Observations From a Tipless Restaurant, Part 2: Money and the Law

This is part 2 of a series detailing what I learned from operating our farm-to-table flagship restaurant, the Linkery, as a “no-tipping” restaurant that instead charged a fixed percentage for service, from 2006 to 2013. We also operated a sister restaurant, El Take It Easy, that followed the traditional tipping model, allowing for a fairly direct comparison.

In order to understand how we ended up switching to a fixed service charge and declining tips, it’s helpful to understand some financial and legal facts about the restaurant industry that are not widely communicated to the public at large.

The first is that, for the most part, the restaurant industry operates on very slim margins. There are some restaurants that are very profitable, but the vast majority are not. I’ve read that the average profit for a restaurant is 4% of its revenue, and in my experience most independent restaurants do not have any real profit at all. More typically the owner is able to pay herself for her work, and that’s about it. This fact is important because it means that, for most jobs in the restaurant industry, the wage paid by the restaurant comes from a limited pool, and there is not a lot of flexibility in the size of the pool — if one job gets a $2/hour raise, that most likely means that another job will have its wage reduced by $2/hour.

Theoretically, it seems like there would be ways to free up more money for labor, and there are some methods that help. In the end, though, restaurant-going is so prevelant that the expectations of the market — for price, quality, ambience, and so forth — are pretty well set in every dimension. And the cost of meeting those expectations is pretty well set, too. So the business model for most restaurants, and what they pay for each job, is tightly constrained.

Now, let’s say that on a typical shift, a restaurant sells $1000 in food and drink. It would be reasonable that, to make that revenue, a restaurant has 2 cooks who work 8 hours each, a dishwasher who works 8 hours, and two servers who work 6 hours each. We can extrapolate from standard industry models that, of the $1000 in sales, there will be $300 available to cover the 36 hours of labor. It just so happens that this math means that everyone in the house will make $8/hour, which is of course both minimum wage and a poverty wage. But that’s just how the pie divides.

And yet, wait! We’ve forgotten something. There are also 220 extra dollars paid by the guests as tips. (This 22% is typical for restaurants like ours in San Diego — the exact amount will change with restaurant style and location.) This tip money could add another $6/hour to everyone’s wage, getting everyone up to $14/hr.  While $14/hr isn’t enough to live well in San Diego, it starts approaching realistic money.

However, to give the tip money to every worker would be illegal.  The law is historically very clear — the $220 in tips belongs to the two servers only, and cannot be distributed to any other employees. So, the two servers make a total of about $26/hour each, while everyone else in the restaurant is stuck at $8/hour.

Now, if you’re a server, you’re probably nodding your head that this seems fine! However, if you’re an owner or manager trying to operate an excellent restaurant, you’ve got some problems. From the perspective of the business, the cook and dishwasher are just as important to the guests’ experience as the server is. In fact, they may be even more important, particularly if your restaurant’s reputation is based on its food. But, now your cooks are making less than 1/3 of what the servers are making. The cooks’ wage is so low that talented cooks won’t be able to live on it, and they will probably leave the industry. Additionally, no matter what cooks you have, there will be a lot of ill will between the producers of the food, who are well below the poverty line, and the servers, who are making white-collar wages.

In some states (big East Coast restaurant states come to mind, like New York), the government balances this situation by offering what is called a “tip credit”. The tip credit allows restaurants to pay their front-of-house employees less than minimum wage (usually about $5/hr less), if tips will make up the difference. In the example above, the servers in a “tip credit” system would most likely end up making about $21/hour and, cooks about $12/hour, and the dishwasher still about $8/hour. To someone like me who puts a lot of demands on his kitchen, this distribution still feels inequitable — but it’s a lot closer to fair. While the “tip credit” is a blunt instrument, it does address the wage inequity problem enough to keep the tipping system at least somewhat viable.

Now, in California and several other states, the “tip credit” is not allowed, and the restauranteur has fewer options. The recourse most commonly used in these states is the “tipout”. Here the servers just give a portion of their tips to the kitchen, usually as part of a cultural expectation created within the restaurant. The problems with depending on a voluntary tipout to equal out the pay on the team are:

  1. Some servers may decide to withhold a tipout, in a sense cheating the system, and the employers is precluded from redressing this; and
  2. Servers may use that option of withholding their tipout, to extract special favors from the kitchen regardless of whether those favors hurt other guests or other servers.

In other words, with the business disallowed from enforcing the tipout system, control is left to the culture of the team in the restaurant; and humans being humans, more times than not that situation leads to hurt feelings, anger, or worse.

In a tipout system, what started as one enterprise (a restaurant selling its food & hospitality to its guests), has now spawned two completely new, concurrent businesses: the business of the server selling the perception of extra attention to the guest for tips; and the business of the kitchen workers selling favors to the servers for tipout.  These additional, parasitic businesses are not focused on improving the quality of the original enterprise. Instead, these secondary businesses are focused on maximizing income using short term, and proven effective, means.  (I won’t go into those means here, but you can find many of them detailed here.)

Over the last couple years, another alternative has started to emerge — the lawful tip pool.  In this model, the business is permitted by the law of the land to distribute its tips among every member of its team, including kitchen workers.  In the Ninth District (the western US), one court decision (Cumbie v. Woody Woo) has paved the way for this, but the US Department of Labor has attempted to nullify that decision by issuing new regulations.  It’s not clear that the Labor department is actually allowed to trump the court, and the matter is not yet clearly resolved. (Update: in February 2015, a ruling by a 3-judge panel of the Ninth Circuit made it once again clearly against federal law to include cooks in a tip pool).

Back in 2006, however, it was still definitely illegal for a tip pool to include back-of-house workers.  The only legal way that tip revenue could be distributed among the whole team was for it to not be tip revenue. Instead, it had to be an amount charged by the business for its service. In other words, a service charge.

So, following the lead of places like Chez Panisse, we added a line item service charge on our bills, and that allowed us to distribute the money paid for table service in a way that seemed more equitable.  Having more equitable distribution in money, of course, led to us having better quality service and food. In turn, we saw a sharp increase in business over the first two months of the new system.  Our servers’ total pay rose to about $22/hour, most of the cooks started making about $12-14 depending on experience, and the diswashers about $10. (Those numbers declined somewhat over the years, as the recession took its toll on restaurant check averages in San Diego).

There was big difference, however, between what we did and what places like Chez Panisse had done — unlike those other places, and unlike every other sit-down restaurant in America, we refused to accept money beyond the service charge.

We had a couple reasons for not letting our guests give extra money beyond the 18% we allocated for the cost of table service.  The first reason was that we wanted to make it clear to our guests that we were not trying to take more money from them — in fact, we were taking less, by switching from a tipped system that averaged 21-22% of the check, to a fixed 18%.  We were reducing our tip income by 15% as a show of good faith to our community.

The other reason we didn’t accept tips was that removing any option for tipping was the only way to remove those two parasitic businesses — the side business between the server and guest, and the side business between the server and cooks — from our own company.  By not allowing these side businesses to exist, we created an environment where all of us were engaged in only one mission, our stated goal of creating remarkable experiences for our guests, around local food and drink.  Being in only one business made us like pretty much every other company in America, except that it also made us unlike any other restaurant in America.

And the tension in that, of course, is where the interesting stuff was.

Click here to continue to Part 3.

Click here to read Part 1 of this series.

Observations From A Tipless Restaurant, Part 1: Overview

This is the first of a multi-part series detailing what I learned from operating our farm-to-table flagship restaurant, the Linkery, as a “no-tipping” restaurant that instead charged a fixed percentage for service, from 2006 to 2013. We also operated a sister restaurant, El Take It Easy, that followed the traditional tipping model, allowing for a fairly direct comparison.

On Sunday, October 12th, 2008, I woke up early and, as was my habit in those days, checked email on my smartphone to see what our sales had been the night before, and if there were any emergencies I needed to attend to before we opened for brunch. The top email in my box was from a man whose name I didn’t recognize.

“You, sir,” read the email’s only sentence, “are a douche.”

With that, I knew that the New York Times Magazine article about us had dropped.

In 2006, you see, we had turned the Linkery into a fairly interesting experiment: a table-service restaurant, serious about the quality of its food and service, that would not accept tips and instead charged a fixed percentage of the bill to cover the cost of table service. Amazingly, while there were other restaurants (most famously, Chez Panisse) that added fixed percentage “pre-tip” service charges to their bills, we appeared to be the only place in the US that combined a service charge with refusing to accept any extra tip beyond the service charge. So when Paul Wachter, the writer of the NYT Magazine piece, went looking to explore what a restaurant might look like if the tip dynamic were completely removed, he ended up with us.

His piece (click here to read it) is great, and gives you most of the background you could want to know about us, so I won’t rehash it here. I will however give an overview of the basic Linkery no-tipping story and our experiences with it, for readers who might not be familiar with it yet.

1) Due to poorly cohering laws in many Western US states, using a service charge has typically been the only legal way for a restaurant business to balance wages between servers, bartenders, cooks and dishwashers. That’s why restaurants like Chez Panisse instituted such a policy. Subsequent court decisions in the Western US have opened up the possibility that other arrangements are legal, but the service charge is still the safest model.

2) Because tips cannot legally, in most cases, be controlled by the employer, they are typically distributed (or not distributed, as the case may be) according to a social compact among the employees. To the extent that the rules of the compact are enforced, enforcement is through social means, like ostracization. In either event, the systems for both acquiring and distributing tips are easily gamed by members of the serving staff who are intent on doing so.

3) The Linkery’s most transgressive act was not in implementing a service charge. Our most transgressive act was refusing to allow our guests to pay our servers anything more beyond the service charge — this is where the angry came out. A certain small number of very vocal men (and it was always men who were vocal about it) resented that we were not letting them try to exercise additional control over our team members. This was true even though compelling research has shown that servers do not adjust quality of service as a result of tips; instead the idea that the restaurant was not offering our servers up as objects of control, was heresy. For these people, the primary service they wanted from the restaurant was the opportunity to pay for favors from the server — much like the patron at a strip club pays the club for the opportunity to dangle bills in front a dancer for individual attention. The idea that a restaurant could legitimately want to be in a different business than a strip club, was not an idea these guests could countenance. Thus, I was ever subject to witty takedowns like you are a douche, along with other well-thought-out gems.

4) Our ability to make sure team members in all parts of the house were taken care of, and to remove tip-related squabbling from our business, gave us a huge competitive advantage in the marketplace; this in turn allowed us to serve a much higher quality of food and take lower margins on it. Basically, it was because of the much-lower-friction monetary flow through the company that we were able to survive as a true, deep farm-to-table restaurant in San Diego for so many years. Other operators in town, fully aware of how tips poison restaurants, knew we were enjoying an edge. Some of our colleagues resented this, and lashed out in some ways, including that of telling local journalists and bloggers that we were lying about the food we were serving. I assume that this is because those restaurants couldn’t serve the kind of food we did and still take tips, because tips are so wasteful. And if they couldn’t do it, than they assumed/said we weren’t doing it.

5) Once established, the tipless/service charge model made us more successful in every dimension. Having a sister restaurant that used the traditional model was helpful in evaluating this — at our second restaurant, for instance, we could never achieve a consistently high quality of service. We believed the block came from the sense that, once the guest delivers a tip, the quality of service has been validated — even though studies clearly show that, across a large sample, guests tip basically the same regardless of quality of service. Meanwhile, our revenue was always higher at the tipless restaurant, I think because quality of food and service were both better due to the more consistent pay system (which at the Linkery was much closer to that of a normal, non-hospitality business than that of most restaurants, where server pay varies with a lot of randomness). With higher revenue and more consistent pay system, our retention was better. This continued to be a “virtuous circle” of benefits we saw from having a tipless/service charge model. On a personal level, it was much more fun to work with the non-tipped team; in that environment it was easier to build a focus on doing great, worthwhile work, and doing it well, when those thoughts weren’t being interrupted every couple minutes by a guest deciding how much to pay a team member for their last few minutes of services rendered.

This is a summary of the experiences I had in our no-tipping lab, and in my next few posts I’ll dig a little deeper into each of them. Then I’ll finish this series by talking about what I’ve learned this year from a couple new friends who are researchers from the University of Guelph, and who have brought me in contact with some deeper thoughts about the tipping issue, from the social justice side. After seeing what they and their colleagues have uncovered, I’ve become convinced that thoughtful cultures who value civil rights will make tipping not just optional but illegal; and that this could actually happen sooner rather than later, when courts assess the reality of the situation.

click here to continue to Part 2.

The Inverted Pyramid

Closing the Linkery last week was mostly a joyous event but it was of course also, in a profound way, sad. The sadness wasn’t a sadness of ego — no, ending a project is more exciting than beginning one, in the same way that high school graduation is more exciting than the first day of ninth grade.

Instead, this sadness was very much about saying goodbye to my closest friends, the people I had worked with for so many years. They were all very generous about it, but we all knew that, by quitting in my job as owner, I was resigning as their advocate — or, as we’ve called it for a long time — firing them as my bosses. And really firing people — a thing which is different than the kind of day-to-day firing and quitting which is part of the fabric of many hospitality companies — is a raw thing to do.

Let me give a little background.

Many years ago, I was searching for a model to explain how I saw the owner/manager’s role in a company, and I ended up settling on what I call “the inverted pyramid”.

To contrast: most conceptions of how a company is organized end up looking like a pyramid. Many of us have seen this type of “org chart” distributed in companies we’ve worked at, it looks like this:

typical org chart (click to embiggen)

It is a command-and-control structure, and I believe its primary function is to perpetuate the illusion that makes us big bosses feel so good: that the success of our companies originates in our personal superawesome noggins, and we then propagate those ideas by instructing the department managers what to do, with them in turn directing the individual workers. The end result that I the big boss, get to watch what seems to be a hive of people, all acting as my extension, helping bring my Jesus-light to my customers who need it so much. Drawing that org chart feels great because it shows how important I am!

Of course, as anyone who was ever worked in a job toward the bottom of that pyramid knows, that chart has little to do with reality. In all but the most Tayloresque organizations — and in any organization that can be competitive in the post-connection economy — the people at the bottom of the chart are the ones responsible for making the decisions which make the company beloved, and doing the work that creates value. In most companies, the front-line workers are getting the job done for their customers, and also doing their best to neutralize the potentially destructive influence coming down the chain of command from the “top”.

Now, I’ve lived on all levels of that pyramid and I’ve been guilty of all sorts of sins, including top-down thinking and, as the saying goes, “taking myself too goddamned seriously”. Because when you’re responsible for meeting payroll, in times of stress it’s really hard to focus on letting each person contribute their gifts — in fact, it’s nearly impossible to *not* get into a command-and-control mentality. Simply put, when there are rocks all around the ship, you don’t want hold a town hall meeting on the bridge. That said, I know that any time I’ve been in command-and-control mode as a manager, our company was not — could not be — anywhere close to its best.

Because I know all that, I tried to change my perception of our company to an upside-down pyramid. That is, a pyramid that looks like this:

actual organizational functions (click to embiggen)

In this model, the front line workers are individual agents who do work that has intrinsic value. If, for instance, they are cooks, they could choose to sell their services independently, as private cooks who come to your kitchen and cook you dinner, one meal at a time. If they are servers, they could sell their services as individual butlers or valets. However, instead of selling their services independently, in this case they have decided to band together to sell their services as a group — say, as a restaurant.

As cooks, servers, dishwashers, janitors, and office workers, they have all the skills they need to operate the restaurant. However, if they’re busy cooking/serving/washing dishes/running spreadsheets, none of them will have the time to actually organize the company — to make sure everyone is meeting their commitments to the group, that revenues are distributed fairly, that adequate capital is obtained, that word is getting out about their product, and so on. So, this group of workers conceptually “hires” an owner/manager to assume those responsibilities, and they each agree to pay that owner a percentage of the revenue they generate.

This decision made by the workers to hire an owner works because, as a group, this band of workers has now greatly increased their earning potential: the 20 of them can easily serve 500 people in a night (25 guests per person), whereas independently they might have struggled to serve 8 people per worker. Out of the extra difference, they pay the salary of their manager, whose job is to provide them the service of maximizing the value of their effort. In other words, they have hired the “owner” to represent them and make their work as valuable as possible.

I believe that this mental illustration of the relationship between owner/manager and front-line workers (ignoring for the moment capital, to which I’ll return shortly) is the only illustration that makes sense. However, we can’t see it because of the language we use.

When a group of workers hires an owner, we say that the owner hired a group of workers.

When a single worker decides that her owner is not doing good enough work maximizing the value of her work, and she fires that owner in order to find a better one, we say that she quit.

When an owner decides that he is not capable of maximizing the value of his team as a whole if one specific person continues to be among his responsibilities, and the owner elects to stop representing the work of that person, we say that the owner fired the worker.

In every case, our language is describing the exact opposite of what happened. But there’s a reason for this: because, in our system, the capital is controlled by the management class. In fact, capital, owner and management are so closely associated that, in practice, we use the terms interchangeably.

In our situation where capital and management are completely aligned, owners of profitable collectives are able to extract far more than the value of their services. Skilled workers in most industries (unless they have very rare talents or are unionized) have extremely limited choice in hiring their owners, and are always at a major financial disadvantage in negotiations with the very managers they are hiring. If the workers want to eat and pay rent, they end up giving their owner almost all the excess value of their services. And they’re usually better off keeping whatever job they have, since any new owner they find is likely to be as bad or worse than their current one — and the worker may go broke while they find a new owner who will work for them.

(It’s interesting to note that, while it seems to us like a given, the confluence between capital, ownership and management is not a necessary condition. We can easily imagine a society where the standard model of business formation is for businesses to be capitalized by guilds. In this model, skilled workers would band together to start a company, and obtain capital from their guild, to which they already paid dues. The capital would be in the form of a loan whose interest is calculated exactly to pay the insurance against some such loans failing. The workers’ primary form of compensation is profit sharing after making their loan payments. The workers hire a manager, whose job it is to maximize the value of their effort. If the workers evaluate that the manager is not doing an adequate job in increasing their value, they fire the manager and get another one. In this perfectly plausible scenario, the guild represents the capital, the workers are the owners, and the manager is an employee. It’s only by cultural happenstance that we now typically experience all three roles as being held by the same entity, or by two or three closely aligned entities.)

Given the power imbalance in the world we actually live in, we let our language reflect the real control that management-capital has over workers. An owner who fires a worker may just be thinking that he (the owner) can perform better representing a different worker; but the truth is that usually the worker will have a much harder time finding a new owner, than the owner will have finding a new worker.

The good/bad news is, I think, that this reality is merely an artifact of the dominant business paradigm in our current society. As community-based businesses, and businesses requiring insignificant amounts of capital, become more prevalent, then capital, management and ownership can start to be seen again as separate functions. When that happens, the true role of management, as an agent of the workers whose job is to add value to their work, can be seen more easily.

Which brings us back to closing the Linkery. I made that decision that we would do so, when it became clear to me that, given the changes in the market over the last few years, I was no longer the right person to maximize the value of the work that our people could do. In fact, it was clear just about any other manager would do a better job at it. And, while there are markets in which I know I can help make people’s work much more valuable, those markets are not in San Diego. Which means I had to fire all these wonderful people — fire my closest friends — as my clients, force them to find a new manager to work for them, and leave town.

All of us will be far better off for it, I am sure. But the act of actually doing it, sucked.