Are Generational Differences Real
The question people ask about generations is what the young are like. The evidence answers a narrower one: which differences remain when people of the same age are compared across decades. Generational differences turn out to be real but specific, near zero in work attitudes, moderate in leisure values and test scores, and often age or timing in disguise.
Answer to the question
Are generational differences real?
Some are, but fewer and smaller than the labels suggest. A 2012 meta-analysis of 19,961 workers found work attitudes differed little between generations, and a large survey of school seniors found small changes in self-views. Moderate cohort gaps appear in leisure values and in IQ scores, so the real ones are specific, not sweeping.
Key takeaways
- Costanza and colleagues (2012) pooled 20 studies and 19,961 workers and found differences between Traditionals, Baby Boomers, Generation X and Millennials in satisfaction, commitment and intent to leave to be modest to negligible, many near zero.
- Trzesniewski and Donnellan (2010) analysed about 477,000 American school seniors from 1976 to 2006 and found little meaningful change in egotism, self-esteem, happiness or loneliness.
- Twenge and colleagues (2010) compared 16,507 seniors from 1976, 1991 and 2006 and found leisure values up by 0.57 standard deviations between the Boomers and Generation Me, with no difference in altruistic values.
- Duffy (2021) reports Ipsos surveys in which Millennials were called materialistic, selfish and lazy in 2017 and work-centric and ambitious in 2019, once Generation Z was available as an answer.
- Flynn (1984) documented large gains in the mean IQ of Americans between 1932 and 1978, a cohort difference that nobody reads as a trait of character.
In the Ipsos Global Trends survey of 2017, Millennials were described as materialistic, selfish, lazy and arrogant. In 2019, with Generation Z available as an answer, they were described as work-centric, well educated and ambitious (Ipsos data analysed in Bobby Duffy's Generations, pp. 245-248). Nobody had changed in two years, so are generational differences real? They are, but narrowly. They appear in a few measured places, such as IQ scores and the work values of American teenagers, and they are close to absent in the work attitudes on which most workplace advice rests. Most of what gets called a generational difference is a difference of age, or of the year the question was asked.
Age, period and cohort blur
A gap between a 22-year-old and a 55-year-old can have three sources. It can be age, the effect of having lived more years. It can be period, the effect of what is happening in the year of the survey, such as a recession. Or it can be cohort, the effect of having been born into one era rather than another. In a single survey everyone is measured in the same year, so period is fixed and birth year is simply age read backwards. The survey cannot say which of the three produced the gap, and most popular claims about generations come from exactly one survey.
Duffy builds Generations (2021) on keeping the three apart, and the method is to plot each birth group's answers across many years (p. 95). In his Spanish life-satisfaction series, recession dips hit every generation at once, which marks them as period effects, and no cohort gap persisted (p. 95). The same pattern read from one year's snapshot would have looked like a difference between the old and the young.
The cheaper repair is a time-lag design, which compares people of the same age at different dates. That removes age and leaves cohort and period tangled, and Jean Twenge and her colleagues say so in their own 2010 paper. A 2015 review in Industrial and Organizational Psychology found little solid empirical evidence for generationally based differences at work, and little theory to explain why there should be any.
Twenty studies, almost no gap
The largest pooled test of generational differences at work is a 2012 meta-analysis led by David Costanza. It combined 20 studies, 18 pairwise comparisons and 19,961 people across four groups, Traditionals, Baby Boomers, Generation X and Millennials, and asked whether they differed in job satisfaction, organisational commitment and intent to leave.
Single comparisons ranged from 0.02 to 0.25 in standard deviations for satisfaction, from -0.22 to 0.46 for commitment and from -0.62 to 0.05 for intent to leave. The authors summarised the relationships as modest to negligible, with many near zero. They concluded that meaningful differences among generations probably do not exist on those variables, and suggested that interventions aimed at generational differences may not work. Each comparison sets one group's average against another's, so even a real gap would say little about any single employee.
A 2021 review in the Journal of Business and Psychology, Generations and Generational Differences: Debunking Myths in Organizational Science and Practice and Paving New Paths Forward, begins from the same place. Its abstract calls the evidence for generations at best scant, sets out to "bust" ten common myths about them, and recommends a lifespan view of how age operates at work in place of a generational one. Duffy reaches the same place from outside the journals. The claim that Millennials hold a 12-second attention span and Generation Z an 8-second one traces, in his account, to a Microsoft Canada report that cites Statistic Brain, which gives no data (p. 84).
One survey, read two ways
Monitoring the Future surveys national samples of American high school seniors, which makes it a natural time-lag design: seniors in 1976 can be set beside seniors in 2006 at the same age. Two research groups have used its data from 1976 onward and read it differently.
The first analysis, by Kali Trzesniewski and Brent Donnellan, covered about 477,000 seniors from 1976 to 2006. It found little meaningful change in egotism, self-esteem, locus of control, happiness, loneliness or antisocial behaviour. Later seniors were more cynical, less trusting, less fearful of social problems and held higher educational expectations, and the authors noted that the cynicism mirrored the wider population, a pattern that fits a period effect better than a cohort effect. The effect sizes were small.
The second asked a different question of the same survey: what do seniors want from work? Jean Twenge and colleagues used 16,507 seniors from 1976, 1991 and 2006 and compared Baby Boomers, Generation X and what they call Generation Me. Leisure values rose across the three, by 0.57 standard deviations between the Boomers and Generation Me, which is moderate. Extrinsic values peaked with Generation X and stayed above the Boomers' for Generation Me (0.26). Intrinsic values (-0.20) and social values (-0.28) were lower for Generation Me than for the Boomers, and altruistic values did not differ, contrary to press reports of a more altruistic generation.
The design has a stated weak point and a stated defence. Boomer and Generation X values were measured decades ago, so they may have shifted with later working life; the authors answer that work values form early and stay fairly stable into early adulthood, which makes the school years a reasonable baseline. They also report group averages, which describe no individual senior.
| Account | Data | Reading | Its own limit |
|---|---|---|---|
| Costanza and colleagues, 2012 | 20 studies, 19,961 workers | Differences near zero in satisfaction, commitment and intent to leave | Covers only those three variables |
| Trzesniewski and Donnellan, 2010 | About 477,000 seniors, 1976 to 2006 | Little change in self-views, small effects | Self-views, not work values |
| Twenge and colleagues, 2010 | 16,507 seniors in 1976, 1991, 2006 | Leisure up by 0.57 standard deviations, intrinsic and social values down | Values measured at school, not at work |
| Duffy, 2021 | Ipsos survey data, 30 countries | Stereotypes follow the newest cohort | Company-reported data, his own analysis |
The two seniors' studies differ more in what they measured than in what they found. A cohort can plausibly differ in what it says it wants at 18 and still match its elders in satisfaction and commitment on the job, and the data here do not connect the two. Both also measure what people say about themselves, not what they later do.
Stereotypes follow the young
The reversal that opens this piece is Duffy's own evidence on stereotypes. It comes from Ipsos data, including a 30-country survey commissioned for the book and the Global Trends results for 2017 and 2019 (pp. 245-248). The Millennials in the 2019 answers were the people described in 2017. The slot they had been occupying had changed.
His reading is that the traits are age-based clichés about the young, which transfer to whichever cohort is newest. He treats the habit of judging the latest young cohort as worse than any before as a constant, and not as a finding about any one of them. The figures are Ipsos publications and his own analysis, so they are company-reported data and not an independent test, but the pattern is the one a cohort theory cannot easily produce. If the traits belonged to a birth group, they would have stayed with it.
The people had not changed in two years. The slot they had been occupying had.
For anyone reading a claim about a generation, this gives a cheap check. If the description would fit whichever group is youngest, it is probably a description of being young. [Your Nostalgia Is Lying to You](https://thewhyweproject.com/articles/your-nostalgia-is-lying) weighs how reliable the memories behind such comparisons are.
Where cohorts really differ
If the description would fit whichever group is youngest, it is probably a description of being young.
The standing example of a real cohort difference is the Flynn effect. In 1984 James Flynn reported large gains in the mean IQ of Americans between 1932 and 1978, and later work found a rise of roughly three points per decade in many countries. Each later-born group scored higher than the one before on the same kind of test at the same age, which is the signature of a cohort effect. It is rarely described as a trait of a generation, because a test score is rarely read as a character judgement. The usual explanations are environmental.
Duffy argues that the cohort gaps which matter most are economic. Income growth has stalled for younger cohorts while housing wealth has accrued to older ones, so what separates a 30-year-old now from a 30-year-old in 1990 is closer to housing than to character. The figures behind that argument are survey and official series that this piece has not rechecked, so it stays as his argument. He reads much of what looks like Millennial rejection of marriage, children and early independence the same way: as delay tied to money and stage of life, with the milestones arriving later and not being refused.
What these real differences share is a measured quantity, taken at the same age in different decades, with the same instrument. That is a demanding standard, and claims about Millennials and Generation Z are often made without it. How to read a psychological claim after the replication audit is the subject of [The Psychology You Learned Didn't Replicate](https://thewhyweproject.com/articles/the-psychology-you-learned-didnt-replicate), and the argument over screens and young people is set out on both sides in [Both Sides of the Screen-Time Argument](https://thewhyweproject.com/articles/screen-time-both-sides).
One comparison settles it
Three tests sit inside one comparison, and each rules out one of the three effects. If a claim would hold for people aged 24 whatever year they were born, it is about age. If everyone alive in that year would show it, whatever their age, it is about the period. Only when a group of the same age has answered the same question differently in an earlier decade has a cohort effect been shown, and claims about Millennials and Generation Z are often made without that comparison.
The leisure result in the school seniors' data is one that passes, since seniors were compared at the same age in 1976, 1991 and 2006. Its authors accept that a change affecting everyone in those years cannot be fully excluded, so it is worth believing at the size it was found, a moderate rise of 0.57 standard deviations between the Boomers and Generation Me.
Common questions
Are generational differences real?
Some are, but fewer and smaller than the labels suggest. A 2012 meta-analysis of 19,961 workers found work attitudes differed little between generations, and a large survey of school seniors found small changes in self-views. Moderate cohort gaps appear in leisure values and in IQ scores, so the real ones are specific, not sweeping.
Why do generational differences look bigger than they are?
A single survey cannot separate age, period and cohort, so a gap between young and old is easy to credit to birth year. Stereotypes add to it. In Duffy's Ipsos data the traits attached to Millennials in 2017 had shifted by 2019, once Generation Z was available as an answer.
What is the difference between an age effect and a cohort effect?
An age effect appears in every cohort as its members reach the same stage of life. A cohort effect belongs to a birth group and stays with it, such as the rise in IQ scores that Flynn documented. Only repeated measurement of the same ages across decades can tell them apart, and a single survey cannot.
Should managers plan around generational differences?
The 2012 meta-analysis found little to plan around. Its authors concluded that meaningful differences among generations on satisfaction, commitment and intent to leave probably do not exist, and suggested that interventions aimed at generational differences may not be effective.
Sources
Costanza, D. P., Badger, J. M., Fraser, R. L., Severt, J. B., & Gade, P. A. (2012). 'Generational differences in work-related attitudes: A meta-analysis.' Journal of Business and Psychology, 27(4), 375-394. doi:10.1007/s10869-012-9259-4
Costanza, D. P., & Finkelstein, L. M. (2015). 'Generationally based differences in the workplace: Is there a there there?' Industrial and Organizational Psychology, 8(3), 308-323. doi:10.1017/iop.2015.15
Rudolph, C. W., Rauvola, R. S., Costanza, D. P., & Zacher, H. (2021). 'Generations and generational differences: Debunking myths in organizational science and practice and paving new paths forward.' Journal of Business and Psychology, 36(6), 945-967. doi:10.1007/s10869-020-09715-2
Trzesniewski, K. H., & Donnellan, M. B. (2010). 'Rethinking Generation Me: A study of cohort effects from 1976-2006.' Perspectives on Psychological Science, 5(1), 58-75. doi:10.1177/1745691609356789
Twenge, J. M., Campbell, S. M., Hoffman, B. J., & Lance, C. E. (2010). 'Generational differences in work values: Leisure and extrinsic values increasing, social and intrinsic values decreasing.' Journal of Management, 36(5), 1117-1142. doi:10.1177/0149206309352246
Flynn, J. R. (1984). 'The mean IQ of Americans: Massive gains 1932 to 1978.' Psychological Bulletin, 95(1), 29-51. doi:10.1037/0033-2909.95.1.29