In 1971 two psychologists, Philip Brickman and Donald Campbell, wrote a chapter about how a good society might be planned, and buried in it was the most deflating idea in the happiness literature. Whatever happens to you, you go back. Money, marriage, disaster, rescue — the belt keeps moving and the scenery does not change. They called it the hedonic treadmill, and the name stuck because the picture is so easy to hold in your head. Seven years later Brickman went looking for the two most extreme human events he could put side by side in a questionnaire, and for half a century his result has been the sentence people reach for when they want to explain why the raise didn’t help.
🧠 The Lottery and the Wheelchair
The 1978 paper ran in the Journal of Personality and Social Psychology under a title that is already the argument: “Lottery Winners and Accident Victims: Is Happiness Relative?” Brickman, Dan Coates and Ronnie Janoff-Bulman assembled three groups — people who had won large sums in a state lottery, people who had been paralyzed in accidents, and ordinary controls — and asked all of them to rate their lives.
The answer that flew out of the lab and into every psychology course since: nobody stayed where the event put them. The winners were not walking around in the glow everyone imagines; the accident victims were not living inside the permanent darkness everyone, including themselves, had predicted. Both had drifted back toward the average. Two engines were doing the work. Comparison, because a winner measures the new life against the old one and against someone who won more, and a paralyzed man measures his body against the body he used to have. Habituation, because the new circumstances stop being news. Brickman later framed it with Helson’s adaptation level theory: you judge what you are receiving against what you have lately been receiving, and the reference point quietly follows you.
The idea escaped the lab and went somewhere uglier. It got used to argue that policy cannot improve anyone’s life, that ambition is a treadmill by definition, that trying harder is a category error. A concept for explaining a machine became a reason to stop running on it. There is a name for the corollary, coined by the writer Gregg Easterbrook: abundance denial, the human talent for constructing an elaborate case that what you have is not really enough.
Two of the three mechanisms are worth knowing by name, because they are different animals. Shifting adaptation levels means the neutral point moves while sensitivity to differences survives: you get used to the raise, and the year-end bonus still lands. Desensitization means the wiring itself gets quieter — people who live for years in a war zone stop flinching at what used to stop their breath. Hold onto the third one; it comes back at the end.
🤔 Then the Panel Data Arrived
In 2006, Ed Diener, Richard Lucas and Christie Scollon published a paper in American Psychologist with a title that reads like a demotion: “Beyond the Hedonic Treadmill.” Their opening line is brutally fair to the theory they are dismantling — the treadmill model implies that individual and societal efforts to increase happiness are doomed to failure. Then they list five repairs. People’s set points are not hedonically neutral. People have different set points, partly because of temperament. One person can have several at once, because pleasant emotion, unpleasant emotion and life satisfaction can travel in different directions. Set points can change. And individuals differ in how much they adapt.
The hardest evidence came a year later, from Lucas, working with nationally representative panel data. Study one: just under forty thousand people. Study two: over twenty-seven thousand. Inside those panels, 679 people acquired a long-term disability, and the data let him watch them for an average of 7.18 years before the onset and 7.39 years after it. Happiness fell by somewhere between 0.40 and 1.27 standard deviations depending on the measure. Then it stayed down. There was little adaptation to watch. Whatever the treadmill was doing, it was not catching everyone.
Marriage tells the same story from the other end. Lucas and colleagues used a fifteen-year longitudinal study of more than twenty-four thousand people to look at what happens to life satisfaction around weddings. On average, people reacted and then drifted back toward baseline, which is exactly what adaptation theory predicts: the wedding glow is borrowed happiness with a repayment schedule. But the average is where the finding gets interesting, because the average hides the population. The people who reacted most strongly were still far from baseline years later. Plenty of individual trajectories ran in the opposite direction from what the theory predicted. Marriage can permanently move the line, and a study that only reports the group mean will never see it. Fujita and Diener had already described the set point as a “soft baseline” — steady for most people, and genuinely mobile for roughly a quarter of them.
Read that twice. “On average” is the most dangerous phrase in this entire literature, and it is the phrase that gets quoted.
🔗 What a Promise Can’t Cover
Anything that sells you a peak is selling half a product. A reading, a launch, a promotion, a gift, the day the numbers finally come in — these are peaks, and they are deliverable. The part that decides whether the purchase was worth it is the flat stretch afterward, which is exactly the part nobody sells and nobody asks about.
That is an uncomfortable brief for anyone in the prediction business. A reading that tells you what you will get is answering the easy question. The valuable version answers the harder one: what will this be worth to you in three years, and what in your life is immune to adaptation? The same split shows up in retention curves. When usage sags, part of that is the product and part of it is the customer being a mammal. Distinguishing experience-driven attrition from adaptation-driven attrition is the difference between a roadmap that works and a year of feature work that changes nothing.
Fiction runs into the same wall, which is why happy endings have to stop early. Write three more chapters past the reunion and the reader is watching a couple get used to each other. The characters who stay with us are the ones who did not adapt: the loss still being carried, the want that never got satisfied, the grudge that outlived its reason. A character who recovers from everything the plot throws at her is a character with no fixed point, and readers can feel that even when they can’t name it.
🎲 The Twenty-First Time
Here is the loophole. Frederick and Loewenstein sorted adaptation into three processes, and the third one runs the other way: sensitization, where repeated exposure makes the response stronger. Wine, music, a craft, a difficult problem you keep returning to — expertise is built out of sensitization. The things that deepen under repetition are not immune to the treadmill; they are facing a different machine.
So run the inventory. Write down three things that still land on the twentieth time — a song, a dish, a walk, a person, a type of problem. Then write three that went flat by the third. Most people find their stated goals on the second list and most of their actual pleasure on the first.
You are going to adapt to whatever you get. That part is not optional. The choice on offer is smaller and stranger than it sounds: what you decide to get used to.