Stay: Exit, Loyalty, and the One Thing Your Delegate Cannot Do for You

You do not answer right away. Everything in the message is true. You asked for this, in a way. Months ago you told your delegate to look after Friday dinners, and it has. It has compared every dining room within reach on every measure a dining room can be compared on, and it has found a winner, and the winner is not the place you have gone for six years.

You know the usual place is not the best. The chairs are hard. The menu has not changed much. Last month the fish was overcooked and you said so, quietly, to the owner, who winced and sent out something else and came by later to ask if it was right. She knows your name. She knows where you like to sit. When your father died she did not let you pay.

None of that is in the comparison. It could not be. Your delegate was asked which table is better, and it answered the question it was asked.

So you type one word. No. Then, because that seems curt even toward a machine, you add: keep the usual place.

The delegate does not argue. It does as it is told. But I want to stay inside that moment, because something happened in it that has not happened before in the history of buying things. For the first time, you were shown exactly what your loyalty costs. Eight minutes. Twelve dollars a person. A higher rating. The price was put on the table, itemized, and you paid it on purpose.

For most of history nobody was handed that receipt. We stayed with the bank, the barber, the insurance agent and the corner restaurant, and we could tell ourselves it was loyalty, and nobody could check. Maybe it was. Maybe we were just tired, or had never looked, or the leaving was too much trouble. From the outside, and often from the inside, those all looked the same.

They are about to stop looking the same. Your delegate does not get tired. It always looks. And the trouble of leaving is exactly the kind of small matter it was built to take off your hands.

This book is about what is left when all the reasons for staying that were never really reasons have been taken away. I think what is left is small, and old, and worth more than it has ever been. I also think it is the one move in this whole business that your delegate cannot make for you. It can leave on your behalf, perfectly, a hundred times a day. It cannot stay.

The scene

Your delegate found a better table and showed you the difference. You kept the old one anyway.

The question

What were you doing when you said no, and why could the delegate not have done it?

Albert Hirschman knew a good deal about leaving. He was born in Berlin in 1915 and left Germany in 1933, when he was barely grown. In 1940 he was in Marseille, helping the American Varian Fry get refugees out of France and over the mountains into Spain, until it was his own turn to go. He crossed, reached the United States, and became an economist of an unusual kind, one who went and looked at things.

In the 1960s he was looking at development projects in poor countries, and in Nigeria he found a puzzle. The country had a railway. It had been built to carry heavy cargo over long distances, which is the one job at which a railway should beat a truck every time. And the trucks were beating it. Worse, the railway did not seem to be getting any better under the pressure. It was getting worse.

That is not what the textbook says should happen. The textbook says competition is a tonic. The customers leave, the revenue falls, the managers wake up, the service improves. Hirschman watched the first step happen and none of the others.

His explanation was simple. The customers with the most at stake, the shippers who cared most about whether their cargo arrived, were the first to move to the trucks. They were also the ones who, with no trucks to turn to, would have made the loudest and most useful fuss. The road took them away. What was left behind was a railway with fewer customers, less noise, and, because the state covered its losses, no particular reason to notice either.

The competition had not cured the railway. It had drained off the people who might have.

He saw the same shape closer to home. When a public school starts to slip, the parents who care most about schooling are the first to move their children somewhere else. They are exactly the parents who would have gone to the meetings. The school loses its best critics at the moment it most needs them, and slips faster.

Hirschman wrote this up in 1970 in a short book called Exit, Voice, and Loyalty. I have leaned on it before, in a piece called Regulate the Door. There I used the first two words of the title. This book is about the third.

Hirschman

A failing railway got worse under competition, because the customers who cared most were the first to leave.

What follows

Leaving is not the only thing a customer can do, and it is not always the thing that helps.

Hirschman said a person facing something in decline has two tools. He gave them plain names.

The first is exit. You leave. You change banks, switch brands, move your child, emigrate. Exit is the economist’s favorite, and for good reasons. It is clean. It needs no meeting and no argument. Nobody has to like you. You simply stop coming, and the thing you left can read the message in its own receipts.

The second is voice. You stay and say something. You complain to the owner, write the letter, go to the meeting, stand for the board. Voice is the politician’s tool. It is messy and slow. It takes nerve, and it can fail. But it carries something exit never can, which is information. A customer who leaves tells you that something is wrong. A customer who speaks tells you what.

Hirschman’s quiet point was that the two tools work against each other. The easier it is to leave, the less anyone bothers to speak. Why sit through the meeting when you can walk across the street? And so an institution with a wide open door hears very little. It learns only that the room is emptier than it was.

He noticed something else, and I think it is the most useful thing in the book for a person who runs a business. A firm in trouble needs two kinds of customer at once. It needs the alert ones, who notice the slip and react, because without them it would never find out. And it needs the inert ones, who do not notice or do not bother, because they are the ones who keep the lights on while it fixes the problem. All alert, and the firm is dead before it can learn anything. All inert, and it never learns.

A firm in trouble lives on a mixture. The alert customers send the signal. The inert ones buy the time.

Hold on to that mixture. It has been the unspoken foundation of every business you have ever dealt with. It is about to change, and not gradually.

Hirschman

There are two responses to decline. Exit is leaving. Voice is staying to say what is wrong. Easy exit weakens voice.

What follows

Every business has survived its bad months on customers who did not leave right away. The question is why they did not.

Exit and voice are the famous words. The third word in the title gets a single chapter, late in the book, and reads a little like an afterthought. I do not think it is one.

Hirschman needed loyalty to solve a problem in his own theory. If exit is so easy and voice so costly, why does anyone ever speak? Why did every unhappy customer not simply go? Yet people do speak. They stay with the slipping school and the slipping party and the slipping restaurant, and they make their fuss, sometimes for years.

His answer was that some people are attached to the thing itself. Loyalty, he wrote, holds exit at bay and activates voice. It does two jobs in one phrase. It keeps the person in the room, and because they are in the room and care how it comes out, it makes them talk.

Notice what kind of loyalty he means. It is not blind. The loyal member has seen the decline as clearly as anyone. It is not helpless either. Hirschman is firm that loyalty only means something when the person could go. Someone who cannot leave is not loyal. They are stuck. And the one who stays freely carries a kind of weight in the room that the stuck never have, because everyone knows the door is there. The complaint of a person who could walk is heard differently.

So loyalty, in his account, is a bet. The loyal person stays through a bad stretch because they judge that the thing can be put right, and that they can help put it right, and that it is worth the cost of waiting. They pay that cost in worse service than they could get elsewhere. They pay it knowingly.

That is where Hirschman leaves it. He is an economist, and he is interested in what loyalty does for the firm. He does not much ask what it is, or where it comes from, or what it does for the person who has it. For that we will need a philosopher, and we will get to him.

First I want to clear the ground. Most of what we have called loyalty, in business and in life, was not this at all.

Hirschman

Loyalty holds exit at bay and activates voice. It only counts where the door is open.

What follows

Staying is loyalty only if you could have gone, knew it, and stayed to make the thing better.

The first counterfeit is the customer who cannot leave.

You know this one from the inside. It is the bank you would have left years ago if leaving did not mean an afternoon of forms. It is the software your whole office runs on. It is the phone plan with the contract, the gym with the cancellation that must be done in person, on a weekday, in writing.

The business books have a polite word for this. They call it retention. A less polite word would be nearer the truth. The customer is not staying. The customer is being kept.

I have written about this at length in Regulate the Door, and I will not repeat the whole argument. The short form is that a company can get worse by exactly as much as it costs you to leave. Put a weight on the door and you can let the room behind it go to seed, and the people inside will stay, and you can point to them and say look how loyal our customers are.

Hirschman saw through that. Loyalty that rests on a locked door is not loyalty, and it does none of loyalty’s work. The captive does not use voice to make things better. The captive mutters. The captive waits. And the day the lock breaks, the captive is gone without a word, because nothing was ever holding them but the lock.

Here is the test. Ask what would happen if leaving cost nothing. If the answer is that they would go, they were never yours.

The counterfeit

A customer who stays because leaving is hard.

The test

Take the weight off the door. Whoever goes was held, not loyal.

The second counterfeit is quieter, and far more common. It is the customer who has not looked.

Most of us are this customer most of the time. We renewed the insurance because the notice came and the number seemed about the same. We have bought the same brand of coffee for eleven years, not from devotion, but because that is where the hand goes. We have a subscription we forgot.

There is no lock on any of these doors. We could walk through whenever we liked. We just have not turned our heads.

This is Hirschman’s inert customer, and I do not want to be unkind about it, because inertia is not stupidity. Attention is the scarcest thing a person has. Nobody can review every arrangement in their life every month, and a sensible person does not try. We look hard at a few things and let the rest run. Letting things run is how a finite person gets through a week.

But a business cannot tell the difference between a customer who is asleep and a customer who is devoted, and many have built their whole plan on not asking. The renewal price that creeps up a little each year is a wager that you will not look. The plan nobody would choose today, still billed to people who chose it once, is a wager that you will not look. A great deal of what gets reported as customer loyalty is this, counted from the wrong side.

The test is the same shape as before. Ask what would happen if looking cost nothing. If the customer saw every alternative laid out side by side each morning, and stayed, you have something. If they would be gone by noon, what you had was their sleep.

The counterfeit

A customer who stays because they have not looked.

The test

Make looking free. Whoever goes was asleep, not loyal.

The third counterfeit has the word right in its name. It is the loyalty program.

You carry these in your wallet and your phone. Buy nine coffees and the tenth is free. Fly enough and you board earlier. Spend enough and you move up a color.

I have nothing against a discount. But it is worth saying plainly what these arrangements are. A loyalty program is a payment for repeat business. It is a price cut, paid late and in a private currency, with the terms kept just complicated enough that you cannot easily tell what it is worth. That last part is not an accident. The fog is the product. If you could see at a glance that your points come to a small rebate, you would weigh them as a small rebate, and shop accordingly.

Paid staying is not loyalty for the same reason a paid compliment is not praise. The moment someone across the street pays more, the bought customer crosses the street, and is right to. There was never anything there but the arithmetic.

And notice which way the bargain runs. In Hirschman’s account the loyal customer pays a cost to stay. They put up with something worse than they could get, for a while, on purpose. In a loyalty program the customer is paid to stay. It is the same word on two opposite things.

The test, once more. Ask what would happen if the arithmetic were done perfectly, every time, to the cent, with every rival offer in view. Whoever stays after that is not staying for the points.

The counterfeit

A customer who stays because they are paid to.

The test

Do the arithmetic exactly. Whoever goes was bought, not loyal.

Three counterfeits, three tests. Take the weight off the door. Make looking free. Do the arithmetic exactly.

Now read those three tests again as a job description. They describe a delegate.

Leaving is hard only because of paperwork, phone calls, waiting on hold, and a form that must be found and filled. Every piece of that is small work, and small work is precisely what you hand down. I teach my students to delegate the small matters and keep what is worthy of their attention. Cancelling a gym membership has never once been worthy of anyone’s attention. A delegate will do it while you sleep, and it will not be worn down by the third transfer to a retention desk.

Looking is expensive only because attention is scarce. A delegate’s attention is not scarce in that way. It can lay every insurer, every plan and every dining room side by side each morning and think nothing of it. I described this in The Monthly Election: when nothing holds you, every provider has to win you again each month, and your own AI casts the vote. The sleeper does not have to wake up. Something is awake for them.

And the fog around the points is only fog to a tired person standing at a counter. To a delegate it is a sum.

So here is what I believe happens, and I do not think it takes long. The captives are let out. The sleepers are woken. The bought are repriced. Every customer a business holds for a reason other than the business itself becomes, all at once, an alert customer in Hirschman’s sense. The mixture that kept firms alive through their bad months, some alert and many inert, tips all the way to one side.

The delegate does not destroy loyalty. It burns away everything that was only passing for it.

That sounds like the end of loyalty, and a good many people will write that it is. I think it is closer to the opposite. When every false reason for staying has been removed, the people who are still there are there for a true one. For the first time a business can look at its regulars and know what it is looking at. For the first time you can look at your own habits and know which of them you mean.

What you mean is the subject of the rest of this book.

The turn

A delegate removes the lock, does the looking, and does the arithmetic. All three counterfeits fail at once.

What follows

Whatever staying survives a free door, a full comparison and an exact sum is the real thing. There will be less of it, and it will be visible.

Josiah Royce was born in 1855 in Grass Valley, California, a gold camp in the Sierra foothills that was barely older than he was. Nearly everyone around him had come from somewhere else, and a good many were on their way to somewhere else again. He grew up watching a town try to become a community out of people who had no particular reason to stay.

He went east, and ended up teaching philosophy at Harvard alongside William James. The two were close friends and disagreed about nearly everything, in public, for decades. James held that an idea is true so far as it works. Royce thought truth needed something larger than any one person to stand on.

In 1908 Royce published a set of lectures called The Philosophy of Loyalty. It was an odd book for its moment and it is an odder one for ours. He took a word that even then smelled of flags and regiments, and argued that it was the center of the whole moral life. Not one virtue among others. The root of them.

His reason was a question he thought every person eventually runs into: what am I for? He did not believe you could answer it by looking inward. Look inward, he said, and you find a crowd of wants that pull in different directions and change by the hour. One want says rest and another says work. One says this city and another says that one. There is nothing in the crowd that can settle the argument, because every voice in it is just another want.

What settles it, Royce said, comes from outside. A person finds something larger than themselves, a family, a craft, a town, a ship, a science, a cause, and binds themselves to it. And then, for the first time, the wants line up. There is something to measure them against. The person becomes one person.

I find that worth a slow reading in a year when each of us is being handed a tireless servant whose entire job is to satisfy the crowd of wants, one at a time, as well as it can.

Royce

A person cannot find a purpose by consulting their own wants. The wants disagree.

What follows

Something outside the self has to be chosen and held to. Royce’s word for the holding is loyalty.

Royce gives his definition early and keeps to it. Loyalty, he says, is the willing and practical and thoroughgoing devotion of a person to a cause. Thirteen words. Each of the three adjectives rules something out.

Willing rules out the captive. Royce is as firm as Hirschman on this. Devotion that is forced on you, or that you fell into and never chose, is not loyalty. The loyal person has, at some point, said yes. This is the open door again, arrived at from the other side. An economist says loyalty needs the possibility of exit. A philosopher says it needs consent. They are describing the same hinge.

Practical rules out the admirer. Loyalty is not a warm feeling. A person who is fond of the place and never goes is not loyal to it. Royce insists that the devotion show up as acts, done steadily, over time. You can see why Hirschman’s two jobs follow. Showing up is holding exit at bay. Speaking up is the practical part of caring how it comes out.

Thoroughgoing rules out the fair-weather friend. It means the devotion does not get reopened every time it becomes inconvenient. It holds across the bad month. This is the adjective that will matter most when we get back to the delegate.

And then the last two words, a cause. Royce means something specific. A cause is not a thing you like. It is something that joins you to other people and that would still matter if you were not there to enjoy it. That sounds too grand for a corner restaurant, and I thought so too at first. But look at what the regulars of a good small place are actually attached to. It is not the fish. It is a room where they are known, an owner whose living they have watched, the staff, the other regulars, the street. It is a small common thing that a few dozen people keep in existence by returning to it. That is a cause in exactly Royce’s sense. It is just a modest one.

He adds a point I have come to think is the deepest in the book. The loyal person chooses the cause, and yet does not feel that they made it up. It was there. It had a claim. The loyal say, in one way or another, that it chose them. I have said for years that ideas have people, people do not have ideas. Royce is describing what it is like when one does.

Royce

Loyalty is willing, so it is free. Practical, so it acts. Thoroughgoing, so it holds.

What follows

Each adjective removes a counterfeit. What remains is a decision, kept up in deeds, not to reopen the question.

Nobody has ever received that note before. I expect it to become routine, and I think it changes what loyalty is more than anything else in this book.

Until now loyalty has mostly been cheap, and it was cheap for a bad reason. We did not know what it cost. The person who stayed with the corner hardware store for thirty years had a general sense that the big store out by the highway was a little less for most things. They did not have a figure. They could not have said what the thirty years came to. The cost was real, and it was hidden in fog, and the fog made it easy to bear.

That fog has done a great deal of quiet work. It let people be loyal without ever quite deciding to be. It also let them believe they were loyal when they were only asleep. Inside your own head, the sleeper and the loyalist were hard to tell apart, because neither one had looked at the price.

A delegate looks at the price. That is what it is for. And having looked, it will tell you, because a delegate that hid the cost of your instructions from you would be a poor one. So every loyalty you keep will come with a receipt. This is what you passed up. This is what staying came to.

Some loyalties will not survive the first receipt, and good riddance. If you see the figure and feel only foolish, you were asleep, and now you are awake, and your delegate can move the account by lunch.

But some will survive it. You will read the figure and feel nothing much, or you will feel something closer to satisfaction, the way a person feels about money well spent. And that tells you something you may not have known about yourself. It tells you what you are for.

Loyalty used to be hidden by not knowing the cost. Now it is measured by knowing the cost and paying it.

This is Royce’s willing, made exact. He said the devotion has to be chosen. For most of us, most of the time, the choosing was vague. The receipt makes it sharp. Every month you are shown the door, the alternatives and the sum, and every month you say again: I know. Stay.

The new fact

Your delegate will show you, to the cent, what each of your loyalties costs.

What follows

A loyalty that survives its own receipt is a known, chosen, paid one. That is the only kind there ever really was.

I have said the delegate cannot stay. It is time to say exactly what I mean, because in one plain sense it obviously can. Tell your delegate never to move your account from the credit union and it will never move it. It will hold that line more reliably than you would.

So let me separate two things. A delegate can keep your loyalty. It cannot have one.

Think about what the delegate is doing when it works well. It is given an aim, and at every turn it picks the option that best serves the aim, with everything it knows at that moment. It does this fresh each time. Yesterday’s choice has no pull on today’s, except as information. That is not a flaw. It is the whole of its value. You want something that re-decides without sentiment every time the facts change.

Now look at Royce’s third adjective. Thoroughgoing. Loyalty is the decision not to re-decide. It is a standing refusal to reopen the question each time a slightly better answer appears. The loyal customer at the slipping restaurant is doing, on purpose, exactly what a good optimizer must never do. They are holding to yesterday’s choice against today’s comparison.

Those are not two settings of one machine. They are opposites. An optimizer that will not switch when switching is better is simply a broken optimizer. A loyalist who switches whenever switching is better is simply not loyal. You can instruct the first to behave like the second in a given case. Then the refusal is yours, and the delegate is carrying it out, the way a clerk carries out a standing order whose reasons are none of the clerk’s business.

There is a second reason, and it is the one Hirschman and Royce share. Loyalty requires that you could leave and that staying costs you something. Your delegate cannot leave you. It has nowhere else to be and nothing else it would prefer. And nothing it does on your behalf costs it anything. The eight minutes are yours. The twelve dollars are yours. The overcooked fish is yours.

People will say their delegate is loyal to them. I understand the feeling, and I would put it differently. Your delegate is not loyal to you. It is yours. A thing that cannot go is, in the terms of this book, the first counterfeit. It may be a wonderful servant. It is not a regular.

So the one who stays has to be the one who could go, who sees the price, and who pays it. In every arrangement in your life that is one person.

The distinction

A delegate can keep a loyalty you hand it. It cannot have one of its own.

Why

Loyalty is a refusal to re-decide, by someone free to leave, at their own cost. An optimizer re-decides, cannot leave, and pays nothing.

If loyalty is a decision not to re-decide, then in practice it is a list. Somewhere in your instructions to your delegate there will be a set of names that are not up for review.

I tell my own delegate you decide on nearly everything, and I mean it. It may compare, move and renegotiate without asking me. But I have always held one line, and I teach it: never send your delegate to deal with a person who is worthy of your own attention. The short list is that same line, extended to the comparison. These are the people your delegate does not get to weigh.

I think everyone will have such a list within a few years, whether or not they call it that. So it is worth saying what belongs on it.

Not much. That is the first rule. Royce spends a good part of his book on what happens when causes collide, and the lesson I take from it is that a person with too many has none. They pull against each other, and you are back to the crowd of wants. A list of forty names is not loyalty. It is the old sleep with a new label. I would be surprised if most people can honestly hold more than a dozen.

People, not products. Nobody owes thoroughgoing devotion to a brand of detergent. The things worth staying for have a face in them: someone who would notice if you stopped coming, and whom you would notice if they were gone.

Places where your voice would be heard. Hirschman’s loyalty activates voice. If you could never say the fish was overcooked and have it land on someone who cared, there is nothing to be loyal to. There is only a vendor.

And every name goes on in the open, with the receipt in view. That is the difference between the short list and the sleep it replaces. Nothing is on it by default. You put the name there, you see what it costs, and you leave it there.

The practice

Keep a short list of names your delegate may not shop. People, not products. Few, not many.

What follows

Everything off the list, let the delegate re-decide freely. The list only means something because the rest is open.

Staying is half of loyalty. The other half is the one people skip. The loyal customer owes the place their voice.

Go back to the railway. What killed it was not that customers left. It was that they left without a word, and the ones who left first were the ones whose words would have been worth the most. Hirschman’s whole book is a plea for the fuss that did not get made.

A free door makes that silence much easier. If your delegate can move you in a minute, the natural thing after one bad dinner is to let it. No awkward conversation. No wince from the owner. Just a different table next Friday. Multiply that by every customer and the owner learns nothing at all, except that the room is empty.

So here is what I think loyalty asks in the age of delegates. For anyone on your short list, you say it before you go. You tell them what slipped, and you give them the chance to put it right. If they do, you have helped keep alive something you care about. If they cannot or will not, you leave with a clear mind, and they at least know why.

And you say it yourself. This matters more than it may seem. Your delegate could draft the complaint in a second, and send it, and track the reply. But consider what the owner is about to be buried in. When complaining costs nothing, everyone’s delegate complains about everything, at length, in perfect prose. Voice by machine becomes free, and whatever is free and unlimited gets ignored. I do not blame the owner who stops reading.

What cannot be manufactured is a regular, at the counter, saying it quietly in person. I have argued in Face Value that the human face is becoming the carrier of trust for exactly this reason. It is also becoming the carrier of voice. The complaint that counts will be the one that someone bothered to bring through the door on their own two feet.

The door is free now, so walking out says nothing. Walking in to say what is wrong says a great deal.

The duty

For the few on your list, speak before you leave, and speak in person.

What follows

Machine complaints will be endless and ignored. The voice of a regular who could have left without a word becomes the most useful thing an owner hears.

I work with restaurants and other local owners in Charleston, so I want to turn the book around for a chapter and speak to the person on the other side of the counter.

Here is the hard news first. A good part of your book of business is one of the three counterfeits. Some of your customers are with you because switching is a nuisance. Some because they have never looked. Some because of the punch card. You do not know how many, and until now you did not need to. Within a few years each of them will have something working on their behalf that removes the nuisance, does the looking and counts the punches. Plan as though all three groups are already gone.

The second piece of news is harder. Hirschman’s mixture was your cushion. When you had a bad month, a new cook, a renovation, a run of slow service, the inert customers carried you while you fixed it. Take them away and a bad month is not a dip. It is a cliff, unless there are people in the room who see the slip and stay anyway.

So the question that matters for a local business is no longer how many customers it has. It is how many short lists it is on.

You cannot buy your way onto one. That is the nature of the thing. An offer to pay for a place on the list is an offer to be the third counterfeit, and the customer’s delegate will price it as one. You cannot lock your way on either. The door is free.

What you can do is the old work, which turns out to be the only work that was ever real. Be there. I tell my students to be the face their customers trust, and I mean a face, in the room, that people can find. Know them. Remember what they said last time. When you slip, say so before they have to. When they complain, treat it as the gift it is, because a regular who complains is doing the one thing that can save you, and a regular who stops complaining has already begun to leave.

And be something that would be missed. Royce’s cause was a thing that joined people together and mattered beyond any one of them. A room where people are known is such a thing. A price is not. Nobody was ever loyal to a price.

For the owner

Assume the captive, the sleeper and the bought are leaving. Count the people who would stay through a bad month.

What follows

You cannot buy or lock your way onto a short list. You get there by being a face, in a place, that a person would miss.

A book in praise of staying owes the reader a chapter on going. Loyalty that can never end is the first counterfeit again, with the lock moved inside your own head.

Hirschman is clear that the loyal person does leave, in the end. Loyalty delays exit. It does not abolish it. The loyalist stays while there is a reasonable hope that the thing can be put right and that their staying helps. When that hope is gone, they go, and their going carries a weight that the easy leaver’s never did. Everyone knows what it took.

Royce gives the other half. He knew the objection: people are loyal to terrible things, and loyal armies face each other across every battlefield. His answer was a phrase that sounds like a riddle and is not. Be loyal to loyalty. Choose and serve your cause in a way that strengthens the loyalties of other people, and does not prey on them. A cause that can only live by destroying what others are faithfully attached to has forfeited its claim.

Bring that down to the size of this book and it gives a workable test. Does the thing you are staying with still keep faith with you and with the others who depend on it? A place can serve a bad dinner and still be keeping faith. It winces. It makes it right. A place that has started treating your staying as a weight on the door, that raises the price because it has worked out you will not go, has stopped being a cause. It has become a counterfeit from its own side. It is betting on your sleep.

So here is the order of things, as I understand it. Notice the slip. Say it, in person. Give it time, an honest amount. Watch what they do with what you said. If they turn toward it, stay, and be glad you spoke. If they turn away, take the name off the list. Your delegate will have a better table by morning.

That last step should cost you something. If it does not, the name was never really on the list.

The limit

Loyalty holds exit back. It does not forbid it.

The order

Notice. Speak. Wait. Watch. Then stay gladly or go cleanly.

Readers of my other work will want to know how this fits the Reality Equation. I will keep it short and mark what is argument.

The Equation says Reality is a ratio, the Actual divided by Expectation. The Actual is what the world hands you. Expectation is what you unconsciously bring: a prediction built from history, and alongside it the pull of an idea. Start by naming whose Reality we mean. Here it is yours, at the table, on the night the fish is overcooked.

That night the Actual came in under what you expected. The ratio fell below one. You felt it, as anyone would.

Your delegate, reading the same night, sees the same shortfall and does the sensible thing with it. It updates the prediction and compares. That is all the message in the prologue was. It is a clean, lawful reading of a ratio, and I have no quarrel with it.

But the Equation holds that once Reality arrives, more than one action remains possible. That is where freedom lives. You cannot reach in and change what you expected, and you cannot change what was served. No one can touch either term. What you can do is act, and your acts leave artifacts, and artifacts become history, and history is the only thing that ever retrains anyone’s expectation, slowly.

Seen that way, leaving and staying are two different histories being made. Leaving makes an artifact of absence, which teaches the owner almost nothing. Staying and speaking makes a better dinner next month, in her Actual and in yours.

And there is the idea. A room where people are known is an idea, and I hold that ideas have people, and not the other way around. An idea that goes unfed does not announce its decline. Behaviour becomes the expected thing, the idea goes unfed, and nobody feels the turn. Each return, each word to the owner, is an act on its behalf. I think that is what Royce saw from his side when he said the loyal serve a cause they did not invent.

What is record: the Equation, its terms, and the rule that only action reaches them. What is argument: that loyalty is best understood as the choice of which history to make after a disappointing night. I am putting that forward. I have not proved it.

The Equation

A bad night lowers the ratio for you and your delegate alike. The delegate compares. You may still choose which history to make.

What follows

Staying and speaking are acts that leave artifacts. Only artifacts change what anyone expects next time.

Five predictions. Each can be checked, and I have tried to say them so that I can be caught.

1. The sleepers wake all at once

   Within a few years most people will have something comparing their recurring arrangements for them, every month, without being asked. Businesses that have lived on customers who never looked will find that margin gone in a season, not a decade. The creeping renewal price is the first thing to die.

2. Points turn back into prices

   Reward programs will be read by delegates as what they are, a rebate with a number on it, and weighed against every other number. The fog stops working. Most programs will quietly become plain discounts, and some will be dropped.

3. Decline stops giving warning

   A business with no loyalists will go from fine to empty faster than its owner can react, because the customers who used to linger are now moved overnight. People will call it bad luck. It will be Hirschman’s mixture, missing its second half.

4. The short list becomes a setting

   Delegates will carry a named list of people and places that are not to be compared, and showing its cost will be standard. People will talk about their list. Owners will know, and say with some pride, who has them on one.

5. The complaint that counts is carried in

   Written complaints, being free to produce, will be answered by machines and read by no one. Owners will learn to weigh what is said to their face by someone they recognize above everything else they receive, and they will be right.

Three claims carry this book. For each, here is what is already true, what still has to happen, and where I am probably wrong.

You sit where you always sit. On the way over you passed the other place, the one that was rated higher and cost less, and it looked very nice through the window. Your delegate was right about it. It is right about most things.

It did everything you could ask. It found the door, and showed you it was open, and told you what was on the other side, and what staying would cost. That is more than anyone has ever been told about their own habits. Then it waited.

What it could not do was the last part. It could not look at the open door and the lower price and the better rating, and decide that none of them was the point. It could not owe this room anything. It could not be missed here. It was not the one who was let off the bill the week of the funeral.

So you did that part. It took one word.

The fish, tonight, is perfect. You tell her so. She says she changed the supplier after what you said last month, and thanks you for saying it, because nobody else had. Six other people, she thinks, just stopped coming.

You could have been the seventh. It would have been free.

Your delegate can leave for you. It is very good at leaving, and for nearly everything in your life leaving is what you want from it. But there are a few rooms in the world that exist only because certain people keep walking back into them, knowing the price.

Keep the table.