How the math works

If you pick local market pay, your salary is scaled by the ratio of the two metros' wage levels. A nurse earning $95,000 in Chicago is modelled at roughly $122,000 in San Jose and $88,000 in Tulsa, because that is what the local market pays for the same work.

If you pick remote, your salary is held constant everywhere. That single switch is usually worth more than any other input on the form, because you keep a high-cost-metro salary while paying low-cost-metro rent.

Leave the salary field blank and we substitute the national median for your occupation, adjusted to your current metro.

Applied in this order, on gross wages:

• Federal income tax — 2025 ordinary-income brackets and the standard deduction, for single, married-filing-jointly, or head-of-household. Filing status is inferred: one adult with kids files as head of household.
• Social Security and Medicare — 6.2% up to the $176,100 wage base, 1.45% with no cap, plus the 0.9% additional Medicare tax over the threshold. The Social Security cap is applied per earner, not per household, which matters for two-income couples.
• State income tax — a per-state model with three shapes: no wage tax, flat rate, or a bracket schedule, each after that state's standard deduction. Joint filers get roughly doubled bracket widths.
• City and county wage tax — flat rates where they exist, which is why Philadelphia, New York City, Baltimore, Cleveland, Columbus, Detroit and Louisville look worse here than on sites that ignore them.
• Child tax credit — $2,000 per child, phasing out by $50 per $1,000 of income over the threshold.

Not modelled: itemised deductions, 401(k) and HSA contributions, self-employment tax, the earned income credit, capital gains, or any state-specific credit. If you max a 401(k), your real effective rate is lower than what you see here — in every city, so the ranking barely moves.

Rent is priced at the bedroom count you choose, not a metro-wide average. That is the single biggest flaw in index-style comparisons: a studio and a three-bedroom in the same city are not the same financial place, and the gap between them is not consistent across metros.

One measured two-bedroom asking rent is stored per metro, and the other sizes are derived with national ratios (studio 0.72×, one-bed 0.84×, three-bed 1.31×, four-bed 1.55×). Those ratios are a national assumption, and a known-imperfect one: the measured gap between single-family and apartment asking rents runs from 1.00× in New York to 1.59× in Austin, so this overstates large units in apartment markets and understates them in sprawling ones.

That two-bedroom figure comes from Zillow rather than from a federal median, and the choice matters. Census ACS medians cover the entire rental stock, including long-held and rent-stabilised leases, so they sit far below a new lease where turnover is slow — 41% below in New York — and close to it where building is fast. That bias runs the same direction every time, which would quietly make the most expensive metros look like the affordable ones. HUD's Fair Market Rents lean the other way: 40th-percentile rents for standard-quality units, set for housing-assistance purposes rather than to describe the market. Both are recorded next to every metro so the spread is visible.

The buy-instead figure on each city page is principal and interest on 80% of the median home at 6.5% over 30 years, plus that metro's effective property tax rate and 1.1% a year for insurance and upkeep.

National monthly baselines, scaled per person and then by the metro's non-housing cost index:

• Groceries — $400 per adult, $250 per child
• Healthcare — $450 per adult, $170 per child, premiums and out-of-pocket together
• Utilities and internet — $210 plus $40 per person
• Getting around — $150 per adult, plus up to $520 more depending on how car-dependent the metro is. This is why New York can beat cheaper cities for a single person: no car is a real raise.
• Childcare — $1,150 per child under five, full time. Frequently the largest line item after rent.
• Everything else — $350 per adult for clothes, phone, entertainment and the rest

Deliberately excluded: debt payments, tuition, travel, and savings goals. Those are yours, not the city's.

Five sliders, each weighted zero to three. Every metro is scored 0–100 on each dimension, then the weighted average becomes the fit score used for ranking.

• Money left over — your monthly leftover, normalised against the best and worst metro for your specific profile. Change your household and this rescales.
• Sun and mild weather — 55% sunny days, 45% a comfort score that penalises July highs above 92°F and January lows below 26°F.
• Outdoors and city life — editorial 0–100 composites. These are judgement calls, not survey data, and they are the softest numbers on the site.
• Short commute — mean one-way commute, scaled between 18 and 38 minutes.

Set every slider to zero and the ranking falls back to money alone.

All 7 of these are free and public. The fetch script pulls the Zillow, HUD and Census figures — last run 2026-09-04; the rest are annual manual updates.

It is not tax advice, a rent quote, or a mortgage pre-approval. It is a model, and models are wrong in the details by construction. Use it to work out which four cities are worth researching properly, then go and check real listings and run your actual numbers past someone who files taxes for a living.