a plumber. In between sits every category people argue about.
Source: Bankrate Consumer Tipping Attitudes Survey, 2025
Tipping varies by state: a 5-point gap
Geography matters. LendingTree’s 2025 analysis of Toast POS data found that the gap between the most and least generous states is nearly 5 percentage points at full-service restaurants. Delaware leads at 22.6%. California trails at 17.8%. That difference on a $100 tab is $4.80 — meaningful money for servers relying on tips.
Delaware: 22.6% full-service
West Virginia: 21.0% overall
New Hampshire: 20.9% overall
California: 17.8% full-service
Washington: 17.8% overall
Nevada: 18.2% overall
The pattern lines up with how states pay tipped workers. Delaware lets employers pay a tipped cash wage as low as $2.23/hour against a $15.00 minimum — and it tips the most. California, Washington and Nevada all require the full state minimum before tips (California’s is $16.90/hour) — and all three sit at the bottom of the tipping table. That is a consistent correlation across the extremes, not a proven cause: nobody has shown that individual diners know their state’s tip-credit rules, and state-level differences in cuisine mix, cost of living, and card-versus-cash habits are not controlled for here.
Federal tax treatment of tip income is a separate lever on the same dynamic — and one that has moved recently enough that the summaries in this article’s sources are already out of date, so this page will not restate the thresholds. What matters here is the distinction: tax treatment changes what a tipped worker keeps, not what any diner in this dataset left. The Toast averages above are card-tip totals from the first quarter of 2025; they say nothing about take-home pay. Our no tax on tips guide covers the mechanics and the state-by-state conformity picture.
Sources: LendingTree Tipping Rates Study, Q1 2025; U.S. Department of Labor, Minimum Wages for Tipped Employees; Bankrate Consumer Tipping Attitudes Survey, 2025
The generational tipping gap
The Bankrate 2025 survey revealed a striking age divide: 49% of boomers typically tip at least 20% at sit-down restaurants, compared to just 16% of Gen Z. That is a 3x difference. The same survey shows 84% of boomers always tip at sit-down restaurants, versus 43% of Gen Z. Across all ages, 35% typically tip at least 20% — down from 37% the previous year.
This is not simply “young people are cheap.” Gen Z came of age during the tip-prompt explosion, and the rules were never clear to begin with: Pew found only about a third of Americans say it is extremely or very easy to know whether (34%) or how much (33%) to tip for different services. Bankrate adds that 14% are outright confused about who and how much to tip, up from 11% a year earlier. Nobody handed any generation a chart — but Gen Z is the first to face the question at every counter, checkout, and kiosk.
When every transaction asks for a tip, tipping everywhere becomes impossible. The survey data shows the age gap clearly; it does not explain it. Two possibilities worth holding loosely: younger diners may be triaging a larger number of tip prompts, and Bankrate separately finds the two youngest generations most likely to say unstable income hurts their mental health. Neither source connects either factor to restaurant tipping specifically, so treat both as hypotheses rather than the reason.
Sources: Bankrate Consumer Tipping Attitudes Survey, 2025; Pew Research Center, “Tipping Culture in America,” 2023
Pre-tax or post-tax? The question that starts fights
This is the question that sparks more group-dinner arguments than any tipping percentage: do you tip on the pre-tax total or the post-tax total?
The conventional etiquette answer is the pre-tax subtotal, on the logic that tax is money owed to the government, not a reflection of the meal’s value. Plenty of people tip on the after-tax total anyway — it is the number the eye lands on first. Whether the card reader itself suggests on the pre- or post-tax figure varies by terminal and merchant, so do not assume it matches your own rule. Anchoring is the usual explanation offered for the pull toward the bigger number, though none of the sources behind this article tested tip-base selection directly, so treat it as the standard account rather than a measured finding here.
You do not need a chart to size the gap, because it is an identity: the difference between the two bases is your tip rate applied to the tax line. Tip 20% and you are adding one-fifth of your local sales tax to the bill. What that comes to in dollars depends entirely on your local tax rate — and none of the sources behind this article publishes sales-tax data at all, so there is no national figure here for “the pre-tax versus post-tax difference.” Run your own. (To settle it with actual numbers, the free tip calculator with split computes both bases and the per-person shares in one place.)
The practical answer: the sourced recommendation here is Norris’s — at least 18% of the pre-tax bill. The gap between the two bases is a fraction of your local tax bill, and it matters far less than the difference between tipping and not tipping. If you are splitting a group bill, the real question is whose share the tip is calculated on. That is where proportional tip distribution matters most.
Why we tip what we tip: the research
Michael Lynn is the Michael D. Johnson and Family Professor of Services Marketing at Cornell’s School of Hotel Administration, and he has spent decades on this one question. For the 2025 paper, he pulled 64 observations of average tip percentage out of 56 published articles, theses and dissertations — real transactions, not hypothetical scenarios — to test whether tips have actually risen. They have.
He and Michael Sturman tackled the more obvious question in 2010, in the Journal of Hospitality & Tourism Research: does better service actually earn a bigger tip, or do generous people just rate service generously? They tracked 51 people across multiple dining occasions, so each tipper served as their own control. Tip size still tracked service ratings after holding the tipper constant, which rules out the “generous people rate generously” explanation. It does not prove service causes the bigger tip — a within-person design cannot exclude everything that varies from meal to meal — but the association survives the objection most often raised against it.
So what else is in there? The honest answer is that it is contested, and the shape of the literature says so. Ofer Azar’s 2007 review in the Journal of Applied Social Psychology exists precisely because the research is “spread over various disciplines: mainly psychology, economics, hospitality, and tourism” — no single field owns the explanation. Lynn’s 2025 result speaks to it indirectly. If tipping were only about avoiding disapproval, Azar argued in 2004, tippers would drift just under the norm and the norm would shrink over time. Instead it grew. Lynn frames this carefully — the rising trend provides, in his words, “some evidence” that tippers get something beyond avoided disapproval, and he offers it as support for a widespread belief rather than as proof. His own candidate is positional competition — tips buy better-than-typical service and status, which pushes the average up, which forces everyone else up.
“Tipping in U.S. restaurants alone amounts to $27 billion annually. Tipping is also common in other occupations and countries, making tipping a significant economic activity.
Ofer Azar, Journal of Applied Social Psychology (2007)
That leaves the screen itself. The default percentages a terminal offers are not neutral, and the survey data says people notice: 38% of Americans name pre-entered tip screens as a specific annoyance (up from 34% the year before), 27% say they tip less or not at all when presented with one (up from 25%), and only 11% say they tip more (down from 14%). Pew finds Americans more likely to oppose (40%) than favor (24%) businesses suggesting tip amounts at all. Whatever those defaults are doing, they are not doing it quietly.
Sources: Lynn & Sturman, “Tipping and Service Quality,” Journal of Hospitality & Tourism Research (2010); Lynn, “How have U.S. restaurant tips changed over time?” International Journal of Hospitality Management (2025); Azar, “The Social Norm of Tipping: A Review,” Journal of Applied Social Psychology (2007); Bankrate Consumer Tipping Attitudes Survey, 2025; Pew Research Center, 2023
Why tipping is harder when splitting
Every tipping question above gets harder when the bill is shared. The group size effect, the pre-tax versus post-tax argument, four people anchoring on four different percentages — they all arrive at once when the check lands in the middle of the table. Here is how each finding maps onto a split bill.
The result: everyone at the table pays the same tip rate on their proportional share. The person who ordered a $40 steak tips more than the person who ordered a $14 salad — automatically, without anyone doing mental math at the table.
