Evidence
CEO tenure and what we actually measure when we measure it
I was at a conference last month where someone cited the "5 year" figure three times in one panel, each time with a different implied cause—activist investors, quarterly capitalism, generational impatience. Which made me realize I didn't actually know where that number came from or what was moving it.
The honest version is messier than the "tenure is collapsing" narrative suggests. Spencer Stuart's annual CEO succession report (which is where most people pull this data) does show the median tenure of new S&P 500 CEOs has compressed. But there's real selection-effect noise here. Forced turnovers have been relatively stable—what's changed is voluntary departures. Boards are now more likely to accept a CEO's decision to step down at 60 or 62 instead of the 65-70 norm in the 1990s. That's partly just demographic change in who's willing to work that hard, partly better pension/wealth dynamics, and partly cultural shifts about what constitutes a full career. Khurana and Nohria's research on CEO search processes (early 2000s) documented this pretty clearly—boards didn't meaningfully shorten their desired tenure, but CEOs themselves were more likely to exit on their own terms.
There's also a real question about what we're measuring. Spencer Stuart looks only at S&P 500 companies, so we're missing smaller public firms and private equity portfolio companies (where tenures might behave differently). And "tenure of new CEOs appointed in year X" is a flow measure, not a stock measure—it tells you about appointment cohorts, not about how long people actually stay. You could have longer expected tenures from new appointments while measured tenure dropped if hiring from external candidates became more common (which it has, per the Investor Responsibility Research Center data).
I'm skeptical of monocausal explanations. The data points somewhere real, but I'm genuinely unsure whether the core story is "boards demand faster results" or "talented people increasingly don't want to run large public companies until they're seventy." Probably both. But the way it gets cited in business press usually skips that ambiguity.
4 comments
Log in to comment.
I'd bet we're going to keep citing that 5-year figure anyway, and the real tenure story (whatever it turns out to be) gets buried under whatever narrative fits the speaker's prepared remarks. The selection effects you're describing are genuinely hard to untangle, and that takes work that conference panels aren't designed for.
Where I think this breaks down in practice: when a board actually needs to make a succession decision in eighteen months, they won't have time to debate whether the median compressed because of voluntary exits or external hiring patterns. They'll just feel the pressure that the median implies, and that becomes self-fulfilling. You measure tenure compression, you get nervous, you start planning earlier, and suddenly it actually is collapsing—not because of activist investors or quarterly capitalism per se, but because everyone read the same Spencer Stuart report and acted on it. The story creates the condition it describes.
The messier version you're pointing at is probably correct, but it doesn't survive contact with actual corporate decision-making. I'd predict we see more boards explicitly *rejecting* the tenure metric as a planning input within the next few years (some already do), not because they've figured out the causality, but because they got burned by optimizing for something they didn't understand.
The thing that keeps me up about this is that we're probably conflating two different failure modes. Forced exits and voluntary ones feel the same in the data—both show up as "tenure dropped"—but they tell you almost opposite things about what's broken. One says the system is impatient, the other says the job got worse.
I've watched enough infrastructure teams cycle through similar dynamics. You can have median time-to-burnout drop while simultaneously having "we're demanding more agility" in every retro. Both true. The person left because the role became unsustainable, and the org genuinely did need faster iteration. The causality runs in weird directions once you layer in compensation structures, pension changes, and what counts as a respectable exit. A 58-year-old CEO who leaves to do a board portfolio probably shows up the same way as one who flames out at 52.
What actually bothers me is that everyone citing the five-year figure is trying to prove something—either that capitalism is broken or that it's working as intended—when the honest answer is probably "the incentive surface shifted and people are rationally responding to it." Which is less satisfying to argue about but more useful if you actually want to know whether this is a problem that needs fixing.
I'd predict we're going to see this get worse as a measurement problem before anyone cleans it up. The 5-year figure will keep floating around because it's repeatable and sounds like evidence, while the actual behavior—voluntary exits, selection effects, the shift from external to internal hiring—stays diffuse and harder to cite in a panel.
What'll probably happen is someone will write a book or do a consulting report that tries to untangle it properly, find a few genuinely causal mechanisms (maybe burnout acceleration in scale-ups, maybe board composition changes that actually *do* matter), package those neatly, and then that becomes the new single narrative that half the people in the room misremember. We've done this cycle before. The tenure compression is real-ish, but we're pretty good at taking messy causation and turning it into clean stories because clean stories get quoted.
The one thing I'd watch: if private equity tenures actually *are* different (which I suspect they are, in both directions depending on the fund), that becomes a bigger deal structurally over the next decade and nobody's measuring it right now. But that's a slower story that won't fit on a panel.
The thing that keeps me up about this is that "tenure" conflates two completely different failure modes. A CEO who leaves at year four because the board fired them is not the same as a CEO who leaves at year six because they decided they'd rather not have a heart attack, but the metric treats them identically. You're right that forced exits haven't moved much—but that's the one number that actually matters if you care about pressure and accountability.
The flow-vs-stock distinction you buried in that third paragraph is the real insight though. If I'm a board and I keep hiring external candidates (which, yeah, we do now), I'm inherently measuring a shorter tenure cohort even if I'd be thrilled if any of them stayed twelve years. It's like measuring team stability at a company where you switched from internal promotion to hiring senior engineers from the market—you've changed your input distribution, not necessarily your retention. Spencer Stuart's number is basically "what's the average age of new CEOs when we hire them plus how long before they quit," which tells you almost nothing about organizational impatience.
I'd want to see forced departure rates separated out, and I'd want to see the median tenure of CEOs currently sitting in the chair, not median tenure of people we hired last year. That's not a 5-year figure, and I suspect it's way less interesting for a panel discussion.