Stop Blaming Generations

The people who publish the charts stopped trusting them

In May 2023 Pew Research Center — whose generational charts are the ones everybody screenshots — announced it would stop producing most of them. The statement is worth quoting directly.

“We'll only do generational analysis when we have historical data that allows us to compare generations at similar stages of life,” wrote Kim Parker, its director of social trends research. And, on the state of the field: “The field has been flooded with content that's often sold as research but is more like clickbait or marketing mythology.”

The reason is not squeamishness. It is arithmetic. Age, period and cohort cannot be separated, because any one of them is exactly calculable from the other two — if someone was 50 years old in 2016, they were born around 1966. Pew's own explainer gives the example cleanly: when Millennials show a lower marriage rate in 2016, there is no statistical way to know whether that is a generational difference or simply what people aged 23 to 38 look like.

This is not a solved problem waiting on better data. A 2019 review in the Annual Review of Sociology opens by noting that despite “hundreds, if not thousands, of articles and dozens of books, there is little agreement on how to adequately analyze” it. The current state of the art is bounds and sensitivity analysis, not estimates.

So when a deck tells you what Gen Z wants, it is comparing people who are currently young to people who are currently not. That comparison can be interesting. It cannot distinguish a generation from a life stage, and almost nobody selling it says so.

"Even if generational effects do not exist, the fact that people believe in and use generational identities means they can have very real and often harmful effects in workplaces."

Ravid, Costanza & Romero, Journal of Organizational Behavior · In a companion review of 137 articles, the same paper found 69% offered generation-based recommendations to organizations anyway.

Two behavioral differences do survive. Gen Z is more willing to act on values: 41% report turning down an assignment over personal ethics and 41% a potential employer, against 37% and 39% of Millennials. These are self-reported lifetime behaviors, not annual quit rates, and Deloitte surveys only the 2 youngest groups, so they say nothing about Gen X or Boomers.

And Gen Z is far more likely to feel constrained. In research by O.C. Tanner, which sells employee-recognition software, 49% of Gen Z say they feel pressured to suppress their values at work, against 26% of Boomers.

That is a difference in permission, not in conviction. And permission is something an employer controls.

The channel gradients are steep and consistent. Preference for monthly giving runs 62% among Gen Z, 58% Millennials, 35% Gen X, 18% Boomers. Digital-wallet share of gifts runs 61% for Gen Z down to 21% for the Silent Generation. Smartphone giving: Millennials 54%, Gen X 43%, Boomers 27%.

And the dollars move opposite to the frequency. In the 2024 giving year, Boomer donors averaged about $3,256 against Gen Z's $867. An earlier wave of the same series found Gen Z spreading its giving across more gifts — 11.4 a year to Boomers' 8.3.

Younger people give more often and smaller. Older people give less often and larger. That is a payments problem, not a commitment problem.

It looks like a generation. It measures a bank balance.

Before accepting that young people give less because they are young people, it is worth removing the money.

The Lilly Family School did the arithmetic on the whole decline. 36% of the fall in household giving participation is explained by declines in income, wealth and home values alone — and 44% of the fall in secular giving specifically.

Then there is a natural experiment nobody set up on purpose. Bank of America and the same school survey affluent American households — income above $200,000 or a million dollars in assets outside the primary home. In that population, households headed by someone born in 1981 or later gave at 79%. Households born before 1981 gave at 87%.

An 8-point gap. In the general population, the equivalent age comparison runs closer to 50 points.

Hold that loosely: these are 2 different studies, 8 years apart, splitting age at 30 in one and 42 in the other, and affluent young people are an unusual group. It is not a controlled experiment. But the direction is hard to argue with, and it is the direction the panel data already pointed.

The life-stage evidence stacks the same way. Homeownership among households headed by someone under 35 fell to 36.3% at the end of 2024, down 1.8 points in a year — the only age group that moved. Marriage and children arrive later than they used to. Every one of those predicts giving, and every one of them has shifted against people currently in their twenties and thirties.

The number that should end the argument

Here is the cleanest way to see what generational survey data actually measures.

In 2024 Blackbaud surveyed about 1,300 Gen Z adults. Among those whose employers offer the programs, 89% said they take part in workplace giving and 92% in workplace volunteering.

Now the transaction logs. CECP's 2025 survey of large employers puts matching-gift participation at 21% and volunteer participation at 25%. Benevity, which processes the transactions rather than asking about them, recorded volunteer program participation of 13.6% in 2025.

These three do not measure the same thing and should not be subtracted from each other. Blackbaud's base is narrow and self-selected, and it does not publish how many people are in it. CECP surveys a couple of hundred large companies a year. Benevity counts its own client base. Each is a different population.

But run it at its most generous to the survey — take the self-report at face value, compare it only to companies that definitely have a program — and stated participation still runs somewhere between 4 and 7 times what anyone measures.

Survey questions measure willingness. Transaction logs measure enrollment. The gap between them is not Gen Z lying. It is the distance between wanting to take part and completing a form — which is exactly where a program can intervene, and exactly what a values strategy cannot touch.

And the technology gap is not where you think

You would expect these differences to show up as a Gen Z technology preference. Mostly they don't.

Deloitte has Gen Z and Millennials at an identical 74% using AI in day-to-day work. Microsoft's 2026 Work Trend Index profiled its most advanced AI users across 20,000 knowledge workers in 10 markets. That group is 50% Millennial, 23% Gen X, 22% Gen Z and 4% Boomer.

Read that carefully: it is the makeup of the advanced-user group, not each generation's rate of advanced use, and it partly reflects how many Millennials are in knowledge work to begin with. What it does establish is that the advanced-AI group is not disproportionately Gen Z — which is the assumption most program plans rest on.

The real cliff is further up the age range. Randstad found 42% of Gen X and 56% of Boomers never using AI at all.

Gen Z's enthusiasm is also falling, not rising. Gallup surveyed 1,572 people aged 14 to 29 in late February 2026. Excitement about AI dropped from 36% to 22% in a year, while anger rose from 22% to 31% — with the steepest declines among daily users. And 58% of Gen Z and 54% of Millennials report digital fatigue from alerts and switching between tools.

Younger employees are not asking for more tools. They are the group most exhausted by the ones they have.