The argument over American mobility is usually settled with tax records. Raj Chetty and his colleagues gave us the canonical number: a child born in 1940 had about a 92 percent chance of out-earning their parents; a child born in 1984 had roughly a coin-flip. That is the objective story, and it is grim. But there is a subjective shadow to it that almost no one tracks — what Americans believe about their own chances — and in the General Social Survey that belief has been falling just as steadily.
The clearest sign sits in a single question the GSS has asked since 1987. The way things are in America, people like me and my family have a good chance of improving our standard of living — agree or disagree? In 2000, near the peak of the long boom, 78 percent of Americans agreed. By 2024 it was 45 percent. A belief held by more than three in four had become a minority opinion — a 32-percentage-point collapse eudaimonic in the sense that ordinary life is getting better.
That is the protagonist of this piece, and the first chart lets you put it next to its quieter siblings.
The belief that you can get ahead has come apart
Weighted share giving the optimistic answer, by survey year. Bands are 95% bootstrap intervals. Toggle the question to see that the noisy “my kids will do better” item is the one that didn’t move.
Three questions, three fates. The headline item — a fair chance to get ahead — fell furthest and most cleanly. The retrospective one — is your standard of living better than your parents’ was at your age? — held up far longer, peaking at 70 percent in 2004 before sliding to 55 percent in 2024, a gentler 16-point erosion of the cushion that lets people feel they have already climbed. And the prospective one — will your children do better than you?, the question that dominates op-eds and stump speeches — barely trends at all. It bounces between the low fifties and high sixties and lands at 59 percent in 2024, almost exactly where it sat in the late 1990s.
That last point is worth dwelling on, because it cuts against the usual framing. The item everyone reaches for — the fate of the next generation — is the least informative of the three. It is noisy, it skips 2021, and it shows no durable trend. What actually moved is more personal and more systemic: people’s read on whether the country still offers them a fair shot, and their sense of how far they themselves have come.
The subjective curve bends with the objective one
Here is the intellectual payoff, and it is why Chetty rides along on the first chart as a separate, dashed line. His series is not a GSS time series and must never be read as one: it is indexed by a child’s birth year, not by survey date, and it comes from linked tax records, not from anyone’s opinion. Plotted on its own scale, though, the shape is unmistakable. The objective probability that a child out-earns their parents falls from about 0.92 for the 1940 cohort to roughly 0.50 for those born in the early 1980s. Over the same decades that those children reached adulthood, Americans’ belief in a fair chance bent the same way.
One line is built from tax returns and the other from survey answers; they share a downward slope, not an axis. That juxtaposition — felt reality tracking measured reality — is the whole point. The country did not just become less mobile. It noticed.
What the belief is worth
A belief that has collapsed would matter less if it were inert. It is not. Line up the General Social Survey’s global happiness question — evaluative are you very happy, pretty happy, or not too happy? — against where people fall on the fair-chance scale, and you get one of the cleanest gradients in the whole survey.
Believing the Dream is alive is worth about half a point of happiness
Weighted mean happiness (1 = not too happy, 3 = very happy) by belief level. The pale lines split the same respondents into thirds of constant-dollar income — the gradient survives inside every income band.
Among the people who most flatly reject the fair-chance idea, mean happiness sits at 1.78 on the three-point scale. Among those who most strongly affirm it, it reaches 2.30 — a gap of about 0.53 of a point, which on a scale this compressed is large. The same pattern holds for the “I did better than my parents” item, from 1.76 to 2.25.
The obvious objection is that this is just income wearing a disguise: richer people are both happier and more optimistic, so the gradient might be composition, not belief. It is not, or at least not mostly. Split the sample into thirds of constant-dollar income and the slope persists inside every band — even in the bottom third, happiness climbs across the belief scale. A weighted regression makes the same point: each step up the five-point fair-chance scale is worth about 0.11 of a happiness point, and adjusting for log income, age, sex and a college degree trims that by only about 6 percent. Income matters too, with a coefficient near 0.10 per log-point — but it does not explain the belief away.
What this does, and does not, show
Four honest caveats, because this is association, not proof.
The arrow runs both ways. Happier people may simply believe more — optimism and good mood feed each other — so nothing here says that restoring the belief would manufacture the happiness. The regression shows the link is robust to income; it cannot show which way it points.
The kids item is the flat one. The single most-cited mobility question — whether your children will do better — is the least informative in this data. Reporting it honestly means resisting the temptation to fold it into the collapse; it didn’t collapse.
The thresholds are choices. These are ordinal questions, and “optimistic” here means the top two of five boxes — agree or strongly agree, better or much better. The shapes are robust to the cut, but the levels depend on it.
The mode change is real, and Chetty is external. Post-2021 GSS points sit partly on a different measurement footing, flagged throughout. And Chetty’s curve is cited, cohort-indexed tax data — a counterpoint, never a shared axis with the survey series.
On partisanship, a brief note and a hand-off: mobility belief did become somewhat more partisan over these decades, with the Democrat-minus-Republican gap widening in years the Democrats held the White House. That is a real pattern, but it is a sibling’s story — Whose America Is Thriving? tells it in full — and it is not the spine of this one.