18 July 20266 min readPhilosophy, Decision-making

Someone Has to Lose

There's only so much happiness in the world, one line goes, so you have to steal it from other people. For a specific, nameable slice of what people mean by happiness, the evidence says that line is correct.

I can't find the line. I remember it as You're the Worst: there's not enough happiness in the world, so you have to steal it from other people. I searched the transcripts and the quote databases. It isn't there. The closest documented version is from Peep Show, a decade earlier, Jez to Mark: "There's only so much happiness in the world and they're hoarding it all." Mark's reply, "That's not how happiness works," is the joke, because the audience already suspects it's wrong.

I'm leaving the misattribution in rather than fixing it, because it's the first piece of evidence for the essay. I didn't invent this line. I borrowed it, misfiled it, and carried it around for years as if I'd sourced it correctly. That's the mechanism this piece is about, working on me before I got to the argument.

Here's the argument, stated cold: for a specific, nameable slice of what people call happiness, the line is a correct description. Not all happiness. A slice. I want to be precise about which one, because the imprecision is where every comforting rebuttal hides.

Start with what the line gets wrong. Whether you have enough food, sleep, safety and freedom from pain is not capped by whether I have those things. Global caloric production could rise without anyone's calories falling. My health does not subtract from yours. Kahneman and Deaton's 2010 finding that emotional well-being plateaus around $75,000 in household income got revised in 2023, when Killingsworth, Kahneman and Mellers ran an adversarial collaboration and found the plateau was real only for an unhappy minority; for most people happiness keeps climbing with income well past that number, in some cases accelerating. More for you does not require less for me. On this slice, the line is false, and self-help is right to say so.

Now the part self-help doesn't look at directly. Status is scarce by definition: there is exactly one first place. Being the funniest person in a room is scarce the same way. Clark and Oswald showed this empirically in 1996, using data on five thousand British workers: reported satisfaction moved with income relative to a comparison wage, independent of absolute income. Fred Hirsch named the category in 1976, positional goods, whose value is defined by relative possession, not absolute supply. Robert Frank formalized what happens next, a positional externality, where someone else's purchase changes the baseline your own position is measured against, so status spending spirals upward while making everyone, in aggregate, worse off. For the goods most people actually mean when they say happiness in the bitter, comparative sense, recognition, rank, being wanted, the total is fixed by definition, and someone rising means someone else, measured against that same scale, fell.

That gets you to "some happiness is positional." It doesn't get you to "steal." Competing for a scarce resource and stealing it are different acts, and the line uses the second word on purpose.

René Girard spent his career on a claim that predicts behavior better than it should: desire is not generated internally and then pointed at objects. It's borrowed. You don't want a thing and then discover other people want it too. You want it because they already do. He worked this out first as literary criticism, in Deceit, Desire, and the Novel (1961), then extended it to anthropology in Violence and the Sacred (1972). If I learned to want this from watching you want it, you are the source of my desire, and the moment we both reach for the same object, its obstacle too. That's why "steal" fits better than "compete": competing implies two parties who arrived at the goal independently and happen to be racing for it. Girard's account says you didn't arrive independently. You caught the wanting from someone else, the way you'd catch a cold, and now the person you caught it from stands in the way of the thing they taught you to want.

Girard's darker claim is the one I think the line is actually gesturing at. He argued that societies under enough mimetic pressure, enough people wanting the same things from each other, resolve the crisis by converging on a single victim, not because that person did anything, but because collective relief has to land somewhere, and it's cheaper to aim it at one person than to keep absorbing it as diffuse rivalry among everyone. The scapegoat mechanism, in his account, describes how a specific kind of unhappiness gets redistributed: taken from many people at once and discharged through one. This is Girard's theory of religion and social order, not a settled empirical science, but it names the mechanism the line describes more precisely than economics does. Positional goods explain why the supply is capped. Girard explains why the transfer feels less like winning and more like taking.

In 2009, Takahashi and colleagues ran two fMRI studies, published in Science. In the first, subjects who learned that another person had something superior and self-relevant showed activation in the anterior cingulate cortex, a region associated with the processing of pain. Envy registers, physically, as pain. In the second, subjects watched something bad happen to the person they'd envied, and the region that lit up was the ventral striatum, the core of the brain's reward circuit, the same territory that responds to food, money and drugs. The stronger the earlier envy, the stronger the later reward. Someone else's loss did not just fail to bother the subjects. It registered as a payment, experienced by the reward system the same way winning something is experienced.

I don't want to leave out the finding that cuts against the mood of this whole piece. Rózycka-Tran, Boski and Wojciszke ran a study across thirty-seven countries in 2015 and found that believing life is a zero-sum game correlates with lower life satisfaction, more negative affect, and, at the country level, worse outcomes on civil liberty and human development. People who hold this belief chronically do worse by it. I'm not resolving that tension, because I don't think it resolves cleanly. It could mean the belief is a distortion that makes its holders miserable independent of whether it's true. It could mean people who are already losing on the positional goods described above adopt the belief because it's the accurate read of their situation, and the belief and the misery share a cause rather than one producing the other. What the data does establish is that "this is sometimes true" and "believing it constantly is good for you" are unrelated claims.

Not all happiness is a fixed resource extracted from a shared pool. Enough of it is, the specific, comparative, who's-ahead kind that most people actually mean when they say they want to be happy, that the theft framing survives contact with the evidence better than the abundance framing does. The mechanism isn't malice. It's structural: desire borrowed from a rival, a payoff registered in reward circuitry when the rival falls, a resource that is genuinely capped because rank cannot be shared. Someone doesn't have to be a bad person to be taking something from you by having what you were taught to want. They just have to be standing where the wanting pointed.

I still don't know where the line actually came from. I know what it's describing.

Further reading