Somewhere right now, a consultant is explaining to a leadership team how to manage Boomers, Gen X, Millennials and Gen Z, four species with different values, work ethics and communication needs. The session costs real money. The research it rests on does not exist.
The most rigorous evidence points the other way. Costanza and colleagues' 2012 meta-analysis in the Journal of Business and Psychology aggregated 20 studies with 19,961 subjects across job satisfaction, organizational commitment, and turnover intention, and found the corrected differences between generations small or near zero, and inconsistent in direction. A 2025 update by Ravid, Costanza and Romero in the Journal of Organizational Behavior, combining meta-analysis with qualitative investigation, reinforced the conclusion: generational differences in work outcomes are small and inconsistent, and what variation exists is better explained by age and period effects, where someone is in life and what is happening in the world, than by membership in a labeled cohort.
Generations, in other words, are horoscopes with birth years. This matters for the late career because the generational frame is the polite container in which age stereotypes travel: "Boomers resist change" does managerial work that "older people are worse" could not say aloud.
The real performance question
Clearing away the folklore leaves the actual question: does age mixing help performance? The honest answer is that it works through mechanisms, not magic, and the mechanisms are specific.
The first is complementary error profiles. Experienced workers and early-career workers fail differently: one over-relies on pattern matching from a world that has partly changed, the other lacks the patterns entirely. Teams that contain both, and that talk, catch each other's characteristic mistakes. Teams sorted by age, which is what happens by default when startups skew young and legacy functions skew old, compound them.
The second is transmission. Judgment does not move through documentation; it moves through shared work. An organization whose age groups never work together has no channel for its most expensive asset to flow through, whatever its knowledge-management budget says.
The third mechanism is the cost of the alternative. The OECD's Employment Outlook 2025 is unusually direct: negative stereotypes about aging harm firm performance, and it recommends age-inclusive hiring as an economic measure. The same report notes that only about a third of adults aged 60 to 65 received training in 2023, versus more than half of those 25 to 44. That gap is ageism with a budget line: a standing decision to let part of the workforce depreciate, then cite the depreciation as justification.
From compliance to design
Treating age diversity as compliance produces the familiar theater: a line in the diversity report, a bias training, a photograph. Treating it as performance produces different behavior, because performance questions get designed for.
Design looks like mixed-age staffing as a default, not an initiative; project teams assembled to contain both pattern libraries and fresh eyes. It looks like training budgets audited by age, on the theory that a 62-year-old with five expected years left still outlasts the average job tenure of the 28-year-old beside her. It looks like pay and promotion systems that can reward a lateral, downshifting, or advisory move without coding it as failure, because rigid ladders are what turn every older worker into an obstruction from the ladder's point of view.
There is a measurement discipline hiding here too. If age diversity is a performance question, it should be examined the way performance questions are: with the organization's own data. Which teams hold a genuine mix of tenure and age, and how do their outcomes, error rates, and retention compare? Where do the age-segregated pockets sit, and were they designed or did they accrete? Most firms have never run the query, which is itself the finding: a variable nobody measures is a variable nobody is managing.
And it looks like retiring the generational language entirely. If the meta-analytic differences between cohorts are near zero, then every hour spent tailoring management styles to generational profiles is spent managing a fiction, and worse, rehearsing the stereotypes that the OECD identifies as economically damaging. The evidence-based replacement is boring and effective: manage individuals, mix experience levels on purpose, and keep investing in everyone who intends to stay.
The compliance frame asks: are we legally exposed on age? The performance frame asks: are we systematically underusing the most experienced third of our workforce, and paying for a folklore that tells us it's fine? For most organizations the honest answers are no, and yes.
Sources
- Costanza, Badger, Fraser, Severt & Gade (2012). Generational Differences in Work-Related Attitudes: A Meta-analysis (Journal of Business and Psychology, DOI 10.1007/s10869-012-9259-4). 20 studies, 19,961 subjects: corrected generational differences in satisfaction, commitment and turnover intention are small or near zero and inconsistent. https://doi.org/10.1007/s10869-012-9259-4
- Ravid, Costanza & Romero (2025). Generational differences at work? A meta-analysis and qualitative investigation (Journal of Organizational Behavior, DOI 10.1002/job.2827). Updated evidence: generational differences are small and inconsistent; age and period effects explain variation better than cohort membership. https://doi.org/10.1002/job.2827
- OECD (2025). OECD Employment Outlook 2025. Negative aging stereotypes harm firm performance; only ~1/3 of adults 60-65 trained in 2023 vs more than half of 25-44; recommends age-inclusive hiring. https://www.oecd.org/en/publications/oecd-employment-outlook-2025_194a947b-en.html