Most Claims About Generations Are Age Effects Wearing a Costume
Three different things get called generational — age, period and cohort effects. Only the third is about the generation, and it is the rarest of the three.
Claims about generations confuse three distinct things — age effects true of everyone at that age, period effects true of everyone alive at that moment, and cohort effects specific to a birth window — and most published generational claims are age effects mislabelled.
Three different things get called "generational". Only one of them is.
The three effects
| What it is | Example | |
|---|---|---|
| Age effect | True of everyone at that age, in every era | Teenagers take more risks. They always have |
| Period effect | True of everyone alive at that moment, whatever their age | A recession, a pandemic, an inflation shock |
| Cohort effect | Specific to people born in a particular window | Schools closing during your examination year |
Only the third is about a generation. The first two get the label attached to them anyway, and then get quoted for years.
How to spot the substitution
The test is simple and almost never applied: was the same thing measured at the same age in an earlier cohort?
A survey of 20-year-olds in 2026 cannot, on its own, tell you anything about Generation Z. It can tell you about 20-year-olds. To turn it into a generational finding you need 20-year-olds from 1996 and 2006 measured the same way — and if that comparison exists, it usually shrinks the claim considerably.
Without it, "Gen Z is anxious" and "being twenty is anxious" produce identical data.
Where this bites on this site's own subject matter
"Gen Z is job-hopping." People in their early twenties have always changed jobs more than people in their forties. That is an age effect with a cohort label. And most of Gen Z has not started work yet — the entry window runs to about 2031 — so the sample is the older slice measured at an age when everyone moves.
"Gen Alpha has a shorter attention span." The attention research does not support the headline version, and the part that is real is closer to a period effect: the media environment changed for everyone simultaneously, including adults.
"Screen time is harming this generation." The measured effects are small and contested, and screens arrived for all ages at once — a period effect being read as a cohort effect.
The ones that are genuinely cohort effects
The distinction cuts both ways, and some things really are cohort-specific.
School closures in 2020 hit birth years at different ages by definition — Gen Z was 11 to 23 and Generation Alpha was 0 to 10, so the same event landed on examinations for one group and pre-school for another. That is a real cohort effect, produced by a period event intersecting with age.
Technology works the same way. Generative AI arriving in 2022 was a period event. It becomes a cohort effect only because it reached each birth year at a different stage — twelve for one cohort, before literacy for another.
The reading rule
When you see a claim about a generation, ask which of the three it is:
- Would this have been true of people the same age forty years ago? Then it is an age effect.
- Is it true of everyone alive right now, at every age? Then it is a period effect.
- Does it require having been born in that specific window? Only then is it a cohort effect.
Most published claims fail at the first question, and the failure is invisible because the label sounds like an explanation.
See how the boundaries are drawn at what generation am I, or which cross-generation comparisons survive scrutiny at which generation comparisons mean anything.