The Tipping Argument, Taken Seriously From Both Sides

I have changed my mind about tipping twice. That is two more times than most people who argue about it in public, and it is the first clue that the argument is in worse shape than the question.

My position, arrived at slowly and held without much comfort: tipping is a bad system that currently pays a minority of workers unusually well, and that second fact is exactly why the first one never gets fixed. The two camps are not disagreeing about the evidence. They are describing different people. The abolitionist is picturing a struggling worker in a quiet room on a wet Wednesday. The defender is picturing a fast, skilled server in a busy room on a Friday. Both of those people exist and their interests genuinely conflict.

What the defenders get right

Start with the side that gets dismissed most casually, because it holds up better than its opponents pretend.

A strong server in a busy room out-earns any flat wage a restaurant would plausibly offer. Not marginally. The realistic replacement wage in most service-included models is a decent hourly rate, and a competent server on a good section beats it comfortably, sometimes by a lot. Telling that person that abolition is in their interest is telling them something false about their own pay.

Variable pay also tracks something real. A good server turns tables faster, sells more accurately, catches the problem before it becomes a complaint, and genuinely improves the evening. That is not imaginary value and a flat hourly rate does not register it at all.

And servers themselves frequently oppose abolition. Waving that away as false consciousness is condescending in a way that should make people uncomfortable. They read their own payslips more often than anyone writing about them does.

What the abolitionists get right

Now the other side, which I think wins overall, but not for the reasons it usually leads with.

The strongest point is not fairness in the abstract. It is that tip dependence makes tolerating bad behaviour economically rational. If a table's conduct is unpleasant but their bill is large, a tipped worker has a direct financial incentive to absorb it. Every floor manager knows this and most of them have leaned on it. That is a structural feature, not an accident, and no amount of good policy elsewhere removes it while pay works this way.

Second, variance. The same job pays wildly differently depending on shift, season, weather and who else was rostered. A worker cannot plan a life around that, and the people hurt most are the ones with the least slack.

Third, the menu price is a fiction. You cannot honestly compare two restaurants' prices when one expects an extra fifth on top. The history of the gratuity is a long record of costs being pushed off the printed page and onto the customer's sense of obligation, which is a strange thing to defend as transparent.

The kitchen problem

Here is the argument I find hardest to answer, and it is oddly underused.

Front of house is paid a percentage of the bill. The cook who actually made the food is on a flat hourly. Same room, same night, same busy Saturday. When the restaurant has a spectacular evening, one side's income scales and the other side's does not, and the side that does not scale is the one standing next to a grill for ten hours. Every kitchen I ever worked beside could do this arithmetic precisely, and it poisoned more rooms than any chef's temper ever did.

Tip pooling and mandated splits soften it. They do not resolve it, and in many places they are legally awkward.

The real crux is the transition, not the principle

Most tipping arguments are conducted as though the question were whether service-included is better in the abstract. It is not. That question is close to settled among people who have looked at it. The live question is who absorbs the cost of switching.

A single restaurant that goes service-included has to raise menu prices by something close to a fifth. It then sits next to a competitor whose menu reads cheaper for identical food. Diners compare printed numbers, because printed numbers are what a menu makes easy to compare, which is its own subject over in the menu tricks piece. Worse, the best servers leave for the tipping room across the street, because that is where the money still is.

So the first mover eats a penalty for doing the right thing. That is a coordination problem, and coordination problems do not get solved one virtuous business at a time. They get solved by industry-wide movement or by legislation. Anybody who frames this as individual restaurants lacking courage has not run the numbers.

Where I land

Service-included is right. Wages should be wages. The employer, not the customer's mood, should carry the cost of the labour, and the kitchen should share in a good night.

But I have no patience for the version of this argument that pretends the switch is costless, or that the servers resisting it are confused about their own lives. They are not. They are correctly identifying that a bad system is currently working for them, and that the proposed replacement is being designed by people who will not bear the risk of it.

In the meantime I tip properly, including in rooms where I thought the service was poor, because withholding money from a worker is a strange way to send a message to an owner. If the room was badly run, that is a management failure, and it belongs in the conversation about what service actually means rather than on the payslip of whoever was closest to me.