A handful of countries publish occupation-level wage data of comparable quality to the US survey — the UK, Canada, Australia and several EU members through the European structure of earnings survey. A larger group publishes only national or sector averages. For most of the world nothing official exists at occupational level, and the sites that appear to fill the gap are almost always self-reported figures presented with more confidence than they can support.
Why this question comes up at all
You are considering a job in another country, or a colleague has moved and you want to know whether they did well, or you simply want to know whether your field pays better somewhere else. In the US you would look up the occupation and the metro area and have an answer in two minutes, complete with a full percentile range. The instinct is that the same thing must exist everywhere, and it very nearly does not. What you can find out varies so sharply from country to country that the first useful step is establishing which situation you are in before you start searching.
That sounds like a technicality and it changes what you should do next. If good data exists, use it and stop. If only national averages exist, you can still bracket the answer usefully. And if nothing exists, the honest move is to say so and go looking for a different kind of evidence entirely rather than accepting whatever a search engine puts in front of you.
The top tier: occupation, region and a full distribution
A small number of countries run establishment surveys of a similar character to the American one, producing wage figures by detailed occupation and by region, usually with percentiles rather than just an average. The United Kingdom, Canada and Australia all publish something in this class through their national statistical offices, and several EU member states do too. Where this exists you can genuinely ask “what does my occupation pay in that region” and get a real answer with a spread attached.
Across the European Union there is an additional route worth knowing about. A structure of earnings survey is conducted on a common basis every four years, which produces figures that are more comparable across member states than anything you could assemble yourself. The trade-off is that it is periodic rather than annual, so the most recent release may describe a year that already feels historical, and you should note the reference period next to any figure you take from it.
The middle tier: averages without the detail you want
A much larger group of countries publishes national average earnings, sometimes broken down by broad sector or by broad occupational group, but not at the level of a specific job title. This is genuinely useful and it is not what you were hoping for. It will tell you roughly where a country sits and how your sector compares to the national average, which is enough to rule things in or out.
The way to use it is to bracket rather than to pin down. Establish the national average, establish how your broad occupational group compares to it where that breakdown exists, and accept that you are working with a range rather than a figure. A bracket you can defend is worth considerably more than a precise number you cannot, and it is usually enough to answer the question you actually had.
The bottom tier: nothing official, and plenty that looks official
For a great deal of the world, no national statistical office publishes anything at occupational level. What fills that vacuum online is a family of salary sites built from self-reported submissions, presented with country flags and clean tables and an air of authority that the underlying data does not support. They are not necessarily dishonest, and they are answering a different question than you think.
The trouble with self-reported data is not that people lie, though some do. It is that the people who submit are not a random sample of the workforce — they skew toward recent job-changers, toward those who found the salary notable, and toward a handful of industries and cities where the site is popular. In a country with a small number of submissions, a single well-paid respondent can move a published figure substantially, and nothing on the page will tell you how many responses it rests on.
The classifications do not line up, and that matters more than it sounds
Even between two countries that both publish excellent data, the occupational classifications are different systems built for different purposes. A job title that maps to one code in the American system may map to two in the British one, or sit inside a broader category in the Canadian one. Comparing a figure from one against a figure from the other is therefore comparing two groups of people that were never defined the same way.
Most international classifications map loosely onto a common international standard, which helps, and the mapping is approximate rather than exact. So when you compare across borders, check what the foreign classification actually includes before you trust the comparison. Reading the definition takes two minutes and occasionally reveals that the two categories differ enough to explain most of the gap you were about to attribute to the country.
What to do when nothing is published
Go to the market rather than to the statistics. Job postings in countries with pay transparency requirements carry real advertised ranges, and even where disclosure is not required a scan of current listings for your occupation will give you a defensible band. Recruiters who place in that market will talk about ranges freely because it is how they qualify candidates, and a fifteen-minute call is often better evidence than any table you will find.
Professional associations and unions are the other route, and they are consistently underused. In countries where sector-wide agreements set pay, the agreement itself is a published document with grades and figures in it, which is stronger evidence than a survey because it is not a sample at all — it is the rule.
A practical test before you rely on any figure
Ask four things of any source you are about to use. Who collected it, and were they an employer, a government agency or the internet? What period does it describe, and how long ago was that? How many observations sit behind the number you are looking at? And does the classification mean what you assume it means?
A source that answers all four is worth trusting within its limits. A source that answers none of them is not evidence, however confidently it is presented, and the confident presentation is frequently the tell. The best sources are usually the ones that go out of their way to tell you what they do not cover.
Common questions
Which countries publish good occupation-level wage data?
The UK, Canada and Australia through their national statistical offices, plus several EU members. The EU also runs a structure of earnings survey on a common basis every four years, which is more comparable across countries than anything you could assemble.
What if only national averages exist?
Bracket rather than pin down. Take the national average, compare your broad occupational group to it where that breakdown exists, and work with a defensible range instead of a false precision.
Are salary comparison sites reliable abroad?
They are self-reported and the respondents are not a random sample — they skew to recent job-changers and to a few industries and cities. In a country with few submissions, one well-paid respondent moves the figure.
Why do classifications matter?
Because a title mapping to one code in the US may map to two elsewhere, or sit inside a broader category. Comparing figures then compares two groups that were never defined the same way.
What should I do where nothing is published?
Go to the market. Job postings in transparency jurisdictions carry real ranges, a scan of current listings gives a defensible band, and recruiters who place in that market discuss ranges freely.
What about unions and professional associations?
Consistently underused. Where sector-wide agreements set pay, the agreement is a published document with grades and figures — stronger than a survey because it is the rule rather than a sample.
How do I test a source before relying on it?
Ask who collected it, what period it describes, how many observations sit behind the number, and whether the classification means what you assume. A source answering none of those is not evidence.
Is EU data more comparable than national data?
For cross-country comparison, yes, because it is collected on a common basis. The trade-off is that it is periodic rather than annual, so note the reference period alongside any figure.