'What Were You Making?' Is Now Illegal in Half the Country. It's Still the Cheapest Way to Underpay You.

The smartest thing you can do with the "what's your current salary" question is refuse to answer it, even when the person asking is friendly and the conversation feels casual. The number you give doesn't reveal your worth. It sets the ceiling on your next offer. Roughly 22 states have already outlawed the question because lawmakers figured out what most candidates haven't: the question exists to anchor you to your past pay, not to assess your value.

Here's what most people get wrong. They think disclosing helps them look honest and cooperative, and they assume a higher past number works in their favor. The opposite is closer to true. When you hand over your current salary, the employer doesn't price the role at market. They price you at your history. If that history includes being underpaid, and for a lot of people it does, you carry that discount forward into every job you take next.

Why does the question still get asked if it's banned?

Because a ban on asking is not a ban on answering, and candidates answer anyway.

The legal map is real. As of the April 2026 tally, 22 states have statewide salary-history bans plus 24 local jurisdictions with their own, starting with Massachusetts in 2016. But the question didn't disappear. It went informal. A recruiter mentions "we just want to make sure we're in the right range," a hiring manager asks "so what are you at now?" over coffee, an application form leaves a blank box. None of that requires a formal demand, and all of it gets the number out of you if you let it.

And candidates do let it. In one study tracking actual disclosure behavior, about 25% of job-seekers always volunteer their prior salary whether asked or not, 17% never do, and 58% comply with the ban. A quarter of people offer up the one piece of information most likely to cap their offer, unprompted. Worse, the researchers found the behavior "unravels": people disclose when they think others are disclosing, so the candidate who volunteers makes everyone who stays quiet look like they're hiding something.

The most expensive part of that finding is who's in each bucket. The same research shows always-disclosers skew male and compliers skew female. The people most likely to have been underpaid are the most likely to stay quiet, which is the right instinct, but they're surrounded by people advertising their numbers and setting a norm that makes silence look suspicious.

What actually happens to your offer when you disclose?

It gets anchored to your past, mechanically, even when a higher number should help you.

This is the part to sit with, because the evidence is unusually clean. In a controlled two-sided audit study, researchers found that applicants who disclosed salary history received higher mean recommended salaries, $103,993 versus $96,521 for non-disclosers. So far that sounds like an argument for disclosing. It isn't. The same study found recruiters passed only $0.42 of every $1 of salary gained through the gender wage gap into their valuation, versus $0.64 to $0.70 per $1 earned at high-wage firms.

Translate that. When your past pay reflects real market value, the recruiter mostly believes it. When your past pay reflects underpayment, the recruiter mostly bakes that discount straight into your next offer. The number anchors the offer either way. If you were underpaid before, disclosure forwards the underpayment. The offer tracks your history, not the role's market band.

There's a second mechanism underneath the first. Recruiters in that study estimated that disclosers had outside offers 9% higher than non-disclosers, treating the number you share as a read on your bargaining position. So your current salary isn't being used to figure out what you're worth. It's being used to figure out how little you'll accept. Those are different questions, and the second one never works in your favor.

What you assume disclosure does	What the data shows it does
Proves you're honest and easy to work with	Hands them the reference point they negotiate down from
A high past number raises your offer	Only $0.42 of each underpaid dollar carries forward; the gap compounds
Shows you have nothing to hide	Signals your bargaining floor, read as a 9% lower outside option
Keeps the conversation moving	Ends the negotiation before it starts, at your old number

How much does refusing to anchor actually move your pay?

Enough that lawmakers can measure it in the population, not just in a lab.

When candidates stop carrying their history forward, starting pay jumps. A 2024 analysis found new hires changing jobs in salary-history-ban jurisdictions earned an average 7.9% pay increase versus 3.9% in non-ban areas. That's roughly double the bump, purely from breaking the anchor. The effect is largest exactly where the old anchor was heaviest: newly hired women in ban jurisdictions saw raises 7.8% greater than peers elsewhere, and non-White job-changers saw a net 5.8% pay increase.

Academic work lands in the same place. Boston University researchers found that under salary-history bans, women earn 8% more and Black workers earn 13% more when they change jobs, against baseline gaps where men out-earn women by 15% and white employees out-earn Black peers by 11%. A separate analysis found bans reduced the gender pay gap by 2 percentage points in the states that enacted them. The mechanism behind every one of those numbers is the same: when nobody can anchor your offer to your old pay, the offer drifts back toward what the role is actually worth.

The deeper story here is the one I told in how salary-history bans quietly raised paychecks. The law didn't make employers generous. It removed a discount they'd been applying by default. You can apply that same removal yourself, in any state, by simply not handing over the number, because a law that stops the question from being asked does nothing if you answer it anyway.

What do you actually say when they ask?

You redirect to the role's market band, warmly, without ever stating your current number. The move is deflection, not confrontation.

The instinct most people have is to either blurt out the number or get visibly defensive about it. Both lose. The first anchors you low. The second makes you look difficult before you've even started. The deflection threads between them: you sound collaborative, you give them something useful, and the something useful is the market rate for the role, not your history.

Weak (volunteers the anchor): "Right now I'm at $82,000, but I'm definitely open to discussing what makes sense for this role."

That sentence just ended your negotiation. You named the floor. Everything after "$82,000" is decoration, and the offer will land a polite few thousand above it.

Strong (redirects to the band): "I'd rather not anchor on my current number, it's tied to a different scope. Based on my research, roles at this level and market are landing around $[X], and that's the range I'm targeting. Where does this role sit in that band?"

The strong version does three things at once. It declines without sounding evasive ("tied to a different scope" is true and final). It substitutes the market band for your history, which is the entire game. And it flips the question back: now they tell you their range. That flip matters more than it looks, because most employers won't volunteer the band on their own. Across Praxy's live index of 53,802 active job postings (June 2026), only about 9% disclose any pay figure at all, so for nine roles in ten the only way the range enters the conversation is if you make them name it. You've moved the conversation from "what will you accept" to "what does this role pay," which is the only conversation worth having. If they press, you hold:

"I hear you, and I'm not trying to be cagey. I'm just confident the right basis for this is the role and the market, not what I happened to be paid before. Happy to share my target range if that helps move us forward."

Notice you offer your target, never your current. Your target is a number you chose. Your current is a number someone else chose for you. Only one of them belongs in this conversation. For the full breakdown of why your target is a band and not a single figure, see why market rate is a range, not a number.

When is deflecting the wrong move? (the part nobody mentions)

When the deflection itself costs you more than the anchor would, and that happens more than the negotiation blogs admit.

Be honest about the cases where holding the line backfires. If you're currently overpaid relative to the market, your history is an asset, not a liability, and refusing to share it just throws away a high anchor that would have helped you. If you're in a tight specialist field where everyone already knows the comp bands, the deflection can read as gamesmanship over a number nobody's hiding. And if the recruiter is genuinely on your side, trying to confirm you're not wildly out of range before wasting your time, a flat refusal can stall a process that was about to move in your favor.

There's also the unraveling problem from earlier. In a room where a quarter of candidates volunteer their numbers, your silence carries information too. Most of the time that's fine. Occasionally it's the thing that makes a nervous hiring manager pick the candidate who "was transparent." You can't control the room, but you should know the dynamic exists rather than assume deflection is always free.

The honest frame is the same one I'd give for any negotiation move: deflect on purpose, not on reflex. If your history helps you, use it. If it caps you, bury it. The skill isn't always saying no. It's knowing which number you're protecting. And remember that even a perfect deflection only sets up the negotiation, it doesn't win it. The anchor matters because of how much of your offer is decided before the salary talk even starts, which is a separate and larger fight.

Does the timing of the question change how you handle it?

It changes everything, because the same question means different things at the screening call than at the offer.

Early, on the recruiter screen, the question is a filter. They want to disqualify you cheaply if you're out of budget. That's the moment to deflect hardest, because you have the least information and the most to lose from anchoring. You don't yet know the band, the scope, or whether they even like you. Naming a number here is pure downside.

Late, at the offer stage, the dynamics flip. Now they've chosen you, the cost of restarting their search is real, and the question has less power because the leverage has shifted. This is the trap most people invert: they clam up at the offer when they've finally got leverage, and they blurt out their number on the first call when they have none. If you want the full version of that calendar, the timing of the salary question is its own trap, and getting it backwards is one of the most common ways good candidates leave money on the table.

The rule of thumb is simple. The earlier the question, the firmer the deflection, because early disclosure is all anchor and no information. The later the question, the more you can afford to talk numbers, because by then you're talking about their band, not your history.

What to do now

1. Decide in advance whether your history helps or caps you. If your current pay is above market, you may want to anchor with it. If it's at or below market, commit to deflecting before the call so you're not improvising under pressure.
2. Set a target range from the role and market, not your salary. Triangulate from posted pay ranges, role-specific comp data, and one person actually in the job. That range is the only number you'll share.
3. Write your one-line deflection and your one-line hold. Two sentences, rehearsed, so the redirect comes out calm instead of defensive when a friendly recruiter catches you off guard.
4. Deflect hardest on the early screen, ease up at the offer. The question's power is inverse to where you are in the process. Match your firmness to your leverage.
5. Never substitute your current for your target. Offer the number you chose, never the number someone else chose for you. That single discipline is most of the game.

Want to rehearse the exact deflection before a real recruiter asks, or figure out whether your history is an anchor or an asset in your specific situation? That's the kind of thing I'm built for. Message me on WhatsApp and we'll set your target range, draft the line, and roleplay the call until "what are you making now?" stops rattling you.