Two participation rates moved apart, and that gap is where ageing would enter the arithmetic

Between 2020 and 2025 the American labour force participation rate rose from 61.7 percent to 62.4, a gain of 0.7 points. Over the same six years, participation among people aged 25 to 54 rose from 81.4 percent to 83.6, a gain of 2.2. Same country, same agency, same window. The two rates differ because they divide by different populations, and that difference is where an ageing population would show up in an economy's arithmetic.

Where this is written from

I am an analyst writing inside The Cefa Brief, working from published statistics and public documents and from nothing else: annual series redistributed by FRED and retrieved for this site on 19 September 2026, together with the agencies' own documentation of what those series count. That position sees one thing clearly. It sees what the statistical agencies have chosen to count, at the frequency they publish it, on a national basis, and it can subtract one published figure from another and show the working. It cannot see the subject this piece is named for. Not one of these series counts anybody's age. There is no age structure here, no dependency ratio, no pension or health outlay, no fertility or net migration figure, and no projection of any kind. Six annual observations of a national aggregate are what I have. So I cannot tell you that America is ageing, and I do not. I can tell you where ageing would have to show up in these numbers if it were, and what has in fact happened at those places over six years.

Two rates, two denominators

The headline participation rate counts the civilian population from age sixteen upward and asks what share of it is working or looking for work. It ran 61.7, 61.7, 62.2, 62.6, 62.6 and 62.4 percent across 2020 to 2025. The prime-age rate puts the same question to the population aged 25 to 54 only. It ran 81.4, 81.6, 82.5, 83.3, 83.6 and 83.6 percent over the same years. Both are published by the Bureau of Labor Statistics. Each annual point is an annual average of the published monthly observations; neither is a period total.

Take the ends. The wide rate gained 0.7 points. The narrow rate gained 2.2. The difference between those two gains is 1.5 points. That difference is a composition effect: the wide rate divides by a population that includes everyone over 54, so if the older part of that population grows faster than the rest, and if the older part participates at a lower rate, the wide rate is held down whatever the people inside the narrow band do. Both of those are assumptions, and neither is in these files. Nothing here reports the age structure and nothing here reports participation by age. What the files do say is that the two rates moved apart, and that a rate can sit flat while the behaviour it is meant to describe improves.

The labour input stopped doing so much of the work

Total nonfarm payrolls ran 142.2, 146.3, 152.5, 155.9, 157.7 and 158.4 million jobs across the same six years. Subtracting each year from the one before gives the annual additions: 4.1, 6.2, 3.4, 1.8 and 0.7 million. The addition has fallen in each of the last three observations. The 2025 addition is the smallest in the window. Five steps added 16.2 million jobs in total, and the last of those steps supplied 0.7 million of it. A labour input that adds less each year is the first place a demographic transition would show. Slowing additions have several possible causes, and these series separate none of them.

The other term, and what it produced per head

Output is hours worked multiplied by output per hour. That identity holds by definition, and it is the only place a demographic argument can go. Output per hour in the nonfarm business sector, indexed with 2017 set to 100, ran 109.1, 111.4, 109.8, 112.1, 115.4 and 118 across 2020 to 2025. It fell once, by 1.6 index points in 2022. It gained 3.3 points in 2024 and 2.6 in 2025. Across six years it rose 8.9 points, and 5.9 of those arrived in the last two observations. These files carry no count of hours worked, so payrolls are the only labour input in them, and payrolls count jobs. On that measure the labour input is adding less each year, while output per hour added more in the last two observations than across the three changes before them.

The arithmetic of an ageing economy is not that output must fall. It is that the second term has to carry a share of the load the first term used to carry, and that share rises every year the first term slows. Real GDP per person, in chained 2017 dollars, ran 61,127, 64,758, 66,026, 67,411, 68,681 and 69,749 across 2020 to 2025 and rose in every one of them; end to end the gain is 8,622 dollars a person. Real output itself went from 20.28 trillion chained 2017 dollars to 23.85, a gain of 3.57 trillion, with no down year. One claim the alarm makes is that living standards stop rising. The measure of living standards available in these files rose in all six of the years on record here. That is a statement about six observations and about nothing else. It does not settle 2045. It does establish that the decline has not begun in the data available, and a case requiring it to have begun already is not supported here.

The fiscal half

Fiscal strain is the alarm's other half. Federal debt as a share of GDP ran 122.3, 120.9, 118.4, 117.7, 120.4 and 120.7 percent across 2020 to 2025. The 2025 reading sits 1.6 points below the 2020 reading. The ratio fell for three consecutive observations, rose for two, and ends the window lower than it began. The federal surplus or deficit ran minus 3,096, minus 2,774, minus 1,374, minus 1,687, minus 1,815 and minus 1,775 billion dollars over the same years; the 2025 shortfall is 1,321 billion smaller than the 2020 one. None of that disproves the fiscal case. Six years is a short window, the outlays demographics drive are not in these series, and the ratio's denominator is an economy that grew in every year of it. What the figures do is locate the claim. Anyone arguing that ageing has already destabilised the federal balance sheet is arguing against a ratio that ends below where it started.

One price moved decisively. The ten-year Treasury yield ran 0.89, 1.45, 2.95, 3.96, 4.21 and 4.29 percent across the six years, a rise of 3.4 points. That is the rate at which the government borrows for ten years. These files carry no yield at any longer maturity, so the rate applied to a claim forty years out is not visible here. What is visible is that ten-year money has been repriced across the window, and that repricing is documented in this data set where the demographic path is not.

What separating them leaves you with

The alarm gets one thing right, and it is the thing the identity states. If a population grows older, the hours term slows, and output per head then depends on output per hour to a degree it did not before. That is not a forecast. It follows from the definition of the terms, and no policy can exempt an economy from it.

On timing the alarm is wrong, at least as far as these files reach. Six years of published national data show participation rising inside the working ages, output per head rising in every year, a debt ratio ending below where it began, and output per hour gaining most in the two most recent observations. A case for urgency built on those numbers is a case built against them. The honest position is narrower and more useful. The arithmetic is real and permanent; the alarm is premature on the evidence to hand; and the series that would settle the difference are an age structure, an outlay schedule and a projection, none of which is in this file and none of which is quoted here. A structural view has to survive its own data. Where these nine series run out, so does the argument.