Two people with the same job title, the same experience, and the same employer type can earn very different amounts simply because they live in different cities. But how much your address matters depends enormously on what you do for a living. For some occupations, moving across the country can reshape your paycheck. For others, pay is close to standardized nationwide. We measured the gap for every major US occupation - and the results contain a genuinely counterintuitive finding.
How we measured it
For every detailed occupation with a published median wage in at least 100 of the 150 largest US metros, we measured the spread between the 10th and 90th percentile of those metro medians. The ratio of the two - p90 ÷ p10 - is a "geography multiplier": how much more the same job pays in a well-paying metro than a poorly-paying one, using percentiles rather than the absolute lowest and highest metro so a single unusual data point can't distort the ranking. 420 occupations met the 100-metro threshold; we excluded 28 catch-all categories like "Physicians, All Other" and "Teachers and Instructors, All Other" (title contains "All Other"/"Other," or SOC code ending 9199/9099/9999), since their job mix can differ metro to metro and their spread reflects that composition as much as any real geographic pay gap. That leaves 392 occupations ranked below.
Three caveats. These are nominal wages, not what the money buys locally - our guide on nominal vs. real wages covers that distinction. For occupations typically paid hourly - waiters, bartenders, substitute teachers among them - BLS derives the "annual" figure as hourly wage × 2,080 hours, a standard full-time year, so these numbers reflect the pay rate more than actual earnings for anyone working fewer hours. And state tipped-minimum-wage law varies enormously, which drives a meaningful share of the spread for tipped occupations specifically - a pay-law difference, not necessarily a difference in what the work is worth.
The jobs where location matters most
| Rank | Occupation | Metros with data | Typical metro median | Spread (p90 / p10) |
|---|---|---|---|---|
| 1 | Substitute Teachers, Short-Term | 138 | $37,670 | 2.16x |
| 2 | Firefighters | 147 | $59,850 | 2.13x |
| 3 | Waiters and Waitresses | 150 | $34,900 | 2.12x |
| 4 | Bartenders | 150 | $33,835 | 2.06x |
| 5 | Cardiovascular Technologists and Technicians | 135 | $74,050 | 2.06x |
| 6 | Tax Examiners and Collectors, and Revenue Agents | 114 | $69,450 | 2.01x |
| 7 | Structural Iron and Steel Workers | 114 | $67,935 | 2.01x |
| 8 | First-Line Supervisors of Firefighting Workers | 144 | $85,210 | 1.98x |
| 9 | Telecommunications Line Installers and Repairers | 131 | $65,210 | 1.96x |
| 10 | Health Specialties Teachers, Postsecondary | 121 | $97,810 | 1.96x |
Substitute teaching tops the list at 2.16x, ranging from $21,520 in Knoxville to $78,840 in Anchorage - the same Anchorage premium for place-bound work we found in our Anchorage guide. Waiters and bartenders land high on the list too, largely for the tipped-minimum-wage reason noted above rather than skill or demand. Firefighters, structural iron workers, and telecom line installers are a more straightforward story - skilled, physically local trades where high-cost, high-demand metros bid pay up sharply above smaller or cheaper ones. These are likely explanations, not proven causes.
The jobs where your city barely matters
| Rank | Occupation | Metros with data | Typical metro median | Spread (p90 / p10) |
|---|---|---|---|---|
| 383 | Cost Estimators | 150 | $77,515 | 1.22x |
| 384 | Pharmacists | 150 | $139,200 | 1.22x |
| 385 | Shipping, Receiving, and Inventory Clerks | 150 | $45,030 | 1.21x |
| 386 | Transportation Security Screeners | 107 | $63,770 | 1.19x |
| 387 | First-Line Supervisors of Transportation Workers | 150 | $61,815 | 1.18x |
| 388 | Physical Therapists | 150 | $102,300 | 1.17x |
| 389 | Stockers and Order Fillers | 150 | $37,015 | 1.17x |
| 390 | Postal Service Mail Carriers | 145 | $61,130 | 1.10x |
| 391 | Postal Service Clerks | 145 | $62,650 | 1.08x |
| 392 | Postal Service Mail Sorters and Processors | 117 | $58,470 | 1.08x |
The single most location-independent salaries in America belong to the federal government. The three lowest spreads in the entire ranking are all Postal Service occupations, and Transportation Security Screeners - also federal - sit close by: federal pay follows a single national scale with regional adjustments far smaller than the open market produces elsewhere. The rest of this list is heavily licensed or insurance-reimbursed professional work - pharmacists, physical therapists - where pay is set against standardized rate schedules that don't vary much by metro either.
The five jobs from our city guides
Here's the finding this study set out to check, using the same five occupations behind our Nashville, Syracuse, Fayetteville, and Anchorage guides:
| Occupation | Spread | Rank (of 392) |
|---|---|---|
| Elementary school teachers | 1.79x | 22 |
| Registered nurses | 1.64x | 65 |
| Electricians | 1.48x | 149 |
| Software developers | 1.38x | 231 |
| Accountants and auditors | 1.29x | 339 |
This is the counterintuitive part. A software developer's pay depends on their metro less than a teacher's does - rank 231 versus rank 22 of 392 - even though intuition runs the other way: developers seem like the quintessential "work from anywhere" job, while teaching feels tied to place. A likely explanation: developers increasingly compete in something closer to a single national, often remote-friendly labor market, which compresses metro-to-metro differences, while teacher pay is set by thousands of local school districts against local tax budgets that don't converge the same way. A plausible story, not a proven one.
But the hypothesis behind this study - that place-bound work varies more than nationally-traded work - only holds up partway. Teachers, nurses, and electricians do show wider spreads than developers, supporting it. Accountants and auditors don't fit: also a locally delivered profession in most cases, they show the narrowest spread of all five, tighter even than software development. One occupation breaking the pattern is enough to call "place-bound work varies more" a tendency in this data, not a rule.
Does the spread survive adjusting for cost of living?
For the top three occupations, we recalculated the spread after dividing each metro's wage by that metro's cost index. The result doesn't point one direction. Substitute teachers' spread shrank from 2.16x to 1.80x and firefighters' from 2.13x to 1.90x - part of what looked like a pure pay gap was really a cost-of-living gap. Waiters and waitresses moved the other way, 2.12x to 2.17x - cost of living doesn't explain their spread at all. Geographic pay spread isn't one phenomenon with one cause; it's several mechanisms that happen to produce a similar-looking number.
What it means for you
If your occupation has a wide spread, where you work is one of the biggest levers on your earnings, and relocating - or negotiating with real local benchmarks - can move your pay substantially. If it has a narrow spread, like accounting or postal work, moving cities won't change your salary much, so cost of living becomes the deciding factor instead: the same paycheck in a cheaper metro simply goes further. Either way, a high nominal wage in an expensive metro can still lose in real terms to a lower one somewhere affordable.
Our national study on where a paycheck goes furthest ranks metros on exactly that basis, and our salary calculator converts a salary between any two US metros. You can also look up your own occupation - for example registered nurses or software developers - to see the full metro-by-metro range.
The takeaway
Substitute teaching shows the widest geographic pay spread in this study, at 2.16x between a typical low- and high-paying metro; Postal Service mail sorting shows the narrowest, at 1.08x, alongside two other postal jobs. For some careers, your city is one of the biggest financial decisions you'll ever make; for others - largely federal work and tightly licensed professions - it barely registers. And the most surprising result, that developer pay depends on location less than teacher pay does, is a reminder that "remote-friendly" and "geographically variable" aren't the same axis at all.