In 1959, two psychologists at Stanford, Leon Festinger and James Carlsmith, paid young men to lie, and then watched what the lie did to them. The men had just spent an hour on a deliberately dull task. Some were paid one dollar to tell the next person in line that it had been fun and interesting. Others were paid twenty dollars to say the same thing.
The result ran against ordinary intuition. The men paid one dollar later rated the boring task as genuinely more enjoyable than the men paid twenty dollars did. The smaller the payment, in other words, the more the lie took hold. Festinger and Carlsmith argued this was not a quirk but a window into how the mind handles its own contradictions, a process they called cognitive dissonance.
The most boring hour they could design
The setup was built to bore. Each of the 71 male students, recruited from an introductory psychology course, was first asked to put twelve spools onto a tray, empty it, refill it, and repeat for half an hour using one hand. Then he was given a board of 48 pegs and told to turn each one a quarter turn, then another quarter turn, on and on for another half hour. The experimenter sat nearby with a stopwatch, pretending this was the whole study.
It was not the whole study. The dull hour existed only to give every participant the same mildly negative experience. The real experiment began when the man was told he could leave.
At that point the experimenter explained, with some staged hesitation, that the assistant who normally briefed waiting participants had not shown up. Could the student fill in, just this once, and tell the next person, a young woman in the waiting room, that the task had been enjoyable? For this he would be paid. One group was offered a dollar. Another was offered twenty. A third group, the control, was never asked to say anything at all.
What changed was the private opinion
After the student told the waiting woman the tasks were fun, he was sent to a separate interview, run by someone who did not know which group he was in. There he was asked to rate, on a scale from minus five to plus five, how enjoyable the tasks had actually been.
The control group, who had told no lie, rated the tasks at minus 0.45, slightly on the dull side, which is about what an honest person would say. The men paid twenty dollars landed at minus 0.05, essentially the same as the control. But the men paid one dollar rated the tasks at plus 1.35, clearly on the positive side. The gap between the one-dollar and twenty-dollar groups was statistically significant, unlikely to be chance.
So the people paid the least came closest to believing their own claim. The people paid well did not. That is the finding the paper was built to explain.
Festinger’s account: not enough reason to lie
Festinger’s theory starts from the discomfort of holding two clashing thoughts at once. A man who found the task boring, and who then heard himself tell a stranger it was fun, is carrying two cognitions that do not fit. The mind, in this account, leans on whatever will ease the friction.
The size of the payment changes what is available to lean on. Twenty dollars in 1959 was real money, enough to justify a small lie on its own. The man could tell himself he fibbed for the cash, and the discomfort had somewhere to go. One dollar gave him no such cover. With no good external reason for the lie, the easiest way to resolve the tension was to revise the memory itself, to decide the task had not been so dull after all.
The smaller reward, counterintuitively, did more work on the person, because it left more contradiction unexplained. Festinger and Carlsmith checked whether the dollar group had simply argued more persuasively to the waiting woman, which might have convinced them through rehearsal rather than dissonance. Independent raters listening to taped conversations found the opposite, if anything: the twenty-dollar group was slightly more convincing. The rehearsal explanation did not hold.
The pattern also held up where it should and faded where it should. The interviewer asked four questions, not one, and they were designed to vary in how closely they touched the lie. A question about whether the men wanted to take part in a similar experiment again moved in the same direction as the enjoyment ratings, with the dollar group most willing, though the effect was weaker. A question about how much they had learned about their own abilities, which had nothing to do with the lie, showed no real difference among the groups at all. The shift was specific to the contradiction the experimenters had planted, which is what gives the result its force.
What this does and does not prove
The experiment is rightly famous, but its reach is narrower than the headline version suggests, and the original paper is unusually honest about that. The effect rested on 60 analyzed participants, twenty per group, after eleven were dropped for seeing through the ruse or refusing the money. The enjoyment shift, while statistically reliable, was modest in absolute terms, a swing from slightly negative to mildly positive on an eleven-point scale, not a wholesale conversion.
The sample is also a product of its time and place: all of them young men, all Stanford undergraduates, all in 1959. Whether the same mechanism operates identically across cultures, ages, and genders is not something a single study of that group can settle.
There is a deeper dispute too. Years later the psychologist Daryl Bem offered a rival reading called self-perception theory, which holds that the men were not soothing inner tension at all. They were simply inferring their attitude from their own behavior, the way an outside observer might: “I said it was fun for only a dollar, so I must have found it somewhat fun.” Self-perception predicts much of the same data without invoking discomfort. Decades of research suggest both processes may operate, under different conditions, which means the tidy story of “low pay equals belief” sits on top of a question psychologists have not fully closed.
Why it still circulates
What keeps the one-dollar study alive is the uncomfortable place it points. It suggests that a person’s stated reason for an action and the action’s actual effect on their beliefs can pull in opposite directions, and that being paid well to do something may insulate your private opinions in a way being paid poorly does not. The volunteer who works for free, the intern who stays late for exposure, the person who defends a choice they were barely compensated to make: the model predicts these are the people most likely to talk themselves into meaning it.
That is a claim worth holding loosely, given everything the study cannot settle. But the core observation has survived more than six decades of scrutiny, which is more than most findings from 1959 can say. The boring hour with the spools and the pegs turned out to be the least interesting part.
