A good faith salary range is the pay band an employer honestly believes, at the time of posting, it would actually pay a qualified candidate for the role. It must reflect real expected compensation, not a placeholder like $50k to $200k, and it is increasingly anchored to market data, typically the 25th to 75th percentile for the role. As of 2026, pay transparency laws in 16 states plus Washington, D.C. require you to post one, and enforcers now flag bands that are too wide as not made in good faith.

That creates two ways to fail at once. Post no range, which is illegal across those 16-plus states. Or post an indefensible one, a band so wide a labor commissioner calls it a placeholder. This guide covers which states require a range, what "good faith" actually means, what changed in 2026, the penalties, and how to build a narrow, market-grounded range you can stand behind, with the methodology saved as your audit trail.

## Pay transparency now covers 16 states plus D.C., and remote hiring exposes you to all of them

As of 2026, **16 states plus Washington, D.C. have enacted statewide pay transparency laws**, a count that converges across multiple trackers including Payscale's March 2026 update. The states most commonly cited as requiring a salary range *in the job posting* include California, Colorado, Hawaii, Illinois, Maryland, Massachusetts, Minnesota, New Jersey, New York, Vermont, and Washington, plus D.C.

The count is definition-sensitive. Comprehensive trackers like [GovDocs](https://www.govdocs.com/pay-transparency-laws/) list roughly 18 jurisdictions because they include states that only require pay disclosure *on request*, such as Connecticut and Nevada, rather than in the posting itself. So "16 states plus D.C." is the figure for in-the-posting requirements; broader counts exist depending on how you define the obligation.

The trap for startups is remote hiring. Most state laws are triggered by where the *applicant* sits, not where your company is headquartered. A fully remote role open to candidates nationwide is, in practice, exposed to every state law at once. You do not get to pick the friendliest regime. You comply with the strictest one that any applicant could trigger, which means a single remote job post is a multi-state compliance decision.

## What "good faith" actually means, and why $50k to $200k gets you flagged

A good faith range is the band you honestly believe, when you post, that you would actually pay a qualified candidate. New York's Department of Labor frames it as a range the employer "believes in good faith to be accurate when the ad is posted." There is deliberately no hard percentage in statute, which is precisely why absurdly broad bands draw scrutiny.

The canonical illustration comes from New York City. In a [Fortune report from February 2024](https://fortune.com/2024/02/13/job-pay-transparency-diversity-nyc-tesla/), the NYC Commission on Human Rights cited News Corp for posting an education-reporter role at **$50,000 to $180,000**, a span so wide it did not qualify as a good-faith estimate, alongside a similar flag for Tesla. That case predates 2026 and ran under NYC's 2022 ordinance, so treat it as the standard-setting example rather than a recent event. But the principle it established is exactly what the 2026 wave of enforcement applies: a range has to be a genuine estimate, not a catch-all.

The practical test is simple. If you cannot explain *why* the floor is the floor and the ceiling is the ceiling, with reasons tied to the role and the market, the band is not defensible. "We picked round numbers wide enough to cover anyone" is the failure mode enforcers look for.

## What changed in 2026: California's SB 642, Maine, Virginia, and a shift to enforcement

Three things define the 2026 landscape.

**California SB 642**, effective January 1, 2026, redefines "pay scale" as a **"good faith estimate of the salary or hourly range that the employer reasonably expects to pay for the position upon hire"** and extends the recovery and lookback period to six years. That codifies the good-faith standard into statute and gives it a long tail. The definition is verified across [Jackson Lewis](https://www.jacksonlewis.com/insights/navigating-2026-pay-transparency-laws-and-employer-obligations), Hunton, and other firm analyses.

**New posting requirements phase in.** Per GovDocs effective-date tracking, **Virginia's law takes effect July 1, 2026** and **Maine's takes effect July 28, 2026**. If you hire in either state, your postings need a range within weeks of this article's publication.

**The dominant theme is enforcement, not new statutes.** The states that passed laws in 2023 and 2024 are now actively penalizing non-compliance, including overly broad ranges. The legal map is mostly drawn. What is intensifying is the cost of getting the *content* of the range wrong.

## The penalties, state by state ($100 to $250,000)

Penalties vary widely by jurisdiction. A handful of verified figures:

| Jurisdiction | Penalty |
|---|---|
| California | $100 to $10,000 per violation (Labor Commissioner) |
| Colorado | $500 to $10,000 per violation |
| New York (state) | $1,000 first violation, $2,000 second, $3,000 subsequent |
| New York City | Civil penalties up to $250,000 for unremedied violations |

Sources: Kelly Services, [Rippling](https://www.rippling.com/blog/pay-transparency-laws-state-by-state-guide), and Fortune. The headline number is NYC's $250,000 ceiling, but the more common exposure for a multi-state remote employer is the stacking risk: a single non-compliant posting seen by applicants in several states can trigger several regimes, each with its own per-violation penalty.

## Do salary ranges actually help, or just create risk?

They help. A SHRM survey found **70% of businesses that included a salary in job posts saw more applicants**, and roughly nine in ten job seekers say salary transparency is a key factor in their search. Adoption backs this up: as of 2025, about **60% of Indeed postings include pay, up from 18% in 2020**, a figure verified via Indeed's Hiring Lab and [HBR](https://hbr.org/2026/02/posting-a-wide-salary-range-can-deter-women-from-applying).

The feared downside, that disclosure scares off employers or shrinks the funnel, does not show up in the data. Academic studies, including work via [ScienceDirect](https://www.sciencedirect.com/science/article/abs/pii/S0165176525002605) and the [Minneapolis Fed](https://www.minneapolisfed.org/article/2024/pay-transparency-in-job-postings-trends-trade-offs-and-policy-design), find that state mandates change posting volume by only about 1 to 6 percent, often a statistically insignificant amount, while improving applicant self-selection. Candidates who see a fair number that matches their expectations apply; those who would have walked at the offer stage screen themselves out earlier. That is a cleaner funnel, not a smaller one.

## How wide is too wide? What the data and HBR say about range width

Too wide is both a legal problem and a funnel problem. Legally, an over-broad band reads as a placeholder and invites a good-faith challenge. On the funnel side, HBR research from February 2026, ["Posting a Wide Salary Range Can Deter Women from Applying,"](https://hbr.org/2026/02/posting-a-wide-salary-range-can-deter-women-from-applying) finds that wide ranges disproportionately deter women, who show greater aversion to the financial uncertainty a broad band signals. So an over-wide range quietly works against diversity at the top of the funnel.

Market practice clusters tight. Among analyzed companies, the most common range width is **30%, meaning roughly plus or minus 15% from the midpoint**: about 23% of companies use a 20% width, 38% use 30%, 27% use 40%, and only 8% go to 50%, per [Pave](https://www.pave.com/blog-posts/how-wide-should-your-salary-ranges-be) and [AIHR](https://www.aihr.com/hr-glossary/range-spread/). Compensation practitioners commonly advise a spread of roughly 20 to 40%, with plus or minus 20 to 25% around the midpoint as a safe default, and 50% or more flagged as potentially problematic for good-faith purposes.

The cleanest way to set the band is to anchor it to market data: **floor near the 25th percentile, ceiling near the 75th percentile** for the role. That gives you a span that is defensible by construction, because the boundaries are tied to what the market actually pays rather than to round numbers you chose for comfort.

## How to benchmark a defensible range from real market data

Every compliance guide tells you to use market data and the 25th to 75th percentile. Almost none hand you the actual number for a specific role, with a documented methodology. Here is the mechanic, step by step.

1. **Classify the role into a job cluster.** "Backend Engineer," "Product Designer," "Data Analyst." A clear, objective job classification is the foundation every good-faith framework assumes you already have.
2. **Pull the market benchmark for that cluster.** Read the median, the 25th percentile, and the 75th percentile, along with the sample size the figures are based on.
3. **Set the posting range to the P25 to P75 span.** That span *is* your good-faith range. The floor and ceiling are tied to market reality, not to round numbers.
4. **Record the sample size and trend.** The number of listings behind the benchmark is your methodology footnote. A rising trend lets you defend the band as *current*, which the good-faith standard requires.
5. **Capture objective differentiators.** Stack, seniority, and location are legitimate, non-discriminatory reasons to set different bands for the same title. A Go-heavy backend role can sit higher than a PHP one; a major-metro role can sit above a lower-cost market.

A worked example. Suppose the market benchmark for a Backend Engineer cluster returns a median around $135,000, with the 25th percentile near $120,000 and the 75th near $150,000. Your posting range is **$120,000 to $150,000**, a roughly 25% spread, comfortably inside the safe zone and well clear of the over-wide territory enforcers flag. The sample size behind it, say several hundred listings, is the line you point to if a labor commissioner ever asks how you arrived at the number.

## Turn the benchmark into your good-faith audit trail

The reason most teams set ranges ad hoc is that the data and the methodology live nowhere. You find a number somewhere, paste it into a job post, and keep no record of how you got it. When the good-faith question comes, you have nothing to show.

This is exactly where **Kit's Compensation Research** closes the gap. Pull a salary benchmark for a role cluster and you get the median plus the P25 and P75 percentiles, *with a sample size attached*. The P25-to-P75 span is your compliant posting range; the sample size is your documented good-faith methodology. Market-trend breakdowns by region and technology supply the objective, non-discriminatory justifications for why two roles with the same title sit in different bands, plus the trend direction that proves the band is current.

The benchmark output becomes the record you keep on file: the median, the percentile band, the sample size, and the regional and stack breakdown for that exact role. Carry that same range into the posting when you [write the job description](/blog/writing-job-descriptions), and compliance and sourcing draw on one shared number instead of two disconnected guesses.

<div class="blog-inline-cta">
  <p><strong>Stop guessing at salary ranges.</strong> Kit pulls a market-anchored P25 to P75 band for the role, with the sample size as your built-in good-faith audit trail.</p>
  <p><a href="/users/sign_up">Start your free trial</a></p>
</div>

## Compliance and a better funnel are the same move

The 2026 patchwork turns "what should this role pay" into a legal forcing function with two failure modes: post no range, or post an indefensible one. Both are avoidable with the same action. Benchmark a narrow, market-grounded good-faith range, put it in the posting, and keep the benchmark output as your audit trail.

When you do that, compliance and competitiveness stop being a trade-off. A transparent, credible range satisfies the 16-state patchwork *and* wins the funnel, because fair numbers attract more and better-fit applicants while cutting late-stage drop-off. The same narrow band that survives a good-faith review is also the one that converts. And since a credible range is one of the few levers that reliably reduces [why candidates ghost you](/blog/why-candidates-ghost-you), the compliance work pays for itself at the top of the funnel. If you hire across the EU as well, the EU Pay Transparency Directive is the international counterpart to this US patchwork, and the same benchmarking discipline covers both.

The compliance deadline is not a tax on hiring. Treated correctly, it is the nudge that finally forces every range to be a real number, which turns out to be the number that performs best anyway.