Start with the published median for your occupation in both metropolitan areas, which gives the gross gap exactly. Then subtract the housing difference at a real address, the state and local tax difference, and the change in commuting and childcare. Finally spread the one-off cost of moving across the years you realistically expect to stay. A move that clears all four is worth taking.
Step one is free and precise
The gap between two cities is published to the dollar. Everything that decides whether it is worth taking is not, which is why this calculation has four steps rather than one. The first step is the free half and it is exact.
Look up the published median for your occupation in both metropolitan areas. Software developers, to take a real pair, have a median of $186,640 in San Francisco and $134,120 in Austin. That is a gap of $52,520, or 1.39 times, and it comes from an employer survey rather than from anybody’s self-reported figures.
Use the same occupation code on both sides. Matching a job title to a different code in each city produces a gap that is mostly a classification artifact, which is the most common way this first step goes wrong. The code is the thing being compared, not the title.
Check how much your occupation actually moves
Geography does not price every job equally, and the range of ranges is the finding. Across states, general and operations managers run 2.35 times from lowest to highest, software developers 1.83 times and registered nurses 1.82 times. Retail salespersons and customer service representatives run only 1.38 times.
Postal service mail carriers barely move at all, because federal pay scales are national. So the first question is not whether the destination city is expensive but whether your occupation gets paid differently there at all. If your answer sits near 1.4 times, the geographic upside is small and the move has to justify itself on something other than pay.
Step two is the one people skip
Housing, at a real address rather than an index. Find three actual listings of the kind of place you would genuinely take, in the neighborhoods you would genuinely consider, and use the middle one. That takes an hour and it is the most important hour in the whole exercise.
An index tells you about the average household in a metropolitan area, which is not you and does not live where you would live. Do the same on the side you are leaving, using what you actually pay now rather than what the market says your place is worth. The comparison has to be between two real numbers rather than one real number and one average.
Step three is arithmetic
State and local income tax, employee-paid state programs, commuting, parking and childcare. Each of these is knowable in advance and each is small enough individually to feel skippable. That is exactly why the total is routinely larger than people expect when they finally add it up.
Childcare is the line that most often decides the answer. It varies enormously between metropolitan areas, it appears in no salary comparison anywhere, and for a household with two young children it can exceed the entire pay gap on its own. Price it at actual local rates for the hours you would actually need.
Step four is the one-off
The move itself, a lease break, a deposit, the gap between final and first paychecks, and anything a relocation package does not gross up. Add those into a single number and then spread it across the years you realistically expect to stay, rather than the number of years you currently intend to.
Two years is the honest planning horizon for most job changes, whatever anybody plans at the outset. A $15,000 net cost of moving is $7,500 a year against a two-year stay and $3,000 against a five-year one. That difference is frequently the entire answer, and it turns on an assumption most people never make explicit.
A worked example
A registered nurse considering a move from Alabama to California is looking at a state median rising from $77,080 to $140,270. The gross gap is $63,190, an increase of 82 percent, and it is the largest such gap in that occupation anywhere in the country. That is a genuinely large number to start from.
Now subtract. California state income tax and its disability contribution, neither of which Alabama levies. The housing difference, which between these two states is the single biggest term and can run to $20,000 a year or more depending on which metros you compare. Then the one-off cost of moving, spread across two years.
What remains is usually still comfortably positive, and that is precisely the point of doing the exercise. An 82 percent nominal increase survives a great deal of subtraction. A 12 percent one frequently does not survive any of it.
Where the arithmetic changes the answer
The moves that fail this test are the middling ones, and they are the reason the test exists. A raise of 15 to 25 percent into a materially more expensive metropolitan area is the dangerous band. The gross figure is large enough to feel decisive and small enough to be entirely consumed by housing and tax.
Run all four steps carefully on any move in that band, because it is where intuition is least reliable and where the published gap is most misleading. Above 50 percent the move usually survives the arithmetic. Below 10 percent it usually is not about the money anyway.
Check the destination market, not just the destination pay
A move is a bet on one employer in a city where you may know nobody. The question worth asking is what happens if that job ends within a year, which is a more common outcome than anybody plans for. In a metro with several employers hiring your occupation, that is an inconvenience rather than a crisis.
Check how many people in your occupation the destination metro actually employs, and whether more than one significant employer exists. A city where your occupation is concentrated in a single company is a much riskier destination than the salary suggests. That risk never appears in a pay comparison and it decides how the worst case plays out. Look it up before you sign the lease rather than after the layoff.
What the published data cannot tell you
There is no published series for what a relocation actually costs. There is no metro-level projection of whether your occupation is growing where you are going. And there is certainly no figure for the value of being near family, or of leaving a manager you cannot work with.
Those things belong in the decision and not in the spreadsheet, and pretending otherwise produces false precision. The arithmetic exists to stop you being surprised rather than to make the choice for you. Once it clears, it has finished its job and the rest of the decision is yours.
Common questions
Where do I get the gross gap?
The published occupational median for both metropolitan areas. Use the same occupation code on each side — matching a title to different codes is the most common error here.
Why not use a cost-of-living index?
An index describes the average household in a metro, which is not you. Use three real listings of the kind of place you would actually take and take the middle one.
How do I handle the one-off cost of moving?
Spread it across the years you realistically expect to stay, not the years you intend to. Two years is the honest planning horizon for most job changes.
Does geography move every job's pay equally?
No. Across states, general and operations managers run 2.35x low to high and software developers 1.83x, but retail and customer service roles only 1.38x, and postal carriers barely move because federal scales are national.
What is usually the largest adjustment?
Housing, then childcare for households with young children. Childcare varies enormously between metros and appears in no salary comparison.
Which moves are most likely to fail the test?
A 15 to 25 percent raise into a materially more expensive metro. Large enough to feel decisive, small enough to be entirely consumed by housing and tax.
Is a bigger city always a raise?
In nominal terms usually, in real terms often not. The pay gap and the price gap are separate numbers and they do not move together reliably.
What can the data not tell me?
There is no published series for relocation cost, no metro-level occupational projection, and no figure for being near family. Those belong in the decision, not the spreadsheet.