Methodology

First principles

• Everything runs in your browser. The number you type is never transmitted or stored on a server. Your ledger lives in your browser's local storage; export it any time.
• Open sources only. Baked-in datasets come from government publications (public domain) and openly licensed tabulations — never scraped commercial databases. Where only estimates are possible, rows are labeled as estimates.
• US dollars underneath. All math runs in USD; the currency selector converts display and entry using fixed reference rates (ECB-style), refreshed with each yearly data pass.

Global percentile

Source: the wealth-per-adult distribution in the UBS Global Wealth Report (≈5.5 billion adults). We anchor the published distribution at fifteen points — from $50 (7th percentile) through $8,700 (median), $141,000 (90th), $1.1M (99th), to the billionaire tail — and interpolate log-linearly between anchors:

percentile(w) = p₀ + (p₁ − p₀) · (log₁₀ w − log₁₀ w₀) / (log₁₀ w₁ − log₁₀ w₀)

Negative or zero net worth maps to roughly the bottom decile, consistent with the report's estimate that about one in ten adults nets at or below zero.

Age and country percentiles

Sources: US age brackets — Federal Reserve Survey of Consumer Finances 2022 (household net worth). Countries — UBS Global Wealth Databook end-2022 wealth per adult, as openly tabulated.

Neither source publishes full percentile tables, so each group's distribution is fitted log-normally from its published median and mean:

σ = √(2 · ln(mean/median))   percentile(w) = Φ((ln w − ln median)/σ) × 100

Validated against known reference points (the fit puts the US all-adult 90th percentile near $1.09M against a published ≈$1M). Good to a few percentage points through the middle; looser in the top 1%. Note the units mismatch we inherit from the sources: age brackets measure households, countries measure adults.

City living costs

US metros — derived from federal data

Sources: BEA Regional Price Parities (public domain), all 387 metros, 2008–2024.

monthly(metro) = Σ category baseline × RPP_category(metro) / 100
baselines (2024): housing $1,450 · goods $900 · utilities $250 · other services $800

The $3,400 national baseline represents one person living comfortably with their own place, rent included. Historical years scale the baseline by the national implicit price deflator. Full outputs are published at /data/us-metros-2024.json and /data/us-metros-timeseries.json.

US wages and home values

Metro median salaries are the BLS Occupational Employment and Wage Statistics all-occupations median (gross annual, May 2025). Median home values are Census ACS 2023 table B25077. Both public domain; archived under /data.

International cities

Rounded editorial estimates in the spirit of crowd-sourced indices, labeled as estimates wherever they appear. City-level cost distributions are not published openly anywhere; we say so rather than fake precision. International salaries are country medians (net, estimates); international home prices derive from m² estimates × 75m².

Salary

Source: BLS Occupational Employment and Wage Statistics (OEWS), May 2025 vintage, public domain. The "years of the median salary" comparison uses each US metro's all-occupations median, gross monthly pay; international rows are country-level net median estimates, labeled as estimates.

Underneath, WorthParty also keeps OEWS's published annual wage percentiles — the 10th, 25th, 50th, 75th, and 90th — broken out by metro and broad occupation group. Placing one salary against that spread uses the same log-linear interpolation as the global percentile above, sliding between the two nearest percentile points and extrapolating their slope past either end:

percentile(s) = p₀ + (p₁ − p₀) · (ln s − ln s₀) / (ln s₁ − ln s₀)

A median wage is a midpoint, not a receipt for what any one person earns, and a BLS occupation group is broad — a specific job can sit anywhere inside its band. Treat a salary placement as a well-drawn map, not a survey marker.

Projections and FIRE thresholds

The trajectory compounds monthly at your growth rate, reduced by your tax-on-growth setting, with contributions added monthly:

w_m+1 = w_m · (1 + r_net/12) + contribution   r_net = r · (1 − tax)

FIRE rungs follow community conventions: FI = 25× annual spending (the 4% rule, per Bengen 1994 and the Trinity study); Lean = 25× of 70% spending; Barista = 12.5×; Fat = 25× of 150%; Coast FI discounts the full number from age 65 at your net growth rate. The FIRE guide covers what these formulas leave out — sequence risk, US-data bias, longer retirements.

Known limitations, collected

• Wealth survey data carries wide error bars; percentile ranks are estimates, not rulings.
• Metro figures are area averages; city cores run higher.
• Salary figures are published medians, not a personal record of what anyone earns; occupation groups are broad.
• Projections know nothing about market volatility or inflation; the tax setting is an effective-rate simplification.
• Currency conversions use fixed reference rates — perspective, not trading data.
• Every figure is self-reported by you; the site verifies nothing and stores nothing server-side.

The open data archive

• us-metros-2024.json — 387 metros, derived monthly cost + RPP
• us-metros-timeseries.json — the same, 2008–2024
• us-metro-wages-2025.json — BLS median wages
• us-metro-home-values-2023.json — ACS median home values

Free to reuse; please attribute the underlying federal sources.