Pay Transparency Isn't a Value. It's a Jurisdiction.

We track 1,073 tech companies and read the details off every role they have open. This week we went looking for salary, expecting to write something about which stacks pay best, and got stuck on a much dumber number first.

Of the 14,865 open roles we could read cleanly, 49.2% state the pay.

That number is useless. Not wrong, useless, because it is the average of two worlds that have almost nothing to do with each other. Split it by city and it falls apart in your hands.

The map

San Francisco and Bangalore are not different industries. They are frequently the same companies: OpenAI, Stripe, Datadog and MongoDB all have open roles in both. The same employer, the same year, the same recruiting team in many cases. One posting tells you the money. The other doesn't.

That rules out most of the comfortable explanations. It isn't company culture, because it's the same company. It isn't sector. It isn't "tech is secretive", because tech is 88% transparent in one city and 0.4% transparent in another.

It's the law. That's the whole finding.

Every city at the top of that table sits inside a pay-transparency statute, and the dates line up like a fingerprint.

Texas has no such law. Austin: 54%. The UK has no such law. London: 18.7%. India has no such law. Bangalore: 0.4%.

We are not the first people to notice that laws move this number. Indeed's Hiring Lab has tracked disclosure monthly for years and watched British Columbia jump from 49% to 76% when its act landed. The New York Fed put the average jump at about 20 percentage points on the day a law takes effect. There is even a proper NBER paper on Colorado finding a 30-point rise, and wages up 1.3 to 3.6% alongside it.

What we can add is the other half of the picture. Those studies are excellent and they are almost entirely about places that passed laws. Our set spans both regimes at once, in one snapshot, with one method: the compliant cities and Bangalore in the same table. That contrast is the thing worth staring at, and it says pay transparency is not a norm the industry is slowly adopting. Where nobody is forced, it sits near the floor. Berlin is 5%. Amsterdam is 3%.

The laws leak, and that's the part we didn't expect

Austin bothered us. If the story is purely legal, Austin should look like Berlin. It's 54%.

So we split the no-law US cities by whether the role is remote.

An onsite job in a no-law city behaves like you'd expect: 38.5%, closer to London than to California. A remote job in the same city, often at the same company, discloses at 84.6%.

The mechanism is in the statutes, and we checked all six: every one of them reaches roles that could be performed in-state, not just roles sitting in an office there. Colorado's is the broadest. So the moment a company writes "remote, US", somebody in Denver or Sacramento can apply, and the law is now in the room. And once you're writing one compliant posting, maintaining a second secretive variant for Texas is more work than just posting the number everywhere.

California didn't only make California transparent. It made "remote, US" transparent.

(Honesty check on that table: n=26 for the remote cell is small, and we're not going to pretend 84.6% is precise. The direction is large and it matches the legal mechanism exactly, which is why we believe it. If it were the only evidence we had, we'd be quieter about it.)

Boston is the interesting one

Massachusetts has the newest law in our set, in force since 29 October 2025. Boston is also the most transparent city in our entire dataset at 93.0%, ahead of Colorado, which has had five years to get there.

We can't prove why from our data (and neither the Fed study nor Lightcast's work covers Massachusetts, so nobody else has published this cut). The obvious guess: the multi-state employers hiring in Boston were already writing compliant postings for California, New York and Washington. When Massachusetts arrived, there was no adjustment period, because the template already existed. A law lands faster when everyone's already been following four of its cousins.

That's an inference, not a finding. But it suggests these laws have a compounding quality, where each new one costs less to comply with than the last.

Europe is not about to catch up

You would think the EU Pay Transparency Directive settles this. Its transposition deadline was 7 June 2026, and it has now passed with 4 of 27 member states actually done. Germany, France, Spain and the Netherlands all missed it, which is roughly where our worst European numbers live.

There's a subtler reason not to expect Berlin to become San Francisco. The Directive asks employers to tell applicants the pay in the vacancy notice or before the interview. Pre-interview disclosure satisfies it. A European employer can comply fully and still post a job with no number in it, which is precisely the thing a job seeker is trying to avoid. Indeed's own read is that the directive alone looks insufficient without national law behind it, and their evidence is that employers move when rules bite, not when they're announced.

So treat the European numbers here as a baseline taken the year the rule was supposed to arrive. We'll run it again.

Where this data is soft

The part where we tell you what's wrong with it.

Our numbers will not match Indeed's, and Indeed isn't wrong. They report India at around 50% disclosure. We report Bangalore at 0.4%. Both are true, because we are not counting the same jobs. Indeed covers an entire labour market, and it's weighted toward high-volume hourly and junior postings. We track curated tech companies, and 64% of our roles are senior. Indeed's own India breakdown makes the point better than we can: senior leadership roles disclose at 13% while junior and mid-level sit at 32%. The more senior the job, the more the pay disappears. Our set is the senior, professional end of one industry, so we see the floor of that pattern. Read our figure as "global tech companies hiring in Bangalore", not "India".

No before and after. We only started recording job history in October 2025, so every law above predates our data (Massachusetts by two days). We did not watch disclosure jump. We're looking at where things stand, not at the moment they moved. Everyone in the links above has the longer time series; we have the wider map.

A mention isn't a range. We count a role as disclosing when we can extract structured pay from the posting. A company that reveals the number later in the process, or in the recruiter's first email, counts as not disclosing here, because the job seeker staring at the posting can't see it either. That's the right definition for a reader, and the wrong one for judging a company's compliance.

Which is why we aren't naming anyone. It's tempting to publish a list of companies posting in California without a salary. We won't. Our scrape can't see an ATS flow, or a headcount exemption, and a public accusation of law-breaking deserves a better evidence standard than "a scraper didn't find a number". Aggregates only.

The floor is not zero by accident. Bangalore at 0.4% means two roles out of 560. We looked twice.

What to do with it

If you're job hunting and a posting has no salary in it, the useful question isn't "what are they hiding". It's "where is this job, legally". A silent posting in Berlin is a Tuesday. A silent posting for a remote US role is stranger, because most of that company's peers decided compliance was easier than secrecy.

And if you want the number before you spend an hour on the application: remote-US roles are, right now, the most transparent thing on the board. Not because those employers are more enlightened. Because California is watching.

Every city above has a page with the companies hiring there (remoet.dev/locations), and you can ask your agent to check your own stack against them.

We'd like to know whether this matches what you're seeing from the inside, especially if you write job posts for a living and have opinions about which of these rules actually changed your process. Come find us on Discord.