The chair came to my office, handed me a letter, and told me to go celebrate. I can still see the office. I can still see the person. The letter said what I had spent years hoping it would say, and the instruction that came with it was kind and sincere and exactly what you say to someone on that day.
I cannot tell you how I celebrated. A nicer dinner, probably. Maybe a bottle of champagne. I’ve tried to recover it and it isn’t there. What I remember instead is that it felt, a little, like business as usual. The thing I had been running toward for years had arrived, and the day went on being a workday.
It would be easy to read that as a failure to appreciate what had happened. I’ve come to think the gap is the most accurate thing I remember about tenure. The moment the uncertainty resolved kept its charge. The enjoyment of having arrived didn’t. And the chair’s instruction contained an assumption neither of us noticed: that the feeling would need to be prompted.
If you’ve crossed any gate in this profession, the qualifying exam, the signature on the thesis, the first offer, the acceptance from the journal you’d have been in awe of as a student, you probably know the shape. Each one forecast, sometimes for years, as an arrival. Each one, within weeks, the floor you’re standing on. The feeling fades. The pace you ran to get there doesn’t.
That second sentence is what this essay is really about. But the first deserves to be understood properly, because the popular version, you’ll adapt to anything, so nothing matters much, is wrong in ways that matter for how you run a career.
The treadmill
Psychologists call this the hedonic treadmill. Our emotional reactions to a change in circumstances, good or bad, fade with time, and we drift back toward where we started. Every gain resets the baseline, so the next lift needs another gain. You run to stay in place.
The mathematician’s version is that the emotional system behaves less like a gauge of level and more like a detector of change. It responds strongly to the derivative of your circumstances and only weakly to where they sit. Put one hand in ice water and the other in hot, then both into the same lukewarm bowl: the water feels warm to one hand and cool to the other. Nothing about the water differs. Only the zero has moved.
Three things move the zero. New circumstances are vivid because they’re new, and whatever stops being new drops into the background. Expectations rise to meet what you have, so the goal becomes the floor. And, the one mathematicians should find most uncomfortable, we adapt to what we can explain. An event that still puzzles us holds our attention and keeps its charge; once we understand it, it gets filed.
In this profession, that last one is the job. A problem is completely opaque until it is known, and then it is obvious to everyone and, somehow, uninteresting. Every mathematician has watched a result they fought with for years look, months later, like something anyone would have seen. In mathematics, explanation isn’t a side effect of the work. It is the work. The field’s product is exactly the thing that dissolves the feeling of having produced it.
What the research actually says
There’s a study that could have been written about my letter. Researchers asked assistant professors at a large university how happy they expected to be in the years after their tenure decision, and compared the answers with former assistant professors who had already been through it. The forecasters expected a denial to leave them unhappy for years. The people who had been denied were happier than forecast, and not much less happy than those who’d been promoted. And the recently promoted weren’t as happy as the forecasters expected to be, either.
It’s a small study, but the pattern has held up: we overestimate how strongly and how long we’ll feel about the outcomes we’re waiting on, partly because we picture the event in isolation and forget the ordinary Tuesday that follows it, when we’ll still be grading, refereeing, and annoyed at the printer. It’s the machinery from the region beta essay, working in the other direction. There, the mind’s capacity to recover didn’t switch on when you’d want it to. Here, it switches on whether you want it to or not, including on the things you worked hardest for.
The popular version overstates this in one important way: not everything returns to baseline. People differ a great deal, too; some rebound fully and some never do. Even the measurement is slippery: if someone rates their life a 7 out of 10 before tenure and a 7 after, it’s hard to say whether they returned to baseline or the ruler moved. And a few things barely fade at all, the ones that stay unpredictable, like noise and a long commute. Academic life has plenty of those.
A staircase of treadmills
Here is the version I’ve actually lived, and it’s about work rather than happiness.
I worked hard as a graduate student. Somewhere in that first year I decided I was behind and that the fix was effort, which I wrote about in The Gap That Closes and the Gaps That Don’t. What I didn’t see then was that the pace I set to catch up would become the floor. I had to work harder as a postdoc, and that reset the level. Harder again as an assistant professor, and it reset again, and again on the way to full professor. Each time, the higher rate stopped feeling like a push and started feeling like simply working.
Picture a staircase where every step is a treadmill. You run at one pace to hold your place. To reach the next step, you run faster. And when you get there, the treadmill on the new step is running faster than the old one.
What makes it hard to step off is the way down. We judge outcomes against a reference point, and losses weigh roughly twice as much as equal gains. Once your sense of a normal week has moved up, working less doesn’t feel like a return to a sensible level. It feels like slipping, like becoming someone who used to be productive. Going up is a gain that fades; coming down is a loss that stings. That asymmetry is the pawl that keeps the gear from turning back. The promotion file adds a second pawl: what you produced in the push to tenure becomes the baseline everything after it is read against. The file is a ratchet with a paper trail.
And part of it isn’t psychology at all. Commitments accumulate at the peak. The students, collaborators, grants, and editorial roles you took on while running flat out are still there when the deadline that justified the pace goes away. Other people are standing on the treadmill with you.
I’ve had a sabbatical, and some parts of the career did ratchet down; the service, most obviously, went away for a while. What didn’t come down was the feeling. I still measured myself against the same standard, just with fewer things in front of me. That’s the most informative thing I know about this. The ratchet has two parts, the load and the reference point, and removing the load leaves the reference point exactly where it was.
Then came an administrative role, and a second treadmill. Now I was running fast on two at once, and the second had a different character. Administration is the commute of academic life: every crisis is new, every personnel case different, and very little of it ever arrives anywhere. It doesn’t just double the running. It adds the kind that never fades into the background, while leaving the first treadmill exactly where the ratchet set it.
High residue, low residue
The reward for good work is more work. You prove things, and they open up more things, which you then have to work harder to prove. You do a good job as an editor and get asked to do more for the journal, and then for another. As far as I can tell from the inside, this only ratchets up.
Those two examples aren’t the same, though. When a proof opens new problems, success creates options: directions you could take, mostly chosen, mostly wanted. When good editorial work brings more editorial work, success creates obligations, assigned by people who’ve seen you do it well, and who, if you make it look easy, tend to assume it costs you little. Most mathematicians I know don’t want off the first ratchet at all. What they want is for the second to stop feeding on it.
Here is the asymmetry I think I’m trying to convey. The feelings adapt in both directions. What doesn’t adapt symmetrically is what each kind of event leaves behind.
A grant rejection hurts, but only for a moment, and then you go back to work on the next one. A paper rejection is annoying and sometimes feels unfair, but it’s data, something to make the paper better for the next submission. Neither leaves much behind. Your obligations are the same afterward, and so are other people’s expectations of you. Failure, here, is low-residue.
Success is high-residue. It moves your reference point, brings new commitments, raises what others expect, and becomes the record the next evaluation is read against. The feeling fades; the obligation stays. So the good events do more lasting work on you than the bad ones, not because success feels bad, but because success is the event that moves the floor. Add the asymmetry from Sounding Sure Is Not the Same Skill as Being Right, and the bad events teach you a lot and commit you to little, while the good ones teach you little and commit you to a lot.
The obligation ratchet doesn’t turn at the same rate for everyone. Women in academia are asked more often to take on committee work and other tasks that don’t count toward promotion, and say yes more often when asked. The reward for good work is more work, but not evenly.
One caution. The setbacks I’ve described fade quickly partly because they’re routine: when funding rates are what they are, rejection is the expected outcome. Not every setback is like that. A second failed job season, or a postdoc ending with nothing lined up, is larger and harder to explain, closer to the events that don’t fully fade, and having been through one before doesn’t seem to soften the next. If you’re reading this from inside one of those, rejection only stings for a moment is a sentence written from the safe side of the table, and you should read it that way.
The field’s treadmill
Some of the treadmill is outside you, and it genuinely moves.
Long ago, a paper on weighted inequalities didn’t need to track how its constants depended on p, or on the characteristic of the weight. Now, if you don’t track that dependence, the paper isn’t seen the same way. Every subfield has its version: an estimate that used to be enough and now has to be sharp. Nothing about my psychology produced that change. The field’s standard rose, and staying competitive means meeting it.
Biologists call this the Red Queen effect, after Through the Looking-Glass, where it takes all the running you can do to keep in the same place, and to get somewhere else you must run at least twice as fast.
Those are two different ways a bar can rise. Some of it is real progress. Once sharp dependence on the characteristic is known to be achievable, a paper that doesn’t track it is leaving information on the table; the bar went up because knowledge went up. Some of it is escalation, standards rising because everyone is competing for the same scarce positions rather than because the mathematics asks for it. From the inside the two feel identical. Only the first is worth embracing without reservation.
And the internal and external treadmills aren’t independent. The field’s standard is the aggregate of what referees, panelists, and hiring committees treat as normal, and those people are senior, each carrying a reference point that ratcheted up over their own career. Individual adaptation, summed over everyone doing the evaluating, becomes the external standard. The junior person meets it as a fact about the field, which it is, even though nobody chose it. It’s the structure from Go Get an Outside Offer, with adaptation manufacturing the standard instead of a belief manufacturing the consensus, and with a generational lag: the people setting today’s bar adapted to yesterday’s conditions and are the least able to feel how high it has become. I include myself.
The lag shows up in advice, too. When I talk with someone applying for jobs or grants, or submitting a paper, I can see the stress, but I’m not sure I really recall what those events felt like. From a calm state it’s hard to feel the pull of an anxious one, even one you’ve been in yourself; it’s sometimes called a hot–cold empathy gap. So I give advice from the facts, then how things seem to work, then my own experience, labeled as that. The person in front of me is in the hot state. Asking them what it’s like is better information than anything I can reconstruct.
It isn’t a staircase
I’ve been using the staircase as though it were a single flight going straight up. That’s how academic careers are usually drawn, and it’s how mine felt. But it’s too simple.
At each step there are many other steps you could take, each with its own treadmill at its own speed. The better picture is the bottom of an empty stadium, with stairways and landings running off in every direction, for a mathematician, a poset rather than a ladder. That includes moves down. Stepping down to take a different path costs more, because it registers as losing ground rather than changing direction. But the stairway is there.
It looks like one flight from the inside partly because, once you’ve adapted to a landing, your attention narrows to the step directly above it. And partly because the stories in circulation all come from the upward flight. The people who stepped down from a role, or moved toward undergraduate teaching and are glad they did, don’t generate stories, for the same reason the quiet cases were missing from the outside-offer folklore.
Adaptation also suggests how to choose among the stairways, and it isn’t by the landings. If every landing becomes your baseline within a year or two, the arrival is a poor basis for choosing a direction, because the arrival is exactly the part that fades. What you’ll live with, every day, is the running. There’s modest evidence that changes in what you do hold their effect longer than changes in what you have, especially when the activity is varied and freely chosen. That fits my career better than anything else I’ve read. Tenure is something I have, and it went flat within weeks. Research is something I do, an unusually varied one, and it’s the one treadmill I’ve never wanted to step off.
So the question isn’t which landing you want to reach. It’s which treadmills you’d be glad to keep running on after the arrival has worn off.
The stadium isn’t equally open to everyone. Graduate students see only a few stairways, and faculty carry the most residue to clear. Postdocs have the least room and the most exposure to the field’s treadmill, but the running-versus-landing question may matter most there, because the job market rewards landings nobody can control. What you can choose is the problem you’d keep working on through a bad season, whatever it does for the application; it’s also the program you’ll still be glad to be running once a job arrives. And when you measure yourself against the CVs in your cohort, remember that the ones you can see are its visible tail and misses those not in the pool.
Choosing the treadmills
Knowing the ratchet exists doesn’t release it, any more than knowing about the spotlight effect makes a room stop feeling watched. Every turn of it is an individually reasonable decision. Saying no matters, but it only blocks the next click; it does nothing about the clicks already taken. Four things work on the rest, and they cost increasingly more.
Audit from zero. Most budgets are incremental: last year’s, plus a change. That’s a ratchet written into accounting. Zero-based budgeting makes every line item justify itself from nothing each year. The personal version: once a year, list what you’re committed to, and ask of each item whether you’d take it on today if it were offered fresh. The ones that survive only because they’re already there are the ratchet’s choices, not yours. For new requests, ask whether you’d say yes if the work started Monday.
Choose the reference point on purpose. This is the one the sabbatical says can’t be skipped. If the standard you measure against is your own peak rate, every reduction registers as a loss, and the load creeps back up to meet it. Set the standard from what the work actually requires, not from the most you’ve ever done, and measure a lighter year against that. It changes what counts as a loss.
Put an end date on what you accept. Nothing ends by default. An editorship, a committee, a standing collaboration continues until someone actively stops it, and stopping feels like a loss. Taking something on with its end attached, three years, then done, reverses the default. Stopping becomes what’s supposed to happen, and continuing is the decision that needs a reason.
And if you have any standing, change the structure. Everything above works on one person’s treadmill; none of it touches the field’s. Make service loads visible, use term limits, and rotate roles deliberately. When you hand off something you’ve done well, check that it’s promotable, or you’ve only pushed the ratchet down to whoever is least able to refuse. And before holding someone’s file to a bar, ask whether your own file, at the same stage, would have cleared it. If not, the difference is either the field’s progress or your own treadmill’s speed, and it’s worth knowing which before it becomes a sentence in a letter.
None of this stops adaptation, and it shouldn’t: some constraints aren’t treadmills you chose, and some of the rising bar is real progress. The goal isn’t getting off; it’s knowing which treadmills you’re on, and why, and whether you’d pick them again.
I’ve told my wife that when I retire, I’ll keep working with my graduate students and postdocs and keep doing the research I want to do. I meant it as a statement about retirement. I’ve come to think it’s the most useful diagnostic I have, and you don’t need to be anywhere near retirement to run it.
Imagine every requirement gone: no degree to finish, no job to land, no promotion to earn, nobody waiting on anything. What would you keep doing? Mine is the students, the postdocs, and the mathematics, and it’s only mine. Some people’s answer would be undergraduates; some people’s would lie outside the department entirely. None of those is the wrong answer. The value is that asking the question separates the treadmills you chose from the ones that arrived as residue from doing something else well.
I still can’t tell you how I celebrated. A dinner, maybe champagne. I can tell you which office I was sitting in when the chair handed me the letter, and I can tell you what I did with most of the years that followed, and it’s the second thing I’d keep. I don’t think the chair’s advice was wrong. Celebrate, it’s the one day the feeling is actually there, and you should spend it. But if I could add a line to that letter for whoever gets the next one, it would be the question underneath the celebration: of everything you’ve been running on to get here, what would you still be running on if nobody asked you to?