Nonprofit, nonpartisan journalism. Supported by readers.


If someone tells you they know how much restaurant servers make, don’t believe them

REUTERS/Patrick T. Fallon
At the heart of the fight in Minneapolis over the “tip credit” is the question of how much tipped servers actually earn.

The next time someone tells you how much money the average restaurant server makes in Minneapolis, you should probably take it with a grain of salt.

The topic’s gotten a lot of attention as Minneapolis debates whether or not tips should count toward a $15-an-hour citywide minimum wage that looks likely to pass in one form or another (as long as it doesn’t get quashed in St. Paul). And lots of people have answers: some cite a Bureau of Labor Statistics survey at an average of about $10 an hour, and others a Minnesota Restaurant Association-affiliated survey that found tipped workers make an average $28.56 an hour.

The chasm between those numbers is bigger than the $15 value of the minimum wage up for discussion at City Hall, which is to say, not trivial. The fight over the tip credit — whether or not employers will be required to pay all employees, tipped or not, at least the minimum wage if it increases — largely hinges on this issue. The problem: it’s a question experts say is almost impossible to accurately answer.

What people say servers make

Currently, Minnesota is one of seven states whose minimum wage does not include a tip carve out. That means workers who are tipped and workers who aren’t collect the same wage at minimum, $9.50 an hour from large employers and $7.75 from small ones.

Pathway to $15, a campaign by the Minnesota Restaurant Association, advocates changing that in Minneapolis, if the city moves ahead with a $15 an hour minimum wage. The organization argues forcing restaurants to pay servers $15 an hour regardless of the tips they make could hurt restaurants’ bottom lines, and ultimately workers if businesses cut back their hours and raise prices.

In March, Pathway to $15 released the results of a survey of 72 Minneapolis restaurants during the month of October that found tipped workers made an average of $28.56 an hour, working an average 19.1 hours per week.

The survey was sent to more than 500 restaurants in the city — “everybody for whom we had email addresses,” both members and non-members of the Restaurant Association, said Dan McElroy, the executive vice president of the Minnesota Restaurant Association. Responses came in from the likes of Red Cow, Crave, Hola Arepa, Manny’s Steakhouse, Pizza Luce and D. Brian’s Kitchen and Catering.

“We chose October because it didn’t have any paid holidays (where people would be off work and dining in restaurants extra days),” McElroy said. “Patios were largely open, but it was away from the holidays.”

The Restaurant Association asked respondents to record wage and tip data based on what they would report for taxes on a spreadsheet and send it in. McElroy said the group had an actuary look over the methodology.

“I am as confident as I can be that these numbers we are reporting are correct,” he said.

But not everyone agrees.

After the report came out, the Restaurant Opportunities Centers United, a national group that advocates for a flat minimum wage, released a statement rebutting the claims the Restaurant Association made. Their complaints included that the lack of information on the survey’s methodology makes it unclear whether the sample accurately represents the cross-section of Minneapolis restaurants, high end to low end. They also argue that in October, a Minneapolis Restaurant Week causes servers to earn more money.

“It’s clear that servers who work in fine dining restaurants earn high wages, it’s true — the higher the bill, the higher the tips are going to be,” said Teófilo Reyes, the Restaurant Opportunities Centers United’s research director. But he and others doubt that number is representative of an average across the profession, pointing out that many restaurant servers live in poverty.

Reyes said he believes the Bureau of Labor Statistics’ estimates (figures compiled by the National Employment Law Project are often cited), which put Twin Cities-area waiters and waitresses at an average $10.29 an hour, are the best available. In February, Minneapolis Mayor Betsy Hodges wrote a post opposing the tip penalty on the city’s website that also cited the BLS numbers.

But there are problems with the BLS estimate, too.

First, it comes from a survey of the whole wide Minneapolis-St. Paul-Bloomington metropolitan statistical area, which includes 14 Minnesota counties and two in western Wisconsin. The minimum wage debate is focused on a relatively small chunk of that area, Minneapolis, and where a concentration of higher-end restaurants likely makes workers a bit more money than in suburban and rural spots.

“To me (that) sounds low,” for Minneapolis, said Celeste Robinson, with $15 Now, a group that’s been advocating to raise the wage to $15 an hour without a tip carve out. “I’ve worked in the service industry and not in a high-end restaurant … (but) you can make a lot of money some days, (and) might go home with very little in your pocket some days, which is why it’s very important to have a baseline wage.” She also pointed out that some tipped workers aren’t restaurant servers, and could be making less.

Second, the BLS estimates are based on a survey of employers who might not have complete information about what their servers make.

Tips that are left on credit cards are automatically reported to employers, but it’s common for servers not to report all their cash tips. That’s becoming less of an issue as more customers are paying with plastic, but it still creates hurdles for people trying to compile accurate data.

“The reporting of wages and tips is dependent upon the employer both knowing the tips paid to the employee and including that in the questionnaire response,” wrote Cassandra Wirth, a BLS economist, in an email. If a profession is clearly tipped, and it’s obvious the employer didn’t include tips, the BLS might leave a survey response out — but that depends on the bureau’s ability to identify that.

The Census Bureau has data for Minneapolis alone, but it includes “food preparation and serving related occupations,” which would include jobs that aren’t tipped, doesn’t break the number down to an hourly wage, which requires making assumptions about how much time workers spend at their jobs, and relies on workers being able to accurately report how much money they make. Wages for tipped workers can vary from week to week, said Sylvia Allegretto, a labor economist and the co-chair of the Institute for Research on Labor and Employment at the University of California at Berkeley who studies restaurant economics (Allegretto is opposed to tip credits).

SurveyWorkersGeographic areaIncomeData source
Bureau of Labor Statistics (compiled by the National Employment Law Project)Waiters and waitressesMinneapolis-St. Paul statistical area$8.98/hr (median); $10.29/hr (mean)Occupational Employment Statistics survey,  2012-2015
U.S. Census BureauFood preparation and serving related occupationsMinneapolis$12/hr (median)*2015 American Community survey
Pathway to $15Tipped workersMinneapolis$28.56/hr (mean)Voluntary survey emailed to Minneapolis restaurants in October, 2016
*The Census Bureau doesn’t estimate hourly wages. Calculation by Teófilo Reyes, the Restaurant Opportunities Centers United’s research director, assuming 30 hrs/week (the median for servers according to the American Community Survey) and 52 weeks of work per year.

Allegretto argues that talk of averages is somewhat misleading in the first place: first, depending on the type of establishment, the amount tipped workers make can vary wildly.

And second, even for the same server, wages can be volatile from shift to shift — because of the weather, because of the time of day (is it Monday afternoon or Saturday night?) and other things out of a server’s control.

“Are there some workers in the industry who make lavish wages due to their tips? I’m sure there are, but you have to think of the worker working at an all-night truckstop or at a Denny’s or at all the lower-end establishments that exist and how are they making ends meet?” she said.

Data quality issues

All of these issues add up to a murky picture of server wages.

“There are some serious data quality issues that make it hard to know the full story behind tipped workers,” wrote Jacob Vigdor, one of the professors at the University of Washington working on the Seattle Minimum Wage Study, in an email, referring specifically to unreported income.

Asked whether there was a way to get a good estimate of average server wages, Vigdor said he’s come to the conclusion it would a tricky undertaking.

It would either require restaurant owners to open their accounting books — showing any credit card tips that were “cashed out,” (credit card tips that are distributed to the server in cash at the end of the night, which don’t necessarily show up in wage records), or require a methodological survey of waiters and waitresses asking about income not reported on W-2s, which “might be unreliable for several reasons,” he wrote.

Comments (4)

  1. Submitted by Patrick Steele on 05/09/2017 - 09:31 am.

    Numbers Game

    The Minnesota Restaurant Association is a lobbying group for restaurant ownership in the area, so their numbers should be taken with a grain of salt. It’s not just the DOL, as a DEED survey independently found similar numbers. It’s also worth noting that an “average” wage skews higher than the median wage, which is always the case with incomes because they have a hard floor and no ceiling. This can be seen in the DOL’s $10.29/hr mean and $8.98/hr median in the cited survey.

    What’s interesting to me is that restaurant ownership in the area wants you to believe both that their servers are living comfortable, middle-class lives AND that gradually folding in a $15/hr minimum wage over many years is an undue burden that will destroy the industry. They’ve got their own set of numbers compiled by their lobbying group, and they seem ready to go to war with them instead of just doing the right thing and paying their employees a living wage. Then when questioned on why it’s such a burden, they suggest proponents of a living wage don’t understand the industry.

    A tip credit is simply a bad idea. It was a bad idea when Tom Emmer proposed it in his gubernatorial campaign and it’s a bad idea when the restaurant industry is proposing it for the city. Wage theft is already a major issue in the restaurant industry (a 2010–2012 study of 9,000 full-service restaurants found almost 84% committed wage violations) and the tip credit is one of the driving factors. Why Minneapolis would want to open that can of worms is beyond me.

    There are a lot of other low-wage industries out there, but they don’t have dedicated reporters to do ownership’s bidding. I mean, think about how many food writers there are in this town and yet how many years has it been since there was a single reporter on the labor beat? A tip credit is a bad idea, but it’s also getting way more discussion than it’s due thanks to the cozy relationship between restaurant owners and food writers in this town. Thankfully the Mayor and City Council seem to be above this noise, much to the benefit of working people in Minneapolis.

  2. Submitted by John Ferman on 05/14/2017 - 10:45 am.


    For me a tip is a reward for good service. What constitutes good service is not amenable to definition and varies among people. What incentive might a server have if tips cut their wages. So the tip carve out is a bad deal. The only people that benefit from tip carve out are the rstuarant owners.

  3. Submitted by Paul Udstrand on 05/16/2017 - 08:22 am.

    Two things…

    First, it wouldn’t be THAT difficult to produce reliable estimates. You can get basic wage data from the BLS, and and the census and you can break that down to geographic locations. That gives you a median salary of around $11 per hour. The only thing you may not be capturing is tip income, but that could be accomplished by conducting anonymous surveys of servers with a balanced sample. You could find out actual tip receipts, AND find out how accurate income reporting is and adjust the BLS and census data. And remember, “median” means a point at which 50% of the wages are higher, and 50% are lower. “Mean” is an average, which can be influenced by high and low samples. So if you have sample of say 30 workers at 30 different restaurants, and three or four workers who get really high tips can skew the average dramitically. For instance with sample of 30, if you have five servers who get $200 worth of tips, fifteen servers that earn $30, and ten servers that earn $20, those high end tips will double the average ($51.00) even though they only represent 16% of the sample. If you take those high end tips out you get an average of $22.

    I’m making these numbers up to illustrate a point but you can see that the method Pathway to $15 used is ridiculous. There’s no denominator, for instance we don’t need to know how many restaurants were surveyed, we need to know how many employees were counted. The sample itself is suspect because of the characteristics of the Twin Cities restaurant environment. In the Twin Cities a large number of restaurants are owned by a small number of individuals, it’s entirely possible that of the 75 restaurants that responded, we’re only looking at 10 to 15 different owners. And we have absolutely no idea how those owners arrived at the numbers they provided.

    Another problem with the Pathway data is that I noticed it assumes work weeks of less than 20 hours, why is that? Are there no full time servers? THAT denominator drastically affects the average as well, a full time or 40 hour work week would cut their estimate in halve, down to $14 instead of $28. Why are we only looking at part time servers? Presumably there’s a mix of part time and full time but that should be reflected in the analysis. Pathway’s “survey” is obviously not a serious attempt, and the fact that it yield wage estimates almost three times higher than others is probably a result of poor methodology.

    I hate to say, but poor methodology isn’t surprising coming from a group that opposes $15 minimum wages but calls itself: “Pathway to $15”. Kind of like anti-labor legislation described as: “Right to Work” laws.

    The second problem with this whole discussion if far more basic: Why are assuming that servers wages should somehow be capped in the first place? So what if some servers earn $28.00 an hour? Why are we assuming that servers can somehow make too much? We don’t make that assumption about the owners do we? Has anyone done a survey to see how much owners make?

    In the end it’s just impossible to take restaurant owners seriously. They say they can’t afford $15 but they won’t tell us what their revenue is or how much THEY make. The real question isn’t how much servers make, it’s how much MORE they would make if they had a base of $15.00? That would tell us how much more it would cost owners, the difference between what they pay now, and what they would pay… but they never tell us that. Business is business, there’s no law of man or nature that says anyone is entitled to profit, let alone a certain profit, so if owners have to pay employees more that’s the price of doing business. Beyond that we have to remember this isn’t a zero sum calculation, restaurants CAN raise the price of a meal or a drink if they don’t want to absorb the cost of higher wages. While we see a lot of whining about high paid servers, what we don’t see EVER is exactly how much the price of a burger will actually go up? And even if we did, could we trust that estimate given it’s origin?

    Federal studies on $15 minimum wages have repeatedly shown that it would produce zero or near zero inflationary pressure on the over-all economy, and negligible inflation in affected sectors like the restaurant industry.

  4. Submitted by Paul Udstrand on 05/15/2017 - 11:32 am.


    I forgot to mention that it’s problematic to compare means and medians, (as is done in this article) they’re different calculations. BLS and the census use a median instead of an “average” for good reasons (as I explained above).

    At any rate what you’d really want to see if you want to get your head around this is a scatter plot that would show you where a majority of incomes cluster. More than likely there is a “Mode” (the most frequently occurring number) or set of numbers clustered around the median somewhere, so that would tell that for instance that 15 of 30 servers (50%) make between $10 and $14 per hour. I’m making that up as an example, but you can see how that provides more useful information than a single number like an average or a median. Or you could break you sample down into deciles or quartiles or something.

Leave a Reply