The Mid-Career Stall Is a Positioning Problem, Not a Performance Problem
You have had five strong performance reviews in a row. Maybe six. Your manager would fight to keep you. Your projects land, your people stay, your name comes up in the right rooms. And your title has not changed since 2021.
There is a name for that combination and a number attached to it. The mid-career stall, and roughly one in four. In June, NYU's School of Professional Studies and the Burning Glass Institute published Sidetracked, an analysis of more than 1.3 million career histories at the ten-year mark. It defines a stall as five or more years with no meaningful promotion and negligible real wage growth, and it finds 24.2% of mid-career professionals sitting in one.
One in four. Steadily employed, well regarded, going nowhere.
The advice that follows a finding like this is almost always about you. Get more visible. Ask sharper questions in your one-on-one. Build the skills the promoted people have. Some of that helps at the margin. But the report's own data points somewhere less flattering and considerably more useful: the biggest driver of a stall is not how well you work. It is where you sit.
How do you know if you are in a mid-career stall?
A mid-career stall is five or more years without a meaningful promotion and with negligible real wage growth, and it usually becomes visible around the ten-year mark. Performance is not the signal. Stalled professionals average 1.5 internal promotions and 30% wage growth across their first decade, against 1.9 promotions and 71% growth for peers who kept moving.
Most people miss the stall because they check the wrong instrument. Your review measures how well you did the job you have. It says nothing about whether that job leads anywhere. You can be rated "exceeds" for five consecutive years inside a role that has no next role, and every one of those reviews will be accurate and none of them will be informative.
The better instrument is your raise curve, not your raise.
Write down your last five increases in a row. Something like 4.2%, 3.8%, 3.5%, 3%, 3%. Every one of those numbers is defensible on its own. Read together, they describe a company that has decided what you are worth and has stopped revising the estimate. Deceleration is the earliest honest signal you will get, and it shows up years before anyone says the word "stalled" out loud.
Our own numbers are small, but they rhyme with the national picture. Across Modern Compass profiles with a current role on file, 39 in total, average time in that role is just under three years, and nine of those people have held the same title for four years or more. That is 23% against a national stall rate of 24.2%. A sample that size proves nothing by itself. What it tells me is that the pattern walks through our door at roughly the rate the research predicts, which means it is not a rare condition affecting unlucky people. It is a common condition affecting ordinary careers.
[INSERT: data table comparing Modern Compass current-role tenure distribution against the 24.2% national stall rate]
Your seat has a stall rate, and you probably do not know yours
The Sidetracked authors identify three forces behind a stall: occupational structure, credential quality and alignment, and early-career trajectory. Notice what is missing from that list. Effort is not on it. Neither is talent.
Occupational structure is the one worth sitting with, because it is measurable and most people have never measured it. Stall rates vary widely by sector. In public administration the rate is 30.1%. In the information sector it is 20.7%. That is close to a 50% relative difference in your odds of moving, decided before anyone says a word about you.
In product and AI leadership roles at Disney, Salesforce, and Royal Caribbean, every organization I worked in had at least one seat like this. The person who owned the integration nobody else understood. Indispensable, genuinely respected, and structurally unpromotable, because the only role above them required a scope of work their seat never gave them the chance to do. We protected those people. We paid them at the top of band. We did not promote them. If you had asked me to explain it in the moment, I would have said something vague about timing.
I was wrong, and the honest answer was simpler: the seat was terminal, and none of us had ever drawn the org chart in a way that made that visible.
Here is a fast, unglamorous way to check your own. Pull up three current postings for the role you believe comes next, and read the requirements against what you have actually shipped in the last two years. If you want the comparison done for you, paste one of those postings into the free analyzer on the Modern Compass homepage and look at which requirements you already clear and which ones your current seat has never given you a chance to touch. It takes about a minute and needs no account.
The gap you find is usually not a skills gap. It is an evidence gap. You could do the work; you have simply never been assigned any of it, because the seat you occupy is not where that work lives. This is the same mechanism behind why strong directors stall out below VP, and it is why "work harder" is such a poor prescription. More output from a terminal seat produces more output from a terminal seat.
The move nobody recommends: sideways onto a steeper ladder
The report's most actionable finding is also the one that sounds least like advice. Strategic reskilling into an adjacent, higher-mobility role can cut stall risk by as much as 86%. Not starting over. Not a bootcamp, not a master's, not a two-year detour through an industry you know nothing about.
Adjacent.
An adjacent role is one where 70 to 80% of what you already know is a hard prerequisite rather than a nice-to-have, and the remaining slice is learnable inside the first year. The compliance analyst who moves into risk product. A clinical operations manager who steps into health-tech implementation, where the credibility that took her twelve years to build is the whole reason she gets hired. Internal comms into change management for a systems rollout.
Most people skip this option because it reads as lateral, and lateral feels like a loss. Sideways is the wrong mental picture. You are not sliding across a ladder. You are stepping onto a different ladder whose rungs happen to be closer together, and the gain comes precisely from the fact that your existing expertise is scarce over there and ordinary where you are now.
There is a test for whether a move is actually adjacent, and it is not whether the work sounds similar. Ask what the hiring manager is buying. If they are buying your domain knowledge and teaching you their function, that is adjacent and you will be productive in a quarter. If they are buying your function and teaching you their domain, that is also adjacent. If they are teaching you both, that is a career change wearing an adjacent costume, and it will cost you three years, not one.
Worth saying plainly: a good share of these moves are internal. The same company that has no room above you often has an entire department where your fifteen years of context is a rare asset and nobody has thought to ask you. Internal adjacency is cheaper, faster, and dramatically less visible, which is why it goes unclaimed. You will usually have to propose it yourself, and the proposal lands better as a specific problem you would go solve than as a request to be moved.
It costs something real. Expect twelve to eighteen months of being less expert than you are used to being, which is genuinely uncomfortable at fifteen years in. That is the trade. It is also a much smaller trade than a full career change at 35 or older, which is the option most stalled professionals jump to after they have spent three more years waiting.
And this is not a call to do more. It is a call to aim the same effort at a target that pays it back.
The market is doing a good job of hiding your stall
Timing makes this harder to see clearly. July's employment report showed payrolls falling by 23,000, with the unemployment rate ticking down to 4.1% mostly because labor force participation kept sliding rather than because anyone was hiring. Layoffs are historically low. So is movement. Workday's data has promotion rates down roughly 25% from their 2022 peak, with declines across ten of eleven industries and internal hiring down 8%.
When nobody is getting promoted, a frozen title reads as weather.
Some of it is. But weather describes this quarter, and the stall definition runs five years. Two years of a genuinely brutal market followed by three years in a terminal seat looks identical from the inside, and only one of those situations resolves on its own. The compounding cost is not theoretical either: the report puts the fifteen-year earnings penalty for a stalled software developer north of $43,000.
The useful part of the finding is the timing. Warning signs are legible at the ten-year mark, well before a stall hardens into a career shape. That is a window, and it is open right now for a lot of people who think they are just having a slow couple of years.
[INSERT: annotated diagram showing the raise-deceleration curve alongside the promotion-window timeline from year 8 to year 15]
Run this audit before your next review cycle
Three questions. An hour, maybe two, and you will know more about your situation than another year of waiting would tell you.
Where did the last three people in my seat go? Look them up. If two of the three had to leave the company to move up, or if none of them moved at all, the seat is the finding. You are not.
What has my raise curve done? Not the number, the shape. Five percentages in a row, written down where you can see them together for the first time.
Which adjacent role treats 80% of my experience as a requirement? Do not brainstorm this one, search it. Read twenty postings in two or three functions next to yours and notice which ones describe your last decade in their "must have" section. That is your list. If you want structure for the wider version of this exercise, the career audit walks the same ground across more dimensions.
Then give yourself a date. Not "sometime this year." A real one, roughly ninety days out, by which you will have had two conversations with people in the adjacent function and applied to one adjacent role even if you are not sure you want it. The point of applying is not the job. It is converting a vague feeling into evidence about how the market reads your experience, which is information you cannot get from inside your current seat.
The question was never whether you are good at your job. You already know the answer to that, and so does your manager. The question is whether the seat you are good at has anywhere left to take you, and whether you would rather find that out this quarter or in year six.
Future-Proof Your Career
The White-Collar Jobs Most Exposed to AI, and the Gap Hiding in the Data
Anthropic measured what AI can actually do at work, not what it could theoretically do. The 61-point gap between the two is the window mid-career professionals have to reposition before it closes.
