Social systems of any complexity tend to develop a dominant ordering along some scalar attribute. Dominance hierarchies are near-universal among social species, and informal status rankings re-emerge even in organizations that aim to be flat. Some total ordering is likely to form whether or not everyone intends it. Age is an especially natural attribute around which such an ordering settles, because it is observable, is monotonically increasing, and does not require adjudicating everyone's competence. Three mechanisms then explain why, once the ordering has fallen on age, it is difficult to dislodge.
In-group solidarity. Once a senior cohort exists, its members share an interest in preserving the current ordering, by vouching for one another, coordinating informally around who rises, etc. The ordering is self-reinforcing because the people with influence over the next round of promotions benefited from the previous round's outcomes. Simply put, humans also feel a visceral comfort and kinship with those closer in age; they just relate.
Payscale insurance. A payscale that rises with age is itself a structural insurance arrangement. Younger employees support it because it promises future upside; older employees support it because they are currently drawing on it and will go on to earn the highest salary when they are latest in their careers—exactly the time when personal competitiveness and competence are mostly liable to decline. This means that if older workers manage to hang on even modestly past what a competence-justified tenure would have been, they reap outsized returns. The result is a shared incentive to preserve the age-based ordering that does not require explicit coordination.
Pulling up the ladder (holdup). Older employees accumulate institutional knowledge, relationships, and procedural access whose value depends on the incumbent's cooperation. A senior who is bypassed by a rookie (or under threat of such) can reverse or arrest that rookie's further ascent without direct sabotage (e.g., by withholding context, declining to make introductions, or being slow to respond). The threat does not need to be exercised to shape decisions; its availability is enough to make organizations reluctant to promote against the age ordering in the first place.
Together, these non-mutually-exclusive forces can sustain what looks like a strong equilibrium around the age-based ordering. That equilibrium is common and well-documented; other equilibria remain possible, and different organizations settle at different points along the competence–age continuum. The simulation below represents each force as a separate term with its own parameter, so that the conditions under which the age-based equilibrium dominates, weakens, or gives way can be examined directly.
A note on age vs. tenure. These are different quantities: a 45-year-old hired from a competitor firm has high age but zero tenure at the new employer. The purely firm-specific pressures (institutional leverage, the ladder-pulling mechanism) are carried by tenure; the other two mechanisms operate along both age and tenure, with a real sociopsychological component to the age dimension in particular, since people feel viscerally that something is strange or awkward when they observe stark exceptions to age-based ranking. For clarity, the simulation collapses the two into a single time variable and calls it age, which readers can interpret loosely as biological age, years in the field, or time at the firm.
Each dot is an employee. Their horizontal position is fixed by age (read loosely as years of relevant experience). Their vertical position is rank, which evolves over time. Color shows competence (blue = lower, red = higher). When the dots align along the diagonal, rank perfectly tracks age.
On the word seniority. Many readers will reach for seniority to describe parts of this model. To avoid confusion, I use rank for hierarchical position and age for tenure. The word seniority will not be used again.
Three forces act on each dot's rank every frame:
The key ratio is $\alpha/\beta$. Below roughly 0.5, age-based ordering prevails. Above roughly 2, competence-based ordering takes hold. Most organizations fall between 0.1 and 0.3. However, it very well may happen that $\alpha/\beta > 2$ at some companies. This might correspond to organizations that have structurally committed to performance feedback, and the simulation explores how that ordering can be self-sustained. Full theory →
Each opens the simulation pre-configured and applies a sudden rank reshuffle after one second so you can immediately observe the dynamics. Use the sliders to explore further.