U.S. adults 65+ set to outnumber children by 2029

For most of recorded U.S. history, children have far outnumbered older adults. The U.S. Census Bureau now projects that this will reverse around 2029, when, for the first time, there will be more Americans aged 65 and older than children under 18.The shift is not a forecast about a distant century. In 1950 there were almost four children for every older adult. By 2024 the two groups were about 12 million apart and closing. The latest official projection puts the crossover at the end of this decade. This page documents the convergence using the Bureau’s own counts and projections, with the full dataset available to download.TL;DR — The number of U.S. adults 65+ has risen roughly five-fold since 1950 (12.3 million to 61.2 million), while the number of children under 18 has stayed near 73 million. On the U.S. Census Bureau’s 2023 projections (middle series), older adults overtake children around 2029 — the first such crossover in U.S. history, and earlier than the 2034 date in the Bureau’s superseded 2017 projections. Eleven states and nearly half of all U.S. counties have already crossed that line.The two lines on the chart move for different reasons. The number of older Americans has risen almost without interruption: from 12.3 million in 1950 to 16.7 million in 1960, 25.5 million in 1980, 40.3 million in 2010, and 61.2 million by 2024. As a share of the population, the 65-and-older group grew from 8.1% in 1950 to 18.0% in 2024. The single fastest decade was 2010 to 2020, when the older population grew 38.6% — the fastest rate the Census Bureau records since the 1880s — as the first members of the baby-boom generation (born 1946–1964) began turning 65 in 2011.The number of children has followed a flatter, wavier path. It climbed during the baby boom to about 69.8 million in 1970, dipped through the 1980s, recovered to about 74.2 million by 2010 — around its peak — and has since edged down to 73.1 million in 2024. But because the total population kept growing, the share of children declined over the long run — from a baby-boom peak of 36.0% in 1960 to 21.5% in 2024 (it was 31.3% in 1950, before the boom). The result is the long narrowing visible in the chart: a gap approaching 50 million in favour of children around 1970, down to about 12 million by 2024.Between the 2020 census and the 2024 estimate, the mechanics are stark: the number of older adults rose from 55.8 million to 61.2 million — an increase of 5.4 million — while the number of children held essentially flat near 73 million. The entire closing of the gap over those four years came from the top line rising, not the bottom line falling.The crossover, and why the date movedThe exact year of the crossover depends on which projection is used, and this is worth stating precisely because the widely-quoted figure has changed. The Census Bureau’s 2017 National Population Projections placed the crossover at 2034. Those projections have since been superseded by the Bureau’s 2023 National Population Projections — the first set built on the 2020 census and incorporating post-pandemic fertility and mortality. In the 2023 middle series, the older population overtakes the younger around 2029, five years earlier than the prior estimate. The shift reflects lower projected fertility pulling the under-18 line down faster than the 2017 series assumed.The bracketing figures make the timing concrete. In the 2023 projection, children still lead in 2025 (about 70.9 million under 18 versus 63.3 million aged 65+). By 2030 the order has reversed (about 71.2 million aged 65+ versus 69.1 million under 18). The lines cross in between — in 2029. The earlier date is also more consistent with what has actually been observed: the gap closed from about 17 million in 2020 to about 12 million in 2024, a pace that points to the end of this decade rather than the middle of the 2030s.This is not, in fact, a frontier the whole country reaches at once. As of 2024, eleven states and nearly half of all U.S. counties already had more residents aged 65 and older than children under 18. The national crossover near 2029 is the point at which the rest of the country, on aggregate, catches up to where many states and counties already are.Three forces drive the convergence, and none of them is sudden. The baby-boom cohort is moving through the older age brackets in a single wave — every member of that generation is projected to be 65 or older by 2030. Life expectancy at older ages has risen over the long run, so each cohort lingers longer in the 65-plus group. And fertility has run below the replacement rate of about 2.1 children per woman for years (around 1.6 recently), which holds the under-18 line down. Looking further out, the Bureau’s middle series puts the 65-and-older share at 24.4% by 2060 and 29.1% by 2100, against 17.7% and 16.4% for children.Two ways to read the same chartThe crossover is a single demographic fact, but it supports more than one interpretation, and the data does not settle which is right.Some read this as a slow-moving fiscal pressure. Programs for older Americans — Social Security and Medicare in particular — are funded largely by working-age contributions. As the ratio of working-age adults (18–64) to older adults falls from about 3.4 today toward roughly 2.4 by 2060, the arithmetic that supports those programs tightens, all else equal. On this reading, the chart is an argument for some combination of higher productivity, longer working lives, higher fertility, or higher immigration to widen the working-age base.Others read this as a marker of success rather than a problem. A growing older population is, in large part, the result of people living longer — a public-health achievement, not a failure. On this reading, the United States is also far from an outlier: in 2020 it ranked 34th among countries by the share of residents aged 65 and over, behind Japan (28.5%) and much of Europe, all of which remain high-income economies. The U.S. is aging later and more slowly than most of its peers, and an older age structure is something many wealthy societies already manage.The chart shows the when and the structure of the shift. It does not show which policy response is correct, or whether the trajectory is a concern at all — that depends on assumptions the data alone cannot supply. Is the U.S. path more like Japan’s — older, slower-growing, and still high-income — or a manageable mid-century adjustment? The same chart is read both ways.Counter-arguments and limitationsThe crossover year is a projection, not an observation. Children still outnumber older adults today. The 2029 date comes from the Census Bureau’s 2023 middle series, one of several scenarios. Higher immigration would widen the working-age and child base and could push the crossover later; lower immigration or fertility would pull it earlier. The Bureau itself publishes high-, low-, and zero-immigration variants alongside the middle series for exactly this reason.Counts and shares can tell slightly different stories. This page plots counts in millions, because the milestone — “more older adults than children” — is a statement about numbers of people. The crossover appears at the same point if the two groups are measured as a share of the population, since they share the same denominator. But share-based and count-based framings of “how much the country is aging” can diverge, and both are included in the downloadable dataset.Age thresholds are conventions. “65 and older” and “under 18” are administratively meaningful (retirement eligibility, school age) but arbitrary as biological or economic categories. A 67-year-old in the labour force and a 17-year-old earner sit on the “wrong” side of those lines. The crossover is robust to small changes in the thresholds, but the precise gap is sensitive to where they are drawn.The drivers are not equally certain. The baby-boom cohort’s path through the older brackets is close to fixed — those people are already alive and counted. Future fertility and immigration are far less predictable, and they are what determine how far past the crossover the gap eventually widens.Common misinterpretationsReading the crossover as already having happened nationally. It has not. As of the 2024 estimate, children still outnumber older adults in the country as a whole (73.1 million versus 61.2 million). The crossover is projected, not realised — even though it has already occurred in eleven states and nearly half of U.S. counties.Treating 2029 and 2034 as competing facts. They are not contradictory; 2034 is simply the older estimate. The 2017 projections placed the crossover at 2034; the 2023 projections, which supersede them, place it at 2029. Citing 2034 today means citing a figure the Bureau has replaced.Equating “more old people” with “fewer working-age people.” The working-age population (18–64) is projected to keep growing into the 2050s and then hold roughly level in the middle series — not to contract. What falls is the ratio of working-age adults to older adults. The pressure is relative, not a literal contraction of the workforce.Reading the chart as a forecast of decline. The total U.S. population is projected to keep growing for decades, peaking near 370 million around 2080 in the middle series. The crossover is a change in the population’s age structure, not in its size.Methodology and SourcesWhat the chart shows. Resident population in two age groups — under 18 and 65 and older — in absolute millions, from 1950 to a projected 2060. The shaded area between the lines is coloured by which group is larger: teal while children lead, terracotta after older adults overtake.Actual data (solid lines). 1950–2010 are decennial census counts; 2020 is the 2020 Census (Demographic and Housing Characteristics file: 55.8 million aged 65+, 73.1 million under 18); 2024 is the Census Bureau’s Vintage 2024 Population Estimates (61.2 million aged 65+, 18.0% of the population; 73.1 million under 18,