Why Does Tipping Still Exist?

Time for a riddle. Name something that many people say they dislike, and yet they do it all the time. Something that many people consider a tax, and others see as a form of altruism. Something that almost always happens in many circumstances and yet, in many similar circumstances, never happens.

Danny MEYER: Everybody, practically, leaves a tip in a full-service restaurant. Nobody leaves a tip when they go to McDonald’s.

Today on Freakonomics Radio: we wade back into the tipping wars. We discuss the lessons learned when one gigantic company that didn’t used to have tipping changed its mind.

John LIST: There’s a lot of social pressure to give a tip in that situation.

What happens when a famous restaurateur goes the opposite direction, and gets rid of tipping.

MEYER: I think the biggest thing we’ve learned is that this is really tough.

And why tipping — as controversial as it may be — isn’t likely to go away.

Michael LYNN: They’re wrong when they say that, but they believe it.

*      *      *

Let’s begin with some history. Courtesy of John List, an economist at the University of Chicago.

LIST: Some people argue that tipping goes all the way back to the Roman era. And some argue that it was the 17th century in England. But either way, tipping ended up coming to the U.S., roughly, in the 18th century. And there was actually a lot of resistance. So you had people like Mark Twain saying, “we pay that tax knowing it to be unjust and an extortion.” And you had The New York Times, in 1897, writing that tipping was the “vilest of imported vices.” And even in 1915, in America, there were actually six states that abolished tipping. And they abolished it because they viewed it sort of like Twain did, as kind of a social-pressure way to extort money.

Uri GNEEZY: Tipping is one of the most interesting behaviors that we have, right?

And that is Uri Gneezy. He’s an economist at the University of California, San Diego.

GNEEZY: I sometimes tip when I go to my coffee place in the morning. I always tip at the restaurant. I never tip in McDonald’s. So why is it that it’s so different for me?

So you can see why economists would find tipping to be a topic worthy of their inspection.

GNEEZY: It’s really a very strange behavior and we seem to be acting according to some kind of norms. And these norms are very hard to follow.

LIST: The social norms around tipping in the United States — it’s like no other that I’ve ever witnessed.

It’s estimated that tipping in America adds up to at least $40 billion a year. That is larger than the entire health-and-fitness industry; it’s double the annual budget for NASA. Forty billion dollars is a lot of money, especially when you consider that tipping is optional. But of course, in some cases — restaurants, particularly — it’s not really optional. If you try to not leave a tip of at least 15 or 20 percent in a full-service restaurant, you may well be told, quite directly, just how not optional tipping is.

There are more than two-and-a-half million waiters and waitresses in the United States; they are often paid below minimum wage, in which case the bulk of their pay comes from tips. Most of us are accustomed to this. Others think it’s strange that restaurants — rather than paying employees a set wage, as most industries do — instead pass along their labors costs to customers in the form of a quasi-voluntary tax. Yes, it is a way to keep menu prices low — but it’s a weird custom, don’t you think?

And once you look more closely, you’ll see there are a lot of weird things about tipping. Why, for instance, is it now customary for casual restaurants and coffee shops even to ask for tips on takeout orders? Or think about going to a nice hotel. A porter carries your suitcases to the room; it takes about two minutes, and that’s his job — and yet we usually tip him, even though he doesn’t ask. Now think about the person who cleans your hotel room. There’s often a little envelope in the room to leave a tip for the housekeeper. Do you know how often that envelope is used? One study that was conducted in an “upscale, independent” hotel showed that for every 100 nights spent there, tips were left on just five of those nights! And cleaning your room takes a lot longer than the two minutes to carry your bag. It also means dealing with the mess you left behind.

So why does the housekeeper not get a tip, while the porter does? A couple possible answers: the housekeeper is usually a woman and the porter a man; probably more important, she’s invisible; he’s standing right there in front of you. Still, it’s weird. Here’s another weird one: why is it that you’re always expected to tip a taxi driver but, when you use a car-service app like Uber, you’re not? Oh, hang on. That’s how it used to be. But if you use Uber, you’ve noticed that now you are asked, after the trip, if you’d like to tip. Would you like to know one of the people who made that happen?

LIST: My name is John List.

Who, you’ll remember, is—

LIST: I’m a professor of economics at the University of Chicago.

But also:

LIST: I am also the chief economist at Lyft.

DUBNER: Now, I understand you used to be the chief economist at Uber.

LIST: That’s correct. I was a chief economist at Uber for nearly two years. And I decided to move over to Lyft and do similar work at Lyft.

There’s one more thing you need to know about John List, and his friend Uri Gneezy. They both love to run huge experiments out in the real world. Economists didn’t use to do this. If they ran any experiments at all, they were likely to be small lab experiments, the kind often run by psychologists.

GNEEZY: You can learn a lot from this, but it’s limited.

Limited in that your research subjects are likely to be a few dozen undergraduates, making artificial choices in an artificial environment. Big field experiments, meanwhile: the new breed of economist sees this as a great way to gather meaningful data. The most recent Nobel Prize for economics went to three economists — Abhijit Banerjee, Esther Duflo, and Michael Kremer — for using real-world experiments to address poverty. A lot of people in the field say it won’t be long before John List gets his Nobel Prize. Here are some of the questions he’s tried to answer by running big field experiments:

LIST: Why do people discriminate against one another? Why do women earn less pay than men in labor markets? How do we convince people to pay their taxes on time?

So it’s not very surprising that John List, when he was working at Uber, would try to cook up an interesting experiment. And what’s more interesting than tipping?

LIST: Even though it does have a spotty history, when you think about the potential economic incentives of tipping, it actually presents an interesting way to induce higher quality amongst your workers, your drivers, your contractors.

In other words: a tip is an incentive for better service. That’s always been the main argument for why we tip the servers in restaurants. Here’s Danny Meyer.

MEYER: I’m the founder of Union Square Hospitality Group.

Meyer is one of the most successful and prolific restaurateurs in New York.

MEYER: We have nine full-service restaurants and then an enormous number of cafes, whether they’re in ballparks or museums or theaters.

He’s also the man behind the global hamburger chain Shake Shack.

MEYER: Correct. We created Shake Shack, but it’s now a public company, and I remain the chairman of the board.

Meyer is a zealot for great customer service — he likes to say his business is hospitality as much as it’s food. But he does not like the idea of tipping as a means to induce hospitality.

MEYER: The theory has always been, in the tipping system, that the only way I could possibly get someone to be nice to me or to bring my food promptly is to create a scenario in which they know that I will either punish them if they don’t or reward them if they do.

So that’s another weird thing about tipping. There’s also this uncomfortable fact:

LYNN: It’s discriminatory.

That’s Michael Lynn. He is a social psychologist at Cornell, and one of the world’s pre-eminent scholars on tipping.

LYNN: Both groups, blacks and whites, will tip a white server more than a black server. And that’s even controlling for perceptions of service quality.

Lynn’s research shows that tipping is an unfair way for workers to be paid, because personal characteristics like a server’s race, gender, and appearance factor too much into customers’ tipping decisions. We interviewed Lynn several years ago for an episode called “Should Tipping Be Banned?”

LYNN: I think I would outlaw it. You could make the argument that tipping is a condition of employment that has an adverse impact on a protected class. And the Supreme Court has ruled that even neutral business practices that are not intended to discriminate, if they have the effect of adversely impacting a protected class, are illegal.

So there’s yet another reason to dislike tipping. You’d think, therefore, that the last thing you’d want to do is to introduce tipping into a business that seemed to be getting along fine without it — a business like Uber. As I’m guessing you know, Uber was one of the most successful startups in history. Before it went public this year, the company was valued at more than $80 billion. Its stock price has since fallen. But, still: a massive company, and a controversial one too. Its founding C.E.O., Travis Kalanick, was forced to resign after a series of acts unbecoming a C.E.O. Another Uber controversy: its drivers, who are independent contractors and not employees, often don’t make much money. So John List, when he was chief economist at Uber, he saw an opportunity.

LIST: Prior to 2017, you know, Uber did not have in-app tipping on the platform. You might have had some situations where people were giving cash tips, but it was something that did not go through the Uber platform.

Actually, List saw two opportunities: to give Uber drivers a chance to earn more money, and to run a huge field experiment on tipping. There was one big obstacle: Travis Kalanick, the company’s CEO. How did he feel about tipping?

LIST: Travis had a pretty strong intuition that was a little bit like Mark Twain. He viewed it more as a tax or an increase in price, because he felt that everyone would feel compelled to tip, because it will become a social norm.

We should note that Lyft, Uber’s biggest rival, did allow drivers to be tipped, to reward good service.

LIST: Travis was fine with rewarding great service, but he wasn’t fine with social pressure leading to people transferring some money to another person. So really our first job was to assuage Travis.

Assuage him that tipping on Uber would be truly optional, and not a social-pressurized default option.

LIST: He made arguments that, look, if say 10 to 20 percent of people tip on the Uber app, then I’m going to consider that a win. So we went back to the drawing board and said, “Okay, what do we need to do to make sure that tipping on the Uber app does not become the social norm?” in that we can get roughly 10 to 20 percent of people to tip.

DUBNER: In other words, that was your target? You weren’t going for 50, 100 percent?

LIST: No. Our goal originally was to make sure that this was not a price increase, and loosely, that meant 10 to 20 percent of the rides.

DUBNER: And if it got much higher than that, it would just be a tax, is that the idea?

LIST: That’s correct.

DUBNER: And I’d like to know — especially because in many cases, workers receive tips because their employers don’t pay them enough, and in some cases, employers have been known to take a cut of those tips — what share of the tip in this case, if any, goes to Uber itself, or does it all go to the driver?

LIST: So the entire tip goes directly to the driver. Uber or Lyft, neither of them take a commission on the tip.

So John List, along with Uri Gneezy and two more economists — Bharat Chandar and Ian Muir — turned the sprawling Uber ecosystem into a gigantic tipping experiment. How gigantic was it?

LIST: So in total, we have over 40 million observations of people.

GNEEZY: All ages, all ethnicities, all income levels almost, all over the country — not knowing that they are taking part in an experiment.

LIST: And of those 40 million observations, we have roughly 800,000 drivers in our experiment.

So how did List, Gneezy, and their fellow researchers set up tipping on the Uber app? Remember, they were directed to not make it automatic.

LIST: Okay, we need to separate the act of tipping in space and time from the actual trip.

And that’s important, why?

LIST: If you take a taxicab trip, the person turns around and you give them the fare. And then 90 to 95 percent of people add a tip on it. It’s face-to-face, there’s a social norm. There’s a lot of social pressure to give a tip in that situation.

So here’s how Uber tipping was set up. The app generally wouldn’t give a passenger the option to tip until they were out of the car, and until the driver had given the passenger a rating — which the passenger doesn’t see, by the way. So neither the rating nor the tip would be contingent on each other, which meant no quid pro quo. Especially because, unlike a waiter or waitress in your favorite restaurant, the odds are you will never see that Uber driver again.

LIST: The fact is that very few people meet twice.

In the biggest cities where Uber operates, there is a less-than-one-percent chance that the same driver and rider will ever be matched up twice. So, Uber tipping is presumably not about social pressure, or reciprocity, or extortion. Instead, it’s supposed to represent true appreciation, in monetary form, of good service. So how appreciative are Uber passengers?

LIST: So a first fact from our data is that roughly 15 to 16 percent of Uber rides are actually tipped.

Think about that for a second. That’s about 1 of every 6 rides that get tipped, versus about 6 of every 6 restaurant meals. So: not great news for Uber drivers thinking they’d be raking in tip money. But good news for John List and Uber — who, remember, were shooting for a 10 to 20 percent tipping rate. Now, let’s get into what List learned about the tippers. There are three categories here: passengers who always tip.

LIST: Only 1 percent of people tip on every trip.

There are people who never tip.

LIST: That’s actually 60 percent of people — six-zero percent of people — never tip once.

DUBNER: So that leaves 39 percent who sometimes tip, yes?

LIST: Exactly. You’re talking about people who are tipping one out of eight times.

Okay, so 1 percent of Uber passengers treat the ride like a restaurant meal, automatically adding a tip. 60 percent are saying, “Um, no thank you, I’ve already paid for the ride, so I’m going to not give you more of my money” — which, if you’re even a little bit anti-tipping in general, sounds pretty sensible. But then 39 percent of passengers tipped occasionally. Does that mean that Uber successfully made tipping about exceptional service?

LIST: So one of the most surprising results is that when you look at the data pattern, it’s actually the rider variables that are roughly three times more important than the driver variables.

Let’s translate what List said out of economist-speak. What he means is that the biggest factor in whether a passenger tips is simply what kind of person the passenger is. That is a bigger determinant than the driver or the car itself or the route.

LIST: So if I unpack that a little bit, the first thing you can think about is the star rating.

Remember, just as Uber passengers can rate drivers, Uber drivers also rate passengers.

LIST: The riders who have a five-star rating, they’re more than twice as likely to tip than riders who have a 4.75 rating.

And what kind of passenger has five stars?

LIST: They’re punctual, or they’re nice in the car, they’re giving respect to the driver.

Okay, and these five-star passengers happen to be the same people who are most likely to tip every time. In other words: they’re just wonderful people! But, there aren’t that many of them — just 1 percent of Uber riders. So who are these generous souls? What kind of person is most likely to tip an Uber driver? For starters, they’re most likely to be male.

LIST: A man is about 19 percent more likely to tip than a female rider.

That’s an interesting finding — especially because it’s the opposite of what John List’s own research shows about charitable giving.

LIST: Women consistently, across the entire distribution, give more money to charities than men do.

And why is Uber tipping so different?

LIST: My intuition is it’s because this is anonymous. The world of charitable giving tends to have a lot of social pressure. And what the literature also has taught us is that women are much more likely to change their behavior in the face of social cues. My intuition would be, if you looked at tipping to taxicab drivers or restaurants, that the difference would not be as stark, and women might even tip a little bit more, conditional on having the same income levels as men in those cases.

DUBNER: Now, to what degree, if any, are you able to control for income of the riders in your study? It may be that the men are wealthier, yes?

LIST: What we do find is that riders from lower-income areas tend to tip less than riders from higher-income areas. But even after we correct or adjust for those income-area differences, we still find that men do tip more than women.

DUBNER: You also write that tips tend to be higher during very early morning hours, between 3:00 a.m. and 5:00 a.m., and that these hours have a disproportionate percentage of airport and business trips. What it calls to mind to me is, those trips may be being expensed and may not be coming out-of-pocket, and maybe people are more generous with tipping if it’s not their own money.

LIST: I think you’re exactly right. I think those can be expensed. What you also find, though, is that tips tend to be high on Friday and Saturday evenings. So this is the going-out-to-eat-and-party crowd.

This may or may not mean that the consumption of alcohol makes people more generous. The Uber data cannot answer that question. So: draw your own conclusions. But there is one huge conclusion the Uber data seems to point to: most people, when given the option to tip, do not tip! And when they do, it doesn’t have much to do with the quality of their experience.

LIST: The bulk of the reasons why a driver will get tipped is because of these things that are outside of their control. But that said, there are many things that the driver can control that do impact the tip.

For example?

LIST: For example, if you look at drivers with a five-star rating, they’re tipped close to 50 percent more often than those with a 4.75 rating. So the quality of the driver matters.

It matters — but, again, not very much compared to whether a given passenger happens to be a tipper or non-tipper. Similarly, there are other trip characteristics that may affect tips. How the driver drives, for instance, which Uber can measure via telematics information from the driver’s phone.

LIST: Hard accelerations matter. And if you have more of those, you receive less in tips. Hard braking matters. If you have more episodes of hard braking, you receive fewer tips.

The age of the car also matters a little bit.

LIST: Drivers in the old cars are tipped slightly less than drivers in the new cars.

And the language the driver speaks — which List measured by seeing whether drivers keep the language on the Uber app in the default setting, which is English, or they change it.

LIST: And what we find is that those drivers who change the default language — those drivers end up getting tipped much less as well.

One way to increase your tips as an Uber driver is to be female.

LIST: What we find is a really strong result that female drivers receive about 12 percent more than male drivers in tips. Now, what’s interesting, though, is that both female and male riders tip female drivers more.

DUBNER: Earlier research of yours at Uber found a gender pay gap, with men earning more, even though the algorithm is gender blind. I’m curious whether tipping substantially shrinks that gap?

LIST: When you include tipping, that difference reduces by about 13 percent. So you still have a pretty large effect. Now, the reason why it doesn’t lessen by more than that is that tips are only about four percent of fares.

Okay, so cumulative tipping on the Uber app is only about four percent of the fare total. Yes, that’s a lot less than the standard 15 or 20 percent tip in restaurants. But still, four percent of extra money on millions and millions of Uber rides. Surely that’s got to be good news for the average Uber driver. Doesn’t it?

LIST: Driver earnings actually do not increase on average because of tipping.

Wait a minute, how can that be?

LIST: What happens is, when you add tipping to the app, more drivers come online, and they supply more hours. So that shifts out the labor-supply curve and just that shift alone, of course, decreases wages. But then you have tips added on to that, that make wages roughly the same as what they were before.

John List and his fellow researchers did find one lever they could press to increase the net amount of tipping: changing the default setting of the suggested tip amount.

LIST: One trip might get a default of, “Would you like to tip $1, $3, or $5?”

Another rider, meanwhile—

LIST: They might receive a preset of, “Do you want to tip $2, $3, and $5?” And those defaults or presets matter a lot. Roughly 16 and a half percent of trips are tipped when you have a low preset versus about 14 and a half percent when you have a higher preset.

DUBNER: What’s the median amount tipped, however, in the different options?

LIST: When you have higher presets, people tip less often but they end up tipping more in the end. So if you want to increase tipping, what you want is a higher preset. Within reason.

If Uber really wanted to increase tipping, it could, of course, manipulate the tipping environment. It could, for instance, prompt a passenger to tip during the ride, which would jack up the social-pressure component, and make it more of a have-to, like in a restaurant.

*      *      *

Danny Meyer grew up in St. Louis, and even though he’s been in New York forever, he still exudes a midwestern sense of hospitality. So, while many of his restaurants — like Gramercy Tavern and Union Square Cafe and The Modern — are well-regarded, high-end restaurants, the style of service is, like Meyer himself, midwestern warm. This combination of warm service and good food translates into big tips for the waitstaff. You might think that Meyer, as the employer of the people getting these big tips, might like this. But you’d be wrong.

MEYER: I decided to take on the tipping system back in 2015 and here we are in 2019, and we’ve learned a lot.

That’s right: Meyer started to eliminate tipping at his restaurants a few years ago. His menus would say “hospitality included,” which meant that prices were higher, but you’d no longer have to add a tip to help pay the server’s salary.

MEYER: We never wanted to hire people who would only be nice to you in expectation of a higher reward.

We made an episode about it back then — it’s called “The No-Tipping Point,” if you want to check it out. As we mentioned earlier, Meyer’s restaurant empire also includes Shake Shack.

MEYER: I will credit Shake Shack with having inspired me to take this step in our full-service restaurants. There are no tip jars at Shake Shack, and there’s no opportunity to tip at Shake Shack. If we can make the economics work where we’re selling a $5, $6, $7 burger, why in the world can we not make this work when we’re serving you a $100 dinner at Gramercy Tavern? It just doesn’t make sense to me.

The haphazard sense of tipping also doesn’t make sense to Meyer.

MEYER: Everybody, practically, leaves a tip in a full-service restaurant. Nobody leaves a tip when they go to McDonald’s. Some people leave a tip in Uber. Almost everybody leaves a tip with a taxi driver. I think that whole social-norm thing is very, very confusing to people. And I’m just happy to opt on the side of saying, “You know that price we put on the menu? That covers everything.”

But Meyer wasn’t only thinking about his customers. One of the biggest challenges of running a restaurant is retaining good employees. It’s a business with a lot of turnover, and turnover is bad for business. One obvious way to keep your employees is to pay them well. The waitstaff and bartenders at Meyer’s restaurants were getting paid well — thanks to all that tipping. But the same couldn’t be said for all the people who actually prepare and cook the food, and wash the dishes, and keep the place running. Meyer found that the tipped employees at his restaurants were earning about two-and-a-half times what the back-of-the-house employees made. Were they really doing two-and-a-half times the work? And the pay gap between tipped and non-tipped kept growing.

MEYER: Every time menu prices go up — and they’ve only gone up over the course of my career — the tipped employee is making more because the tip is just a multiplier of the menu price.

This pay gap had a variety of effects. One was that the people who spent time and money training to be chefs didn’t want to work as chefs.

MEYER: We were seeing more graduates from the Culinary Institute of America, Johnson and Wales, the Institute of Culinary Education, applying for dining-room positions than I had seen in my entire career because they realized they couldn’t afford to live in New York and pursue the thing that had always been their dream.

The pay gap between cooks and servers also created ill will, and led to high turnover in the kitchen.

MEYER: I increasingly saw this divide growing and growing and growing between the compensation possibilities between tipped employees and employees that did not legally qualify to make tips.

What does Meyer mean when he says that kitchen workers don’t “legally qualify to make tips”? Because you’d think one easy solution to this problem would be to simply pool the tip money and redistribute some of it to the kitchen workers. But you can’t do that. One quirk of the hospitality industry is that restaurants are allowed to pay servers what’s called a “tipped wage” — which is below the minimum wage — as long as tips make up the difference. This caveat is called the “tip credit,” and restaurants use it to lower their labor costs. But one restriction of the tip credit is that it applies only to employees who spend at least 80 percent of their time in a customer-facing role. So a restaurant can pool their tips and distribute them among the waitstaff, bartenders, and hostesses — but not the kitchen crew.

MEYER: I might think twice if I believed that the legislature was going to create a new law that permitted tips to be shared everywhere. That has not happened. But even if they did do that, it would still lead me to say this is not a professional way to treat people. Tipping, in and of itself, I don’t think fosters the professionalism that we’re looking for in our industry. If you think about every industry where tipping happens, it’s often not a full-time career. I don’t tip my airline pilot or my lawyer or my teacher or the person who’s interviewing me on their podcast. That’s a profession.

Ha! Podcasting — a profession. Anyway: for all these reasons, Danny Meyer decided to eliminate tipping at all his restaurants:

MEYER: We’ve converted about one every four to six months. We just have one more to go.

DUBNER: Okay. So let’s talk about what you’ve learned.

MEYER: Well, I think the biggest thing we’ve learned is that this is really tough, and it’s tough because it’s really a three-legged stool that we’re trying to balance. How do you create a menu price that doesn’t frighten away a prospective customer who’s just casually looking up your restaurant online and may not understand that there will be no tipping on top of that? And then did you actually compensate your team to the degree that you hope to? And now the third leg of the stool is: did you actually make enough money to cover all of the costs involved?

DUBNER: What can you tell us about your restaurant revenues before and after “Hospitality Included?”

MEYER: Well, I can tell you that our revenues are decidedly higher. But keep in mind we started charging 21 percent more, roughly, on all of our prices. So if they were not at least 21 percent higher, that would be a big fail. What I can also tell you is that the compensation for our staff members, both formerly-tipped employees and never-tipped employees, has also gone up quite a bit.

DUBNER: So if one of the goals was to even out the pay between the people cooking the food and the people serving the food, how much have you closed that gap?

MEYER: Here’s what I know, Stephen: the earning potential before we started “Hospitality Included” was 2.4 times more for tipped employees than for back-of-the-house employees. Today it’s 1.9 times more. So that’s big. I can also say that the line-cook wages — so the people who cook your food — have gone up 37 percent. Front-of-house compensation, formerly tipped employees, has gone up eight percent. So we’ve been able to increase both, but we’ve obviously, in order to make some headway in terms of this discrepancy, have increased back-of-the-house pay by a lot more.

DUBNER: And what’s that mean in actual dollar terms for kitchen versus waitstaff? Maybe you could use The Modern as your example?

MEYER: When we started “Hospitality Included,” in the aggregate, the front-of-the-house employees of The Modern were making somewhere around $22 an hour. And today that number is about $25 an hour. Believe it or not, a cook, in the aggregate, in 2015 was making just over $11 an hour. Today they earn $16.50 an hour. So that’s a serious hike.

Meyer didn’t share with us any profitability figures, so we don’t know the degree to which these higher wages are eating into the restaurants’ bottom lines. But at least in terms of wages, the hospitality-included model sounds like it’s working. Kitchen wages went up a lot; even waitstaff wages went up a little, on average. But not everybody likes to be average.

MEYER: Whenever we’ve converted a restaurant, we would experience as much as a 30 to 40 percent turnover in our legacy staff, and I completely understand why that is. Because the way that you got a raise in a tipped house was to have stayed at the restaurant longer than anyone else, which meant that you got the lucrative Thursday, Friday, Saturday night shifts, and it may have taken you six or seven or eight years to earn that schedule.

But the new, no-tipping model made those shifts relatively less valuable. Which, for shorter-tenured workers, was a positive.

MEYER: One of the real benefits is that you can make as much money working on a Monday night as on a Friday night, because the way we compensate our team is a higher hourly rate plus a revenue-share model that averages out the entire week.

And there was another benefit.

MEYER: I feel great about the fact that I no longer need to worry that the only way to get a great tip is to have a woman subjecting herself to leering men or if I’m from a different ethnic group than the table that I’m serving, that I need to worry that they’re not going to give me a good tip.

But if you think you hear a hint of disappointment in Meyer’s voice — well, I’m not sure that’s what we’re hearing — but when he set out to eliminate tipping, he thought he was starting a revolution.

MEYER: Yeah. Well, we did not succeed at inspiring an entire industry to change its tune.

The no-tipping initiative got a ton of media coverage; it was easy to think that a lot of restaurants were following suit.

LYNN: Not as many restaurants have moved away from tipping as you would get the impression.

That, again, is Michael Lynn, the Cornell tipping scholar.

LYNN: There are about 200 restaurants across the United States. It’s just not as big a phenomenon as people say.

Nor is this the first time the no-tipping trend seemed to be catching on.

LYNN: In the 70’s and 80’s, there was a comparable tendency for a couple of restaurateurs that got a lot of press for it, to eliminate tipping. And some of them survived and continued that policy, but the vast majority ended going back to tipping.

Danny Meyer gets that.

MEYER: I’ve been pretty zealous about how I think tipping actually holds people back. But I also understand that movements happen in a big way when the economics work for everybody. And I don’t think we’re quite there yet. I completely understand it’s not working yet for everybody.

DUBNER: If you were just having the idea today, having gone through what you’ve gone through the first few years of trying it, would you try it again? Or would you consider this an experiment in progress that, while noble and smart, etc., is maybe too difficult? That there’s too much status quo to overturn?

MEYER: Well, I think that’s a fair question and the honest answer is I’m not sure. I think largely because of some of the notoriety of our making this initiative, that has actually brought an awareness to the legislature, which is getting very close, as I understand, to creating a situation where tips can be shared amongst everybody.

Even if such legislation were to happen, it wouldn’t eliminate tipping — in fact, it might strengthen the tipping model. But Meyer sees another development that could kill the model.

MEYER: Each time minimum wage goes up, restaurants that are continuing to have a tipping system have all raised their prices to keep up with the minimum wage increases, and then the guest is tipping 20 percent on top of that. So we’re about one minimum-wage hike away from a pricing advantage over other restaurants while we’re paying our staff more. And I think that’s the point when you’re going to start to see many restaurants eliminate tipping.

But Michael Lynn at Cornell — who once told us he’d like to ban tipping — he’s not so sure we’ll ever see the end. Why?

LYNN: Look, when I started doing research on tipping, I thought tips would be strongly related to service quality. But what I found was that about four percent of the variance in tips left by different groups can be explained by their ratings of service quality.

That is, in restaurants — as in the tipping of Uber drivers — there was a weak correlation between quality of service and tip. But Lynn discovered something interesting in his research: a lot of restaurant servers think there’s a strong correlation:

LYNN: About half of the servers in this country will say that they think their tips are moderately-to-strongly affected by the service they leave. They’re wrong when they say that, but they believe it. And because they believe it, tipping in fact does provide an incentive to deliver better service for at least half of the servers in this country. So yes, my opinion about the incentive value of tipping as a system of compensation has changed over time, although the facts haven’t changed. I just learned more about it and began to think a little bit more sophisticated about it.

Okay, did you follow that? Waiters and waitresses think their service significantly affects their tip, even though the data shows it doesn’t; but because waiters and waitresses think it does, they act as if it does. And Michael Lynn found a similar dynamic from the customer side.

LYNN: Tipping increases customer satisfaction. Customers expect service to be better under a tipping system, and that can bias their perceptions. I’ve done a couple of studies based upon recent restaurants that have eliminated tipping. So for example, Joe’s Crab Shack recently, at several of its restaurants, eliminated tipping. Most of them went back to tipping. Their online service ratings declined during the period that they had no tipping and went back up when they re-instituted tipping.

Does that mean the service actually got worse when there was no tipping?

LYNN: I don’t know. I can’t nail down precisely why it occurred. I believe it has to do with customers expect service to be better. I know that customers generally like lower prices and Joe’s Crab Shack, when it eliminated tipping, it raised its prices.

This suggests that many restaurant customers are price-sensitive — and that they’re so accustomed to seeing menus carrying a pre-tip price that they’d prefer to see the low price and then tip even though they’ll end up paying the same or even more. The human mind is a funny thing, is it not?

LYNN: I’m less likely to want to just ban tipping now, because the data I have suggests that customers are happier with tipping. The press mostly focuses on this from the standpoint of servers. But by far the largest group that’s impacted by tipping is customers. As much as people might grouse about tipping, overall it seems to increase consumer satisfaction. And since they outnumber employees by a wide margin, likely it’s in society’s benefit.

So Michael Lynn’s research suggests that Danny Meyer’s no-tipping revolution will have to wait even longer. But Lynn’s research did identify one category of restaurant that’s ready for the transition.

LYNN: And that is really upscale restaurants that replace tipping with higher menu prices.

Any examples?

LYNN: Danny Meyer’s restaurants fall in that category. I think that the price sensitivity of his customers is lower. I also think that he’s got a strong enough reputation that people won’t anticipate lower service quality. “It’s a Danny Meyers restaurant.” People will still expect good service even if there is no tipping.

So what have we learned — other than the fact that tipping is among the most unpredictable economic activities on earth? I went back to John List to ask whether his research on Uber tipping says anything larger about the human condition.

LIST: I think a first point that the data really highlight is that very few people will be continuously generous and consistently generous. I think that that just jumps out in the data. I think a broader point number two is the heterogeneity across people. We oftentimes have people talking about — well everyone’s different — but we actually have a really good glimpse of an important part of our economy. And it shows you just how important things like, if you’re a nice person and you get a nice rating, that that spills over to this generosity that you can exhibit privately.

I think that’s sort of interesting, because in psychology and in economics, we found that a lot of behaviors don’t spill over to different areas in our lives. So you smoke but then you wear a seatbelt. You’re nice in one setting but you’re mean in the other. This sort of gives you a glimpse that there is some consistency in this generosity, or being nice to another person, that is much stronger than what I’ve seen in the past.

DUBNER: John, what kind of a tipper are you?

LIST: I’m an always tipper. I’m the one percent.

DUBNER: And why is that? Is it that you want to give some of your money to those people? Or is it that you want to be for yourself the kind of person who does that?

LIST: I think it’s both. I think that this is an important time to send wealth or income to another person who’s doing a job such as this. But also I feel good about doing it. I have a warm glow. And it makes me feel good inside that I’m not stiffing the person.

If John List is telling the truth, and I have no reason to think he’s not — this is interesting. That an economist, of all people, always tips. Economists — famous for knowing, as it’s said, “the price of everything and the value of nothing.” Economists — who lecture the rest of us about utility- and profit-maximizing, about putting aside our pesky emotions when it’s time to make a decision.

LIST: I have a warm glow.

“A warm glow?” The fact that an economist like John List always tips is, I believe, the best possible evidence of just how weird tipping truly is.

*      *      *

Freakonomics Radio is produced by Stitcher and Dubner Productions. This episode was produced by Matt Hickey. Our staff also includes Alison Craiglow, Greg Rippin, Harry Huggins, Zack Lapinski, Daphne Chen, and Corinne Wallace; our intern is Ben Shaiman. We had help this week from James Foster. Our theme song is “Mr. Fortune,” by the Hitchhikers; all the other music was composed by Luis Guerra. You can subscribe to Freakonomics Radio on Apple Podcasts, Stitcher, or wherever you get your podcasts.