Six Pay-Transparency Laws Took Effect in 2026. Almost None Say How Wide a "Range" Can Be.

By NiceHire Team

Four new salary-disclosure regimes crossed from statute book to enforcement this year: the EU's Pay Transparency Directive (deadline June 7), Virginia's law (July 1), Maine's (July 29), and Connecticut's rewritten one (October 1). A fifth, Delaware's, is queued for 2027. Add Ontario, which isn't EU or US but moved on the same calendar (January 1) and is the one jurisdiction in this group that actually tried to close the gap the others left open. Count them up with SHRM's tracker and the US alone now has eighteen states plus DC requiring a posted range in some form — other trackers count fewer, because they draw the line differently on what counts as "requiring" a range, which is itself a small preview of the bigger problem: almost nobody defines what a compliant range actually has to look like.

That's not a rhetorical flourish. It's the specific, checkable thing four of these five 2026-vintage laws have in common: each requires an employer to disclose a pay range. None of them — except Ontario — specifies how wide that range is allowed to be. And new research published this year, on an unrelated track from any of the legislating, found that the width of the range is not a neutral detail. It's the part of the disclosure that actually does the work, or doesn't.

Two different legislative philosophies, the same blind spot

The EU's Pay Transparency Directive (2023/970) set a transposition deadline of June 7, 2026, and missed it almost everywhere. By the end of that month only four of 27 member states — Slovakia, Italy, Lithuania, and Malta — had fully transposed it, per a tracker compiled by WageIndicator and corroborated by Lewis Silkin's post-deadline compliance note, which lists the same four as fully in force and flags Belgium and Poland as partial. Sweden is the outlier in the other direction: its government announced on March 26, 2026 that it opposes the Directive outright and wants it renegotiated at the EU level, rather than simply running late (per L&E Global's member-state tracker). What the Directive actually requires, where it has landed, is that candidates get the starting pay or pay range before an interview (Article 5), that employers can no longer ask about current or past pay at any stage, and that workers can request their own pay level and the average for comparable work broken down by sex — with the burden of proof shifting to the employer in a pay-discrimination claim. Nowhere in that list is a ceiling on how wide the disclosed range can be.

The US pattern is procedurally different but lands in the same place. Virginia's law (HB 636 / SB 215, signed April 22, effective July 1) is the most aggressive enforcement mechanism of the three 2026 entrants: it covers employers of every size, bans asking about pay history, and gives applicants a private right of action with only a 15-business-day cure period before a lawsuit can proceed — one of just two states, alongside Washington, to go that far (Trusaic; Cooley). Maine's (LD 54, codified at 26 M.R.S.A. § 622-A, effective July 29) requires a pay range in every posting from employers with 10 or more employees, with a narrow carve-out for commission-only roles and a current-employee right to request the range for their own position (Morgan Lewis; JDSupra). Connecticut's HB 5003 (Public Act 26-12, signed May 2026, effective October 1) rewrote the state's 2021 on-request disclosure law into a proactive one: the wage range and a description of benefits must now appear in the posting itself, not wait for an applicant to ask, and the range must be set "in good faith" against a pay scale, prior range, current comparable employees' pay, or the budgeted amount for the role (Seyfarth; Foley). Delaware's law, effective September 26, 2027 for employers with more than 25 employees, is written the same way. All four define a range. None defines a maximum spread.

Ontario is the exception, and it's worth looking at precisely because it shows what closing the gap actually requires. The Working for Workers Four Act and its implementing regulation (O. Reg. 476/24), effective January 1, 2026 for employers with 25 or more employees, don't just mandate a range — they cap it. A posted range for a role paying under $200,000 a year cannot span more than $50,000 from bottom to top; a $100,000–$140,000 posting is compliant, $100,000–$160,000 is not (corroborated by Law Times News and HCAMag). It's a deliberate response to the exact failure mode the EU and US laws haven't addressed yet: a range so wide it discloses nothing.

The research that caught up with the loophole

None of this would matter much if a wide range were just a wide range. It isn't. A 2026 Journal of Applied Psychology study by Lee, Park, and Chang — hosted as a working paper by Cornell's ILR School and summarized by both SHRM and the Harvard Business Review store page for the research — ran four studies and found that women show a measurably stronger preference than men for jobs posted with narrower ranges, a pattern the authors attribute largely to higher risk aversion around an uncertain number. The effect doesn't stop at who applies. Among candidates who did pursue a role, the ones drawn to narrower ranges also negotiated less assertively: they reported greater satisfaction with a midpoint offer, were less likely to counter, and when they did counter, asked for less. Both SHRM's and HBR's summaries converge on the same mechanism: a wide range doesn't just fail to inform, it actively discourages the people most sensitive to pay uncertainty from applying or pushing back once they do.

The same research identified what actually works instead of a narrower number alone: pairing the range with plain context — the typical starting salary within it, and the criteria used to land on a final offer — substantially reduced the gap in both application behavior and negotiation confidence, according to the HBR write-up. In other words, the fix the data points to isn't only "post a tighter range." It's "tell candidates what the number means and how you got there," which a bare range, however narrow, still doesn't do on its own.

What employers are actually doing, a year into this

Ontario is the test case, because it's the one place with both a cap and nearly two years of before/after data. Disclosure itself jumped hard: the share of Ontario postings carrying explicit pay information rose from 49% before the law to 71% by June 2026, per Indeed's tracker as reported by HR Reporter (a single-source figure we could not corroborate against a second outlet, flagged as such). But within that mandatory disclosure, the median posted range widened, not narrowed, after the law and its $50,000 cap took effect — from roughly 16% spread in the first half of 2025 (the equivalent of $100,000–$116,000) to roughly 18% in the first half of 2026 ($100,000–$118,000), per the same reporting. Even under an explicit numeric ceiling, employers moved toward more room inside it, not less. Meanwhile Indeed's own US national tracker shows the share of postings disclosing any pay information growing at a decelerating rate — 50.4% in July 2026, up only 1.9 points from 48.5% a year earlier — even as more states joined the mandate list in between. More jurisdictions are requiring a number. The number itself isn't obviously getting more useful.

Where we come out on this

NiceHire sits in the part of the hiring process this split exposes most directly — job postings and the structured interview pipeline that follows them — without a stake in which legislative model wins. What the research above says is worth taking seriously regardless of which side of the Atlantic, or which US state line, an employer is hiring across: a number alone, however it got onto the page, isn't what makes a disclosure requirement work. The context behind it is. A posting that satisfies Virginia's, Maine's, or the EU Directive's letter while leaving that context out has cleared the statute, not done what pay-transparency law was written to accomplish. Employers preparing for Delaware's 2027 deadline, or a second wave of EU transpositions, have a cheaper and already-evidenced option sitting in the same research: don't just narrow the range. Explain it.