The term comes from Samuel Stouffer's work on American soldiers in the Second World War, and the finding that produced it was not what anyone expected.

Military police were promoted slowly and were relatively satisfied with their promotion prospects. The Air Corps promoted quickly and its members were markedly dissatisfied. The objective conditions and the reported satisfaction ran in opposite directions, and the explanation was the comparison: in a slow-promoting branch, not being promoted is normal and carries no information. In a fast-promoting one, it means you were passed over.

Why it is not a flaw

The standard response is that people should judge their circumstances in absolute terms. This is not available, and it is worth being clear about why.

There is no absolute scale. Asked whether an income is good, a house adequate, a career going well, there is nothing to answer against except some reference — your own past, which is the hedonic treadmill, or other people, which is this. Comparison is the mechanism of the judgment, not a bias laid over an otherwise objective assessment.

Which means the instruction to stop comparing is like an instruction to stop using a scale. What is actually available is choosing which scale.

The reference group is the variable

This is the entire practical content, and it is more actionable than it sounds because reference groups are not given.

They are set by proximity. You compare against people you see, which means neighbors, colleagues, whoever is on the same feed. None of that is a considered selection.

They narrow as you succeed. The reference group updates upward with every move, which is why the discontent survives the improvement. Someone who has moved into a more successful room is now at the bottom of it, and the objective gain is invisible from inside.

They are engineered by anyone selling something. An industry exists to set reference points, and the products designed to be visible are designed that way because visibility is what makes them work on other people's comparisons.

They can be chosen. Not by an act of will, mostly, but by changing what you are in front of. The reliable interventions are about exposure rather than attitude — who you spend time with, what you read, what is on the screen. This is also the honest reason so much advice about social media works: not because comparison is bad, but because those feeds are unusually efficient at supplying a reference group composed entirely of other people's best days.

Why it explains so much about money

The income–happiness literature is easier to read through this lens. Absolute income matters a great deal up to the point where basic needs are securely met and then matters much less; relative income keeps mattering, because it feeds a comparison that never resolves.

Which produces the pattern where a whole society becomes richer without becoming more satisfied. Everyone's absolute position improves, nobody's relative position does, and the part that was doing the work was relative. It is also why positional goods — things whose value is largely that other people do not have them — behave so strangely: the satisfaction they supply is consumed by other people acquiring them, which makes the whole category a status game with a receipt.

What it does not license

It is not an argument that material conditions do not matter. Stouffer's soldiers were not imagining their promotions, and people in genuine deprivation are not suffering from a comparison problem. The finding is about the region where needs are met and satisfaction still moves.

Nor is it a counsel to surround yourself only with people doing worse than you, which is the cheap reading. The deliberate version is narrower: notice what you are measuring against, ask whether you chose it, and change the exposure rather than trying to change the reaction.