The Real Pay Premium of a Target Metro, Scored Against Home

You are deciding which metros to aim your job search at, and you need to know whether a metro's higher pay for your role actually beats what you earn at home once cost of living and taxes are stripped out. This guide is a repeatable scoring model you apply to each candidate metro before you apply, built from the government's own price index, role-specific wage data, and the tax layer that popular calculators skip. By the end you will convert a target metro's nominal pay into a real, purchasing-power-and-tax-adjusted premium over home and land on an Aim, Watch, or Drop verdict you can defend with public numbers.

Why the popular answer to "is the higher salary worth the cost of living" is usually wrong

Most cost-of-living tools answer a different question than the one a job seeker has. They tell you "what salary keeps my lifestyle" from a vendor's proprietary index, then stop. They rarely use the government's own metro price index, they almost never layer in local income tax, and they never amortize the cost of the move itself. So the headline you get compares apples to a different orchard.

The right question for someone deciding where to aim is narrower and more useful: does this metro, for my specific role, leave me with more real money than home after I strip out prices, taxes, and the one-time cost of getting there. That is a scoring model, not a lookup. You apply it to every candidate metro on your list, and it produces a verdict rather than a lifestyle-equivalent salary.

The model rests on three public datasets. The Bureau of Economic Analysis publishes Regional Price Parities, an index where 100 equals the national average price level. The Bureau of Labor Statistics publishes role-specific wages by metro through its Occupational Employment and Wage Statistics program. The Tax Foundation aggregates the local income taxes that state-only calculations miss. None of these are behind a paywall, and none of them require a vendor.

What Regional Price Parity is, and how to convert pay between metros

Regional Price Parity (RPP) is a price-level index published by the BEA where 100 equals the national average. If a metro's RPP is 120, prices there are 20 percent above the national average; if it is 90, prices are 10 percent below. Take the ratio of two metros' RPPs and you have the difference in their price levels.

The conversion is one line of arithmetic. Equivalent pay in a target metro B equals your nominal pay in home metro A multiplied by the ratio of their RPPs:

Real (national-equivalent) pay  = nominal pay / (RPP / 100)

Equivalent pay in metro B       = nominal pay in A x (RPP_B / RPP_A)

Example:
  Home metro A: $120,000 nominal, RPP 100.0  -> real $120,000
  Target metro B: $140,000 nominal, RPP 115.6 -> real $121,107
  Real premium before tax and move: +$1,107 (+0.9%)

Use the "all items" index. It covers all consumption goods and services including housing rents, and BEA notes that housing rents are often the main driver of differences between metros. Housing-rents-only and goods-only sub-indexes exist, but each one understates or overstates total cost. All-items is the correct default, and swapping in a sub-index is one of the traps covered later.

One nuance in your favor. Because rents drive the spread, a reader whose housing cost differs from the metro norm - roommates, a remote arrangement, an inherited property - should expect their real premium to beat the headline. California's rents RPP was 154.3 in 2024 against an all-items figure of 110.7, so the housing component is where an individual's circumstances swing hardest.

Why the statewide RPP misleads exactly where it matters

The single most common error is substituting a state's RPP for a metro's, and it is one-directional: it makes expensive metros look cheaper than they are. For 2024 the San Francisco-Oakland-Hayward metro all-items RPP was 115.6, while California statewide was 110.7. Using the state figure for a San Francisco job understates the local price level by about 4.4 percent, because the state average is pulled down by cheaper inland counties.

The mechanism is worth understanding so you can spot it anywhere. RPP is a population-weighted average across a state's counties. The richest, most expensive metros are a minority of the state's geography, so the state number always sits below the top metro's. That means the state proxy misleads most precisely for the highest-value targets, the ones where getting the verdict right matters.

There is spread inside metros too. Among D.C.-area counties, Arlington County, Virginia is 37 percent more expensive than Madison County, Virginia. BEA finds that 86 percent of metro areas have a price differential of less than 10 percent between their most and least expensive counties, but 54 metros have a differential of 10 percent or more. For most of your comparisons the MSA figure is fine; if you are weighing a specific suburb in a wide metro, check whether your target county diverges.

Getting role-specific pay, not a metro average

The authoritative source for what your role pays in a given metro is the BLS Occupational Employment and Wage Statistics program. It produces annual employment and wage estimates for approximately 830 occupations, covering the nation, all states and territories, and approximately 530 metropolitan and nonmetropolitan areas. It publishes the 10th, 25th, 50th (median), 75th, and 90th percentile wages, plus the mean.

Match on percentile, and match on one SOC code. BLS explicitly does not recommend averaging percentile wages across a combination of occupations, because the percentile for a blend typically will not equal the average of the individual percentiles. If you average two related occupations to build a "median," you have produced a number BLS warns is invalid. Pick the single SOC code that best fits your role and read its published percentiles.

Watch for top-coding on senior roles. OEWS does not release some percentile wages for especially high-paying occupations; for these it publishes only a note that the wage is equal to or greater than a cutoff value. If your 90th percentile reads as a floor rather than a true wage, you are looking at a censored number that understates the real premium. Note it and lean on the 75th percentile instead.

The tax layer most calculators skip

State income tax is table stakes, but local income tax is where "similar" metros separate, and it is the layer most cost-of-living tools omit. Local income taxes are levied by 4,943 jurisdictions in 17 states, and local governments in 16 states tax income in some form, covering over 23 million Americans. Two metros with identical RPPs can differ by 3 to 4 percent in take-home once a city tax applies, a swing large enough to flip a Watch to a Drop.

The named rates you are most likely to hit:

Run each candidate metro against the Tax Foundation's local-income-tax list before you trust a take-home figure. The order in which you apply tax and cost-of-living adjustment does not change the verdict much - some tools apply cost of living first and tax second, others reverse it - but you should document which order you used so the number is reproducible. What matters is that both layers are present.

Relocation is the silent premium-killer

The move itself is a one-time cost that a per-year premium has to pay off, and it is large enough to decide many verdicts on its own. A full-service interstate move averages $4,300 per the American Moving and Storage Association, but that is only the truck. All-in domestic relocation averages around $21,800 for renters and $63,700 for homeowners, with executive moves running well past $90,000.

The dominant line item for a homeowner is selling the home. Realtor commission is typically 7 to 8 percent of the sale price, roughly $35,000 on a $500,000 home. Against a typical annual pay gap of a few thousand dollars, that single cost can consume several years of premium. So the deciding variable is often your expected tenure, not the salary: a $5,000 annual real premium looks like a win until a $35,000 commission erases seven years of it.

Amortize it. Divide the total move cost by the number of years you expect to stay, and subtract that slice from the annual premium before you score. A renter spreading $21,800 over three years carries about $7,300 a year; a homeowner spreading $63,700 carries over $21,000 a year. Either can turn a positive nominal premium negative.

The scoring procedure

Here is the full procedure, in order. Each step names who does it, roughly how long it takes, and what "done" looks like. Work through it once per candidate metro.

For thresholds, calibrate against documented reality rather than guessing. Bankrate's analysis, which adjusts OEWS wages by RPP, found that cost of living cuts the average San Francisco worker's buying power by more than 15 percent, with Los Angeles at 13.4 percent, Seattle at 11.5 percent, New York at 11.1 percent, and Miami at 10.6 percent. San Antonio, St. Louis, Charlotte, and Detroit were the only large metros where the average worker's buying power grew after adjustment.

A workable default: net real premium above roughly 8 percent of home pay is Aim, between 0 and 8 percent is Watch, and at or below 0 is Drop. Set the bands to your own risk tolerance, but write them down before you run the numbers so you cannot move the goalposts to justify a metro you already like.

How this goes wrong: the failure modes

Most bad verdicts trace to one of seven specific errors. Each one is a false positive - it makes a metro look better than it is - and each has a check you can run in seconds. Treat this as the part of the model you return to when a number surprises you.

• Statewide RPP substituted for metro. An SF offer looks 4.4 percent more affordable than it is because California (110.7) is used instead of the SF metro (115.6). Check: confirm you pulled the MSA series, not the state.
• A rent-only or vendor rent index used as "cost of living." A metro with cheap goods but expensive rent, or the reverse, scores wrong. Check: use all-items, which is rent-driven but not rent-only.
• Averaging OEWS percentiles across related occupations. You produce a blended median BLS explicitly warns is invalid. Check: use one SOC code's published percentile.
• Top-coded high-wage occupations. A senior role's 90th percentile reads as a cutoff floor rather than a true wage, understating the premium. Check: look for the "greater than or equal to cutoff" note.
• Local income tax ignored. An NYC or Philadelphia offer nets roughly 3 to 4 percent less than a state-only calculation shows. Check: run the metro against the Tax Foundation local-tax list.
• One-time relocation cost never amortized. A $5,000 annual premium looks like a win while a $35,000 realtor commission erases seven years of it. Check: divide total move cost by tenure and subtract.
• Comparing nominal to nominal. "Higher salary is better" fails when SF's roughly $97,460 mean equals about $77,000 of San Antonio purchasing power. Check: never compare offer-letter numbers directly.

The through-line: every one of these makes a metro look better than the real number would. If your model keeps returning Aim on the priciest metros, suspect one of these before you trust it.

Aim where the jobs actually are for your role

A high real premium is worthless if the metro has almost no openings in your field, which is why the second axis of the verdict is market thickness. Aim where the jobs are is not a single national map; it is role-specific, and the clusters differ sharply by occupation.

In Refolk's index of professional profiles, the US Software Engineer pool totals 348,402 while the Data Scientist pool totals 28,732 - so the engineer pool is about 12 times larger. The two roles also cluster in different places: software engineers concentrate in San Francisco, Seattle, and Sunnyvale, while data scientists concentrate in Los Angeles, the SF Bay Area, and New York. A data scientist and a software engineer weighing the same shortlist should reach different verdicts on the same metro, because the market underneath them is different.

Before you commit to a metro that scores Aim on pay, confirm the market is thick enough to give you real shots. One way to gauge that quickly is to look at who has already made the move you are considering and where they landed. Refolk can find them from public history, so you can see whether your target role and route are well-trodden or rare.

Before you call a metro scored

Run this checklist against each candidate before you write down a verdict. If any item fails, the number is not yet trustworthy.

Keeping the model current

The arithmetic is evergreen; the inputs are not. RPP is released annually - the 2024 figures were released in February 2026 - so re-pull the all-items index each year and confirm you are on the latest vintage for every metro in your comparison. OEWS wages update annually too, with a May reference period, so refresh your role's percentiles on the same cadence rather than carrying a figure from a prior cycle.

Local tax rates change more often and more locally than the price index. Before you rely on a take-home number, re-check the metro against the Tax Foundation local-income-tax list, because a city rate can move without any change in the price index or wage data. Relocation costs are the least standardized input, so treat the averages here as anchors and replace them with real quotes once a specific move is on the table.

The point of a scoring model is that it survives new numbers. When the indexes refresh, you do not rebuild the method; you drop in the new RPP, the new percentile, and the current tax rate, and the verdict updates itself. Keep this document open, keep the checklist honest, and let the public numbers, not the headline salary, decide where you aim.