A senior person resigns, citing an offer that "values them properly". The counter-offer is generous; sometimes it's even accepted. And a year later they leave anyway, or stay and disengage. The pattern repeats because the diagnosis was wrong from the start: the problem was rarely the money. One of Elliott Jaques' most practical findings is that pay is felt to be fair when it corresponds to the true weight of the role — and when the feeling of unfairness appears, it's usually the role that has moved, not the market.
Fairness tracks the weight of the role
Jaques' empirical work on what he called felt-fair pay found something stranger and more useful than "people want more money". People carry a strong, consistent sense of what their work is genuinely worth — and it tracks the time-span of the role: how far into the future its decisions have to hold. People carrying long-horizon work feel underpaid when their pay reflects a shorter horizon. And people in the reverse position — paid for a longer horizon than their role really carries — tend to feel quietly uncomfortable rather than delighted. The sense of fairness runs in both directions, which is what marks it as an observation about work, not a negotiating posture.
Roles creep; calibration doesn't
The corporate version of the problem almost always arrives through role drift. Over a couple of years, a capable person accumulates scope — a cross-functional initiative here, a budget there, decisions that used to go upstairs now stopping at their desk. The actual time-span and complexity of the role grows. The title, the grade and the pay stay where they were last formally set. The person now carries the weight of a bigger role with the recognition of the old one — and what they feel is not a number, it's an injustice they can't quite name. "Under-valued" is the word they reach for. A recruiter is simply the first person to agree with them.
Cash counter-offers don't fix felt-fair problems, because the money was never the thing that felt unfair.
This is why the counter-offer fails even when it succeeds. Matching the rival number treats the symptom while confirming the insult — it proves the organisation could have paid properly and chose not to until threatened. The durable response is structural: look honestly at what the role has become, and if it has genuinely grown, recalibrate the role itself — the spec, the title, the pay, all three, openly. The conversation "your role has grown and we're recognising that" lands entirely differently from "we'll match it", because it repairs the actual breach: the gap between the work and the recognition.
The manager once removed
The structural question is: whose job is it to notice the drift before the resignation letter? Jaques' answer is the manager once removed — your manager's manager. Your immediate manager manages your work. The manager once removed manages your career, your role-fit and your felt-fair position, precisely because they stand far enough above the role to see it in proportion — to compare its true weight across several roles at the same level, which no immediate manager can do from one step away. Where that layer is active, calibration drift gets caught early as a structures conversation. Where it's absent, the organisation finds out about role drift the way most do: from an exit interview.
A working test for whether the layer is real: can people name their manager's manager, and can they recall a moment in the last year when that person engaged with them about something beyond the immediate work — their trajectory, their role, their development? If the answer across the organisation is mostly no, then nobody is holding the felt-fair question — and the next "they left for more money" resignation is already in the post.