TheJobsMarket
What Jobs Pay

Pay by Experience and Level

Short answer

Experience is worth a great deal early and much less later. Pay typically rises fastest in the first five to ten years, then flattens — not because people stop improving, but because internal raises are calculated from an existing salary while starting pay tracks the market. That gap is why changing employer produces larger increases than staying.

The shape of an earnings curve is one of the most useful things to understand about your own occupation, and almost nobody looks at it. Pay rises steeply at first, flattens somewhere in the middle, and in some occupations turns down at the end.

None of that is a judgement on the person. It is the arithmetic of how pay is set.

Why the early years move fastest

Each additional year adds a large proportion to what you can do when you have only had two or three. The tenth year adds much less proportionally than the second, and pay follows the proportion rather than the count.

Why the middle flattens

Two mechanisms, and the second matters more than the first. The marginal value of another year of the same work declines. And internal raises are budgeted as a percentage of your existing salary, while the price of hiring someone new is reset against the market every time. Over several years those diverge, and the person who has stayed ends up below the person who has just arrived.

That is pay compression, and it is the single most common source of a gap that feels unjust and has nothing to do with performance.

Where the curve is not a curve

An occupation's ladder is not the same as its percentile spread. The distance from the 10th to the 90th percentile includes industry, employer size and location as well as experience, so reading the top of the range as "what I get after twenty years" overstates what tenure alone delivers — usually by a lot.

What the data can and cannot tell you

Published wage surveys do not break figures out by years of experience. What exists is the distribution, plus separate series on earnings by age, and neither is a clean answer to "what is another five years worth". This section is honest about which claims rest on solid ground and which are inference.

Articles in this section

Common questions

Does published wage data show pay by years of experience?

No. Wage surveys report the distribution for an occupation, not a breakdown by tenure. Experience is one of several things that produce the spread, alongside industry, employer size and location.

When does pay growth slow down?

For most occupations somewhere between year five and year ten. The marginal value of another year declines, and internal raises key off your existing salary while starting pay tracks the market.

Why do new hires earn more than long-tenured staff?

Because starting pay is reset against the market at each hire while internal raises are a percentage of an existing salary. Over several years those diverge. It is called pay compression and it is not about performance.

Is the 90th percentile what I get after twenty years?

No. It reflects industry, employer size and location as much as experience. Tenure alone rarely reaches it in any occupation.

Does a degree still pay off?

On average and across a lifetime, yes, but the average conceals enormous variation by field, institution and whether the degree is finished. The average is not a prediction about any particular person.

What happens to pay late in a career?

It usually plateaus rather than declines, though measured earnings can fall where hours reduce or people move to different work. Distinguishing a fall in hourly pay from a fall in hours matters here.

Where to go next

What Jobs Pay Around the World

The last piece is how any of this compares outside the United States.

Read it →
CS

Charles Slocs

Data and research

Charles Slocs builds the data side of this site — pulling the federal wage and employment series, matching job titles to occupation codes, and working out what the numbers do and do not support. He writes the pages that are mostly a question about evidence: what a survey measured, how wide the spread really is, and which published figure is out of date.

All articles by Charles Slocs →