Few findings in social science have traveled as well as the U-curve. Ask people in almost any decade how happy they are, line the answers up by age, and the same gentle smile appears: high in youth, a sag through the forties and fifties, a recovery into old age. Economists found it in dozens of countries; it even shows up in great apes. The midlife dip became the closest thing happiness research has to a law.
The law is failing — but not in midlife. In the General Social Survey's two most recent rounds, the under-30s report less happiness than every older group. The left arm of the U, the one that was supposed to point up, now points down.
Happiness by age, four eras
Survey-weighted shares within each age group. Each line pools one era of GSS rounds. Hover for exact values and sample sizes.
Through 1989, 12 percent of under-30s called themselves "not too happy" — indistinguishable from their grandparents. In the rounds from 2005 through 2018 it was still 12 percent. In 2021–2024 it is 28 percent. No other age group comes close: among Americans 60 and over, the figure is 20 percent. The share of young adults saying they are very happy fell from roughly 27 percent in the 1970s to 18 percent now.
This is not a story of everyone getting sadder at the same rate. Stack the young directly against the old, decade by decade, and the lines hug each other for nearly fifty years before splitting in the most recent data.
Young vs. old, 1972–2024
Share "not too happy": adults under 30 vs. adults 60 and over, pooled into multi-round periods. The shaded band is the youth excess.
What broke? The timing in this survey is blunt — the GSS skipped 2020 and changed how it interviews people during the pandemic — but the pattern matches what researchers see in other data with finer resolution: a deterioration in young people's reported mental health beginning in the early 2010s and accelerating through the pandemic years, in the United States and across dozens of countries. Whether the cause is economic precarity, the migration of social life onto screens, or something else, the GSS adds a long-baseline confirmation that the change is real, large, and concentrated under 30.
Two honest qualifications. First, the 2021–2024 rounds were conducted partly by web rather than in person, and people admit more unhappiness to a screen than to an interviewer — some of the level shift is likely mode, though mode change cannot explain why the young moved so much further than the old, since the change applied to everyone. Second, the GSS samples roughly 1,400 under-30s across the two newest rounds combined; the era estimates here pool years and never rest on cells thinner than 50 respondents.
The U-curve described a life in which things get better just when they feel hardest. Today's data describe something darker: a cohort starting adult life from the lowest happiness floor the survey has ever measured for its age. Whether they will climb the right side of the U — or drag the whole curve down as they age — is now one of the most consequential open questions in social science.