How each country measures pay
Every country on this site measures pay differently. None of them is wrong, but they are not interchangeable, and a comparison that ignores the difference will mislead you. Here is exactly what each figure is.
SOURCE·National statistics agencies (7 countries)·REVIEWED JUL 2026
The five things done to a wage figure before it appears on a page
- Take the occupation row exactly as the national agency published it, in local currency.
- Carry the whole distribution where one exists. Lead with the median, never the mean, wherever both are published.
- Annualise anything published monthly, and label it as annualised on the page rather than silently.
- Read the wage twice — once at the European Central Bank reference rate, once at the World Bank purchasing-power conversion factor — and print both.
- Print the agency, the series and the reference period under the figure. No figure without a source line; no source line without a date.
converted = local wage / FX(local -> USD) adjusted = local wage / PPP(country) gap = adjusted / converted
By country
| Country | What the figure is | Source & period |
|---|---|---|
| USAUnited States | Employer payroll survey of ~1.1m establishments | US Bureau of Labor Statistics, Occupational Employment and Wage Statistics (OEWS), May 2024 Public domain (US Government work) · MAY 2024 |
| GBRUnited Kingdom | 1% sample of PAYE records, full-time employees | ONS, Annual Survey of Hours and Earnings (ASHE) Table 14, 2025 provisional Open Government Licence v3.0 · 2025 PROVISIONAL |
| CANCanada | 2021 Census wage distributions by NOC and economic region | Employment and Social Development Canada, Job Bank wage data 2025 Open Government Licence – Canada · 2025 |
| AUSAustralia | Individual tax returns — includes part-year and part-time filers | Australian Taxation Office, Taxation Statistics 2023-24, Table 15 CC BY 4.0 · 2023-24 |
| SGPSingapore | Full-time resident employees, private establishments with 25+ staff | Singapore Ministry of Manpower, Occupational Wage Survey, June 2024 Singapore Open Data Licence · JUN 2024 |
| NZLNew Zealand | Household survey — published only at ANZSCO major-group level | Stats NZ, Household Labour Force Survey (income), June 2025 year CC BY 4.0 · JUN 2025 YEAR |
| INDIndia | ILO-harmonised national survey data, reported for ISCO-08 occupation groups. Covers employees only — a minority of India's workforce, most of which is self-employed or informal. | ILOSTAT, average monthly earnings of employees by occupation (ISCO-08), 2025 CC BY 4.0 · 2025 |
Seven agencies · JUL 2026
The three traps
Mean versus median. A handful of very high earners drag an average upward. The median — the middle person — is almost always the more useful number, and it is what we lead with everywhere it exists.
Tax data versus payroll surveys. Australia's figures come from tax returns, so they include anyone who filed with that occupation, including people who worked part of the year. American and British figures come from employer records for full-time jobs. The Australian number will read lower for that reason alone in some occupations, and higher in others where side income is common.
Exchange rates versus purchasing power. Converting a salary at the market rate tells you what it is worth if you move the money out. It does not tell you how well someone lives on it. Every comparison page shows both.
What we do not do
We do not collect self-reported salaries, we do not model or estimate a figure a statistics office declined to publish, and we do not fill a gap with a near-enough occupation from another classification. Where a country suppressed an estimate — as the ONS did for police officers in the 2025 provisional release — the page says so and shows nothing.
