Internal increases are calculated from your existing pay and bounded by a budget. Outside offers are calculated from the market. Across the 391 published occupations with at least 50,000 workers, the step from the median wage to the 75th percentile is a median of 26 per cent — a distance ordinary annual increases do not cover. That is a fact about the published distribution, not a promise about any individual move.
Why the claim has real force
Internal increases are percentages of your current salary. External offers are priced against the market rate for the role. When your salary sits below today’s market — which it usually does after a few years — the external route produces a larger number for the same work.
Bands are also refreshed annually against survey data describing a period a year back, so the internal structure is always slightly behind the market it tracks. That gap compounds.
The arithmetic in one example
Somebody hired five years ago at $70,000 with three per cent a year now earns about $81,100. The same role advertised today might start at $88,000. Both receive three per cent next year and the gap widens by another $200.
No annual increase closes that. Closing it requires a step: a market adjustment, a promotion, or a move.
Why the claim is still overstated
Because market adjustments exist as a category and are granted more often than people expect. Replacing you costs recruiting, a market-rate replacement salary and months of lost output, so adjusting you is frequently cheaper — and managers know it.
Most people never ask. The comparison that gets made is between leaving and doing nothing, when the actual choice includes a conversation almost nobody has.
What leaving costs that the salary hides
Unvested equity and employer retirement contributions, which vest no slower than three-year cliff or six-year graded. Any bonus requiring employment on a future payment date. A coverage gap if the new plan starts after 30 or 60 days.
Then the first year: no internal credit, no relationships, no knowledge of where decisions actually get made. That is a real cost even when the number is better.
The pattern that beats both extremes
Alternate. Take an internal promotion when it crosses a genuine level, then reprice externally when internal progression flattens against your band.
People who only move externally accumulate no institutional standing. People who never move accumulate a compounding gap. The mixed approach is what the pay mechanisms actually reward.
Run the internal conversation first
Bring the published median for your occupation in your metro, your employer’s own advertised ranges for your level, and a record of what you delivered. Ask for a market adjustment as a category separate from merit.
If the answer is a clear no with the reason that the band sits below market, you have found your ceiling and the external move is now an informed decision rather than a gamble.
When leaving genuinely is the answer
When the band itself is below market and will not move. When two cycles have passed with criteria met and nothing changing. When the only repricing mechanism the employer uses is a resignation.
Those are structural facts about the organization rather than about you, and no amount of performance changes them.
The honest summary
Leaving is frequently the largest single raise available, and it is not the only one. The people who do best treat it as one instrument among several rather than as the only lever that works.
Which means running the internal conversation properly first — not out of loyalty, but because it is cheap, it sometimes works, and it tells you exactly what you are choosing between.
The move that is neither leaving nor waiting
Changing employer or industry at the same level, which frequently pays more than the next internal promotion, costs less time, and is available now rather than when a position opens.
Most occupations exist across several industries, and the same function is priced differently by who buys it. Moving sideways into a sector that is hiring is one of the largest single moves available and it feels like the smallest.
Timing the departure around what you would forfeit
List every unvested grant with its date, the employer retirement contributions not yet vested, and any bonus requiring you to be employed on a payment date.
Where a cliff or a large tranche sits within a few months, waiting usually beats the increase from moving sooner. Where the next event is a year out and refreshes keep arriving, waiting is an indefinite commitment and the schedule is doing its job.
Common questions
Does changing jobs always pay more?
No. The distribution shows the room exists; it says nothing about a particular move. Plenty of moves are lateral or worse.
Where does the 26 per cent come from?
The median step from the 50th to the 75th percentile across the 391 occupations with at least 50,000 workers in the May 2025 national wage data.
Why can't internal increases close that gap?
Because they are a small percentage of your existing pay, so they compound a starting point rather than resetting it against the market.
What is staying worth?
Context, relationships, vesting, accrued leave, and avoiding the real risk that a new job is worse. None of it appears in a salary comparison.
What should I do first?
Get one genuine offer, whether or not you intend to take it. Until then both sides of the argument are guesses.
Is leaving the only way to get a real raise?
It is frequently the largest single increase available, but market adjustments exist and are granted more often than people expect — most never ask.
Why do external offers pay more?
Internal increases are percentages of your current salary; external offers are priced against today's market rate for the role.
What does leaving cost?
Unvested equity and retirement contributions, any bonus requiring employment on a payment date, a possible coverage gap, and a first year with no internal standing.