The Free Coke That Bought a Yes

In 1971, a Cornell graduate student named Dennis Regan ran an experiment at Stanford, where he had done his doctorate, recruiting freshmen from their dormitory rooms. Eighty-one of them signed up for a study of “aesthetics” and were paid $1.75 each. While they waited for it to begin, the phone on the secretary’s desk rang, and the other participant in the room picked it up. That call was the experiment’s first manipulation. A secretary on the line read from a random number table and told the man whether to be pleasant or unpleasant, in a conversation the subject could hear but not see. Half the room heard a polite stranger explain that the secretary had stepped out and suggest the caller try again later. Half heard him snap “Nah, there’s no secretary here… just call later, lady,” and hang up in the middle of the caller’s sentence. Neither the secretary nor the confederate knew, until the phone was answered, which of them the room had been assigned.

Then the two of them were seated in adjoining rooms and given the real task: rating reproductions of paintings. During a break, the confederate went out and came back with two bottles of Coke. “I asked him if I could go get myself a Coke, and he said it was OK, so I bought one for you, too.” He put one on the subject’s desk and refused payment. It was offered very seldom. In a second condition the Coke came from the experimenter instead, which meant it came from someone who would not be asking for anything. In a third, no drink appeared at all.

Near the end of the session, the confederate wrote a note and slid it over. Would you do me a favor? He was selling raffle tickets for his old high school back home, to build a new gym. Twenty-five cents each. The prize was a new Corvette, and if he sold the most tickets he’d take home fifty dollars. If you’d buy any, could you write the number on the note and hand it back?

🧠 What the Numbers Said

Raffle tickets purchased, averaged across the six conditions:

Seventy-seven subjects had usable data. All but two of them bought between zero and three tickets. The two who spent their entire $1.75 on seven tickets each had their scores rescaled to four, and Regan notes in the paper that none of the conclusions change if you leave those sevens alone.

The favor roughly doubled compliance, forty minutes after the fact, for the price of one soft drink.

🤔 The Two Things That Weren’t Supposed to Happen

Regan’s hypothesis going in was that the Coke would work through liking. A favor makes you like the person who did it, and liking makes you say yes to that person. So he built in a control to separate the two mechanisms: sometimes the drink came from the experimenter. Same cold bottle, same interruption of the break, no follow-up request attached to it.

That version bought almost nothing. Just 1.08 tickets against 0.92 for no gift at all. The debt doesn’t attach to the gift. It attaches to a person, and specifically to the person who is standing in front of you asking.

The second surprise is the phone call. The unpleasant condition was designed to make half the subjects dislike the confederate, and Regan then measured whether that dislike hurt compliance. The interaction between liking and the favor came out below chance (F < 1). A man who had just been rude to a stranger on the telephone could hand over a Coke and get the same obedience as a pleasant one.

The mood ratings tell the same story from the other side. Subjects who received the drink did rate the confederate as more generous, more honest, more altruistic, more likable — all of it, with t-values to spare. The favor produced the goodwill. The goodwill did not produce the tickets. What produced the tickets, Regan concluded, was normative pressure: the sense of an outstanding balance, what the sociologist Alvin Gouldner had named the norm of reciprocity eleven years earlier.

Robert Cialdini opens Influence with reciprocation as the first of his principles of persuasion. Liking comes fourth. The ordering is the finding: being owed moves people more reliably than being enjoyed.

🔗 A Free Sample Is a Small Invoice

The reason this keeps working is that the debt is never priced. The giver spends a fixed amount and the receiver takes on an open-ended entry, which is why free samples convert: the supermarket tray of cheese is not a taste test, it is the first line of an agreement nobody remembers signing. Same structure in the restaurant mint that travels with the bill, in the address labels charities mail out, in the trial tier that asks for a card anyway. The value of the object barely matters, because the object is what the giver spent and not what the receiver owes.

You can watch the same mechanism in reverse in any negotiation. Somebody opens with an unreasonable number, then retreats to a reasonable one, and the retreat itself creates a pull to answer it. That’s the reciprocity rule wearing a suit. Making a concession is a way of putting the other side in debt.

🎲 The Only Clean Exit

Regan’s subjects had exactly one way out. Paying for the Coke on the spot would have converted it from a gift into a purchase, and a purchase carries no obligation. In seventy-seven people, almost nobody tried. The confederate stood there ready to refuse the money, and the money essentially never appeared.

So here’s a small thing to try this week. Think of the last favor you accepted that you didn’t ask for — a drink someone set down in front of you, a ride, a phone call out of nowhere, someone handling something that was yours to handle. Then look at the next request that person made of you, or the next one they will. The gap between the two is usually much larger than the favor was worth, and it is the least discussed line item in your calendar.