You've been drafting the resignation email in your head for about six weeks. Not sending it — drafting it. Revising the second paragraph on the way to work.
At some point the drafting becomes its own activity, and the sentence underneath it is always the same: I need a change.
The feeling is real and it's worth taking seriously. What it isn't, is a diagnosis. "I need a career change" is already an answer — and, like most advice-shaped thoughts, it arrived without the question it's answering. Several genuinely different problems produce that identical sentence, and they don't have the same solution. On our reading, two of them are solved by leaving — and two of them follow you.
Four unhappinesses that sound identical
Here's the sorting exercise, and it's most of the work.
Safety. The job is precarious, or underpaid relative to what your life costs, or the sector is contracting and you can feel it. The background arithmetic runs at night. This is a real problem with a real solution, and the solution often is a move — though it's worth checking whether the gap is income or commitments, because those need opposite responses and both present as money stress.
Esteem. You're competent and nobody can tell, including you. The work is invisible, or credited elsewhere, or so far from anything that produces a result you could point at that your sense of being good at something has slowly gone quiet. This one is frequently misread as boredom.
Belonging. You like the work and you're alone in it. Remote and unanchored, or on a team that has no shared context, or the two people who made the place tolerable both left in the spring. Nothing about the role changed; the room emptied.
Self-actualization, or something near it. The work is fine. You're good at it, paid adequately, on decent terms — and it isn't yours. There's nothing to complain about and nothing that feels like it's going anywhere you'd choose.
Four different thin tiers, one identical sentence. And the point of separating them isn't tidiness — it's that, on our reading, a move solves the first and the last far more reliably than the middle two. Change jobs to fix an esteem deficit and you take your relationship to your own competence with you. Change jobs to fix isolation and you arrive somewhere you know nobody at all, which is worse before it's better.
The forecasting problem
Now the harder part, because this is where career-change decisions get made badly even by people who diagnosed correctly.
Wilson and Gilbert's work on affective forecasting is direct about it: people systematically overestimate the intensity and duration of their emotional reactions to future events (2005). We're reasonably good at predicting whether something will feel good or bad, and consistently poor at predicting how much and for how long.
One cause they identify is focalism — when you imagine a future situation, you focus on the thing that changed and neglect everything else that will also be going on. Applied here, that's the whole trap. When you picture the new job, you picture the new job. You don't picture the commute, the unfamiliar tools, the three months of not knowing how anything works, the manager whose style you can't yet assess, or the fact that your Tuesday evenings will feel much the same as they do now.
So the mental comparison isn't new job versus current job. It's new job, imagined without its problems, versus current job, experienced with all of them. That comparison is not close, which is why the pull can be so strong and so uninformative.
None of that means don't move. It means the pull's intensity is not evidence about the decision's quality, and treating it as evidence is how people arrive somewhere new and find the restlessness waiting for them, on a slightly different salary.
There's a cheap correction available, and it's more useful than any amount of introspection. Find two or three people who made roughly the move you're considering — and specifically people who made it eighteen months ago, not three. At three months everyone is in the honeymoon or the panic, and both are unreliable. At eighteen months the novelty has worn off, the new job's own problems have surfaced, and they can tell you what actually changed and what didn't. Ask them the unflattering version of the question: not "are you glad you did it" but "what did you assume would improve that didn't?" That answer is the part focalism deletes from your own forecast, and it costs you a coffee to borrow it.
And if the restlessness is annual — if this is the fourth September in a row it's arrived, and the previous three passed without a move — that pattern is itself information. A recurring seasonal pull that never converts is less likely to be a career signal than a signal about something the year is missing.
The third career option, between staying and leaving
Between "stay and endure" and "leave" there's a third move that gets skipped because it lacks drama.
Wrzesniewski and Dutton described job crafting: the ways people reshape a role from inside it, across three boundaries — task (what you actually do, and how much of it), cognitive (how you understand what the work is for), and relational (who you deal with and how). Their 2001 paper is a model of how crafting works rather than proof that any given attempt succeeds, but the three boundaries are a usable checklist, and they map onto the tiers neatly.
An esteem deficit is often a task and relational problem — the work you're good at isn't visible to anyone who'd recognize it. That is sometimes fixable by taking on one thing with a legible output, or by getting closer to the people who see the results. A belonging deficit is almost purely relational, and it responds to two lunches a week with people whose company you like far more reliably than to a new employer.
The reason to try crafting first isn't that leaving is wrong. It's that crafting is cheap to test and reversible, and it produces information a resignation can't: if you fix the relational boundary and the restlessness lifts, you've learned the diagnosis was belonging. If you fix it and nothing changes, you've ruled something out for the price of a few coffees, and you can go to market with much better data about what you're looking for.
What a change actually costs, tier by tier
The lens is at its most useful here, because the costs of a move are distributed across tiers that nobody totals up.
Safety. Obvious, partly. Income gap, notice periods, probation, the fact that the new place's finances are opaque from outside. If there's an unemployed stretch involved, it's worth knowing that Lucas, Clark, Georgellis and Diener followed more than 24,000 German adults over fifteen years and found unemployment shifted average life satisfaction in a way that didn't fully reverse even after re-employment. That's a population average and emphatically not a prediction about you — but it's a reason to treat a deliberate gap as a real cost rather than a free interval.
Esteem, and this is the one that rarely gets priced. You are currently senior at something. In the new place you will be a novice — slower than you're used to, asking questions you'd have been embarrassed to ask five years ago, briefly worse at your job than you have been in a decade. It's easy to know this abstractly and still not plan for it. If your esteem tier is already the thin one, arriving into a competence reset is the single worst timing available, and it is a common reason for a good move to feel like a mistake in month two.
Belonging. You are leaving your people. Some of those relationships survive the change and most quietly don't, because they were held together by proximity. That's a real subtraction, and it lands about six weeks in.
Cognitive. A large upfront learning load, which is a gain if that tier was thin and a genuine tax if you're already depleted. The same six months of steep learning reads as the best part of the move or the reason you can't think straight, depending entirely on which of those two states you arrived in — and that's knowable in advance, which is the argument for reading the tiers before handing in notice rather than after.
The pattern worth noticing: a move spends safety, esteem, and belonging in order to buy self-actualization and cognitive room. That's often a good trade. It is a much worse trade if the tier that sent you looking was one of the three you're about to spend.
Four questions before you go
1. Which tier sent you? Name it before you write anything. If you can't, that's the work — the diagnostic takes about twenty minutes.
2. Can crafting test it more cheaply? Task, cognitive, relational. If a low-cost change to one of those would tell you something, run it first.
3. What does the new thing cost in the tiers you're not thinking about? Specifically: how thin is your esteem tier right now, and can it afford a competence reset?
4. Is this goal yours? Sheldon and Elliot found that self-concordant goals — the ones connected to your own values rather than adopted under pressure — attract more sustained effort and pay out more wellbeing on attainment (1999). A career move made because a peer's trajectory looks better than yours is structurally someone else's goal, and the tell is that imagining telling people about the new role is more vivid than imagining doing it.
Probes that cost less than a resignation
Between the audit and the leap there's a set of cheap tests, and they're worth running because they generate the one thing you can't get by thinking harder: contact with the actual thing.
Do a small piece of the work. A freelance project, an evening course, a volunteer version of the role. Not to build a portfolio — to find out whether the day-to-day is what you imagined. This is the single most informative thing available, because it puts you in the ordinary hours rather than the highlights.
Ask for a secondment or an internal move first. It's a smaller version of the same change with the safety tier intact, and organizations grant it more often than people assume, mostly because nobody asks.
Have four conversations with people doing the job now. Ask what the worst part is, and what they'd have wanted to know. People are candid when asked specifically and vague when asked generally.
Each of these substitutes evidence for imagination — and imagination is precisely the faculty the forecasting research says is unreliable here.
Where the frame does not apply
Two things, and the first matters more than the rest of the post.
If the workplace is harmful, leave. Bullying, discrimination, a manager who makes you ill, an environment that has already cost you your sleep or your health — none of this is a diagnosis exercise. Get out, take the imperfect landing spot, sort the career story afterwards. A framework that talks someone into carefully assessing their tiers while they're being harmed has failed at the only thing that mattered.
And restlessness isn't always about work. Sometimes the job is the most visible variable in a life that's thin somewhere else entirely, and it takes the blame because it's the thing you could plausibly change. That's worth ruling out before you resign, not after.
Where NexTier fits, briefly
Career change is the decision where the eight-tier reading earns its keep, because the costs and the benefits land in different tiers, and there's no natural place where they get added up together.
If you want it structured, the assessment reads all eight, and it's worth retaking two months into a new role, if only to find out whether the shape matches what you predicted. And if you'd rather not hand us anything, question one carries most of it: name the tier that sent you looking, then check whether the move you're planning feeds it or spends it.
—
This post is part of a series on personal development organized around what we call Maslow's extended hierarchy — our synthesis of his later work. Sources: T. D. Wilson & D. T. Gilbert, "Affective Forecasting: Knowing What to Want" (Current Directions in Psychological Science, 14(3), 2005); A. Wrzesniewski & J. E. Dutton, "Crafting a Job: Revisioning Employees as Active Crafters of Their Work" (Academy of Management Review, 26(2), 2001); K. M. Sheldon & A. J. Elliot, "Goal Striving, Need Satisfaction, and Longitudinal Well-Being: The Self-Concordance Model" (Journal of Personality and Social Psychology, 76(3), 1999); R. E. Lucas, A. E. Clark, Y. Georgellis, E. Diener, "Unemployment Alters the Set Point for Life Satisfaction" (Psychological Science, 15(1), 2004) — a population average, not a prediction about any individual. This post is educational content, not career, financial, or legal advice.