Take the salary, subtract state income tax to get take-home, then divide by the local price level to express it in constant purchasing power. Do that for both places and compare the results. The number you want is what you would have to earn in the new city to stand still — if the offer beats it the move pays, and if it does not, a larger gross figure is not an improvement.
The quick version, in four steps
An offer in another city is higher by twenty thousand dollars and you have no idea whether that is a raise. The arithmetic that answers it takes about a minute and almost nobody runs it before accepting. It is one subtraction and one division, done twice. What comes out is the only number that compares two cities honestly.
Start with the gross salary and subtract state and local income tax, using effective rates rather than headline ones. Then divide what remains by the local price level, expressed as an index where 100 is the national average. The Bureau of Economic Analysis publishes regional price parities on exactly that scale for every metro. What you get is take-home expressed in constant purchasing power.
Then do the same for the other city, which is the step people skip. Adjusting one side and leaving the other in raw dollars is the single most common error in these comparisons and it always flatters the familiar city. Finally, compare the two adjusted figures directly. If the new offer does not beat standing still, a larger gross number is not an improvement.
Why an index alone will mislead you
A price index averages over a basket of groceries, transport, healthcare, housing and utilities, weighted for an average household. Your household is not average in whichever direction your life actually differs from the mean. The index is built for comparing places, not for comparing your specific situation between places. It is orientation rather than an answer.
Housing is the component with by far the widest variation and the widest range of personal circumstances. Somebody renting a studio and somebody buying a four-bedroom house face completely different multipliers between the same two cities. One index number cannot describe both situations, and it does not try to. That is where the quick version needs a refinement.
Replace the housing component with a real listing
Find an actual place you would live in each city and use its actual monthly cost. That single substitution changes the answer more often than any other adjustment in this whole calculation. It converts an average into your average, which is the entire point of the exercise. It takes twenty minutes on two rental or listing sites.
It also forces an honest conversation with yourself about what you would genuinely accept. People compare a city they know well against an idealized version of one they do not, and a real listing is what stops that. If nothing you would actually live in appears at the price you assumed, that is the finding. Better to meet it now than after signing a lease.
The tax layer people skip
State income tax ranges from nothing at all to over ten percent at the top of the scale. Eight states levy none on wages: Alaska, Florida, Nevada, South Dakota, Tennessee, Texas, Washington and Wyoming. Several cities and counties add their own income tax on top, which almost never appears in a state-by-state comparison. That local layer can outweigh the difference between the two states.
Use effective rates rather than headline ones throughout. Most states have brackets, so the top rate applies only to income above a threshold and overstates what you would actually pay. A payroll calculator that takes your filing status and income will produce the real figure in a minute. The difference between the headline and the effective rate is frequently several percentage points.
How much of a premium survives
How much survives depends almost entirely on your own housing situation, which is why no published figure can tell you. Registered nurses earn a median of $216,740 in San Jose against $97,550 nationally, and that premium does not survive intact. It does not vanish either, which is the mistake people make in the opposite direction. Where it lands turns on whether you rent a room or buy a family home.
That is not a dodge, it is the actual structure of the problem. Two people taking the identical offer in the identical city can end up with genuinely different answers. Anyone quoting you a fixed fraction that survives has averaged over households that include neither of you. Run it with your own numbers and the ambiguity disappears.
The costs that move the wrong way
Cheaper places are not uniformly cheaper, and the exceptions cluster. Two vehicles instead of one, longer distances, higher per-unit prices at smaller retailers, and less competition in local services all push against the housing saving. None of those show up in a salary comparison and several of them are large. They tend to arrive as a surprise in month three.
Childcare and healthcare access are the two that most often catch people out. Both can cost more in time than in money, and time does not appear on any index. A specialist appointment that requires a two-hour drive is a real cost with no dollar figure attached. Price them honestly before deciding the cheaper city is cheaper for you.
The line that is not a cost
Retirement contributions reduce your take-home and increase your assets, which makes them a transfer rather than an expense. Treating them as a cost distorts a city comparison, especially where the employer match differs between the two offers. A four percent match on a six-figure salary is thousands of dollars a year that never appears in a take-home comparison. It is real money and it compounds.
Compare the employer contribution as a percentage separately from everything else. Over a long horizon it is frequently worth more than the difference in base that you are agonizing over. Ask for it as a number at offer stage, since it is an ordinary question. Then set it beside the base difference and see which one is actually larger.
What the arithmetic cannot settle
None of this tells you whether the market for your occupation is deep in the destination. A comfortable net figure in a metro with two employers who need your skills is a worse position than a tighter figure in a metro with fifty. The arithmetic describes this job, and the employment count describes what happens after it. Those are different questions and only one of them is about money.
Check the employment level for your occupation in the destination metro before you check the cost of living. It decides what happens if the job ends, and no amount of monthly surplus compensates for having to move again in eighteen months. It is one lookup on the same table you are already reading. Do it first rather than last.
A worked shape
Suppose an offer is $40,000 higher in an expensive metro. State income tax takes several thousand of that before anything else happens. Housing, priced from real listings rather than an index, takes most of what remains for a family buying and considerably less for one person renting. The same offer is therefore a clear gain for one person and roughly neutral for another.
Both answers come from identical starting numbers, which is exactly why a published index cannot do this for you. The honest version of this calculation is always personal. It is also not difficult, and it takes less time than reading the offer letter carefully. Run it before you accept rather than after you arrive.
Common questions
What is a regional price parity?
A published measure of the local price level where 100 is the national average, built from prices actually collected in that metropolitan area. It covers the whole basket rather than housing alone.
Should I use rent instead of the price level?
Only if housing is an unusually large share of your own spending, and then deliberately. Rent is one line in a budget and using it as a proxy for everything distorts both directions.
Do I need the exact tax rate?
An effective rate is enough for a comparison. Headline top rates mislead in progressive states, and several states have no income tax at all, which is where the largest single differences come from.
What is the most common mistake?
Adjusting the new city and not your own. If you deflate one side only, you have made the move look worse by exactly the amount you failed to adjust.
Does this tell me whether to move?
No. It tells you whether the offer leaves you better off in purchasing power. Commute, career and whether you would like living there frequently outweigh it — but you should know the number before deciding to ignore it.
How do I adjust a salary for cost of living?
Gross, minus effective state and local tax, minus your actual housing cost from a real listing, minus anything else that genuinely changes.
Why not use a cost-of-living index?
It averages over a basket weighted for an average household. Housing has the widest variation and depends on whether you rent or buy and how many bedrooms you need.
How much of a big-city premium survives?
Typically a third to two-thirds after housing and tax — so it neither vanishes nor arrives intact.