You send the protected veteran self-identification invitation. The applicant picks "I do not wish to answer." The box sits blank in your system.

Now what? This is one of the most common VEVRAA questions, and one of the easiest to get wrong.

Some HR teams treat a blank as a failure. They chase the applicant for an answer. Others quietly fill the gap themselves. Somebody spots a DD-214 in the file and marks the person as a protected veteran. Both moves create real risk.

A decline is a legal outcome. The applicant has the right to say nothing at all. VEVRAA builds that right straight into the rule. Your invitation has to say the answer is voluntary. It also has to say that saying no brings no bad treatment. So when someone takes you at your word, the system worked.

What matters next is what you write down. Then what you never assume. Then how the blank flows into your benchmark math. Then how you ask a second time. Those four steps are where contractors slip.

This guide picks up where the invitation leaves off. Still building the invitation itself? Start with our companion piece on how to invite applicants to self-identify. Come back once the declines start stacking up.

Everything below tracks 41 CFR Part 60-300 and the live OFCCP figures as of August 2026. Dollar thresholds and the benchmark percentage move. Check the source pages before you lock anything into a plan.

Is a decline actually a compliance problem?

Usually not. A decline is the rule working the way it was written.

Look at 41 CFR 60-300.42. The invitation must tell the applicant six things. You are a federal contractor required to take affirmative action for protected veterans. The invitation must also summarize the law and your affirmative action plan. The request is voluntary. The answer stays confidential. Refusing brings no adverse treatment. And the answer will not be used in a way the law forbids.

Read that list again. You cannot promise all of that and then treat a blank as a defect. Your duty under VEVRAA is to ask, and to ask correctly. Your duty is not to land an answer.

This matters most for covered contractors. Basic VEVRAA coverage now starts at a single federal contract of $200,000 or more. That threshold rose from $150,000 on October 1, 2025. A written affirmative action plan usually kicks in at 50 or more employees plus that same contract size.

VEVRAA itself sits in statute at 38 USC 4212. It stayed in force through the 2025 changes that ended the old Executive Order 11246 program. If your compliance team thinks the whole thing went away, they are wrong on the veteran side. Our VEVRAA compliance overview walks the full duty set.

What do you record when an applicant declines?

You record the decline. That sounds obvious. Most systems still get it wrong.

The common failure is collapsing three different answers into two buckets. Your data needs to hold all three states separately.

The first state is a person who self-identified as a protected veteran. The second is a person who answered and said they are not one. The third is a person who declined or left it blank. A decline is not a "no." Rolling declines into the "not a veteran" column corrupts your numbers.

Alongside the response, keep proof that the invitation went out. Log the date, the stage, and the version of the form used. If your career site changed forms in March, you want to know which applicants saw which version.

Retention rules live in 41 CFR 60-300.80. The general period is two years from the record or the personnel action, whichever comes later. It drops to one year for smaller contractors. That means fewer than 150 employees, or no government contract worth at least $150,000. Either one on its own is enough.

Some records run longer. The annual applicant and hire computations under 60-300.44(k) get three years. So does your outreach documentation and your benchmark documentation. Keep those on a separate clock.

What can you never infer about veteran status?

This is the section worth reading twice. Self-identification is self-reported by design. You do not get to supply the answer.

A DD-214 in a file is not a self-identification. Neither is a resume that lists an Army unit and a rank. Neither is a four-year gap you decided to read as service. Neither is a recruiter's hunch from a hiring event. A profile photo in uniform is not one either. None of these is the applicant's answer, so do not enter one.

The data rule at 41 CFR 60-300.44 counts applicants who self-identified. It adds those "otherwise known as protected veterans." That phrase trips people up. The rule never defines it. The safe read is narrow. Count a person whose status you already hold from a clear disclosure. A current employee who self-identified at onboarding is the clean example. Do not count a status you guessed at from a document.

There is a second trap. Veteran and protected veteran are not the same thing. Someone can serve honorably for six years and still fall outside all four protected categories. Only certain veterans count for VEVRAA purposes. Our breakdown of the four protected veteran categories covers who does and does not qualify.

So even a hiring manager who knows for certain the person served cannot fill in the box. Service alone does not tell you the category. The applicant is the only reliable source.

How do declines change your benchmark math?

Here is where a lot of teams quietly panic. The math is simpler than it looks.

Your veteran hire rate has two parts. The top number is protected veteran hires. The bottom number is total hires. A hire who declined never adds to the top number. That person still adds to the bottom number.

So a high decline rate pushes your measured rate down. That is arithmetic, not a hiring failure. You may be hiring plenty of veterans and still show a low percentage.

The benchmark you compare against comes from 41 CFR 60-300.45. You may adopt the national figure or build your own from five factors. The national figure sits at 5.1 percent, effective July 30, 2025. That was still the operative number as of this writing. Pull the current one from the OFCCP hiring benchmark page each year rather than reusing last year's slide. Our explainer on why the benchmark changes covers the pattern.

Say it plainly to your leadership: the benchmark works as a yardstick, and falling under it does not put you in violation by itself. It flags the years you should look harder at your outreach.

Keep those five figures for three years and write down the comparison you made. Our guide on documenting the benchmark every year shows the paper trail a reviewer expects to see.

Does a pre-offer decline block the post-offer invitation?

No. Those are two separate duties, and you owe both.

Subsection (a) of 60-300.42 covers the pre-offer ask. It goes to applicants before any job offer. Subsection (b) covers the post-offer ask. It goes out after the offer and before the person starts job duties.

Nothing in the rule lets a pre-offer decline cancel the post-offer invitation. Suppressing the second ask because of the first is a configuration mistake, and it is a common one. Someone builds conditional logic in the applicant tracking system to be polite. The result is a missing invitation at a stage the rule covers.

The second ask often lands differently. At the pre-offer stage the applicant does not know you yet. The question can feel like a screen. After an offer, the person has the job and the stakes drop. Many people who skipped the first form answer the second one.

One more note for that stage. Say someone self-identifies as a disabled veteran post-offer. The rule says you should ask whether an accommodation is needed. Then work with them on it. That conversation belongs to onboarding, not to the hiring decision.

What does a low self-ID rate usually mean?

Most of the time it points at the invitation, not at your applicant pool.

Think about what the applicant sees. They are 20 minutes into an online application. A screen appears asking about military service, disability status, and a law they have never heard of. There is a wall of statute language about campaign badges and separation dates. The fastest path forward is "I do not wish to answer."

A few specific things drive that behavior:

• Placement: the form shows up after a long application, when patience is gone.
• Legal wording: the four categories are written in statute language nobody parses on a phone.
• Missing why: the applicant sees no reason you are asking, so they assume it is a screen.
• Buried confidentiality: the promise is there but sits in small gray text below the buttons.
• No default clarity: the person genuinely does not know whether they qualify, so they skip it.

That last one is bigger than most teams think. Plenty of veterans have no idea they hold a campaign badge that puts them in a protected category. They are not hiding. They just do not know the answer, and the form does not help them find it.

How do you fix an invitation people skip?

Fix the form first, then fix where it sits. These changes cost little and move the number.

OFCCP publishes a sample self-identification form you can start from. You may use your own wording as long as it carries the required statements. Whatever you build, keep it consistent across every job and every location.

What does OFCCP look at when self-ID rates are low?

Reviewers rarely open with the percentage. They open with your process.

A compliance evaluation looks at your veteran program on paper. Reviewers ask for your hiring benchmark for the current plan year. They ask for your outreach and recruitment assessment. They ask for the applicant and hire data 60-300.44(k) requires. OMB approved discontinuing the OFCCP scheduling letter in July 2026. So do not plan against the old itemized listing. Work from whatever request you actually receive. The question underneath all of it is whether you ran the program at all.

Subsection (f)(3) of 60-300.44 is the one to know. You have to review your outreach and recruitment each year and judge whether it worked. If it did not work, you have to pick other methods and actually use them. A low self-ID rate two years running with the same unchanged outreach plan is the finding. The percentage itself is not.

So build the paper trail. Show the invitation wording and when it changed. Show the annual data. Show the assessment you ran and what you did after it. Our walkthrough of the OFCCP compliance evaluation phase by phase lays out the sequence. The piece on building good-faith-effort outreach records covers what those files hold.

One caveat on timing. OFCCP has proposed changes to the Section 503 and VEVRAA rules. That rulemaking was still moving as of mid-2026. The current VEVRAA regulations stay in effect until anything is finalized. Watch the agency pages rather than acting on a proposal.

What should you do this quarter?

Start with a short audit of your own system. Pull 50 recent applications and check four things.

Does the pre-offer invitation carry all six required statements? Does the post-offer invitation still fire after a pre-offer decline? Does your applicant tracking system store declines as their own value? And can you produce the five annual data points without a manual rebuild?

Then look at your veteran section. If your plan has not been touched since last year, the review is overdue. Our guide to the veteran section of a VEVRAA affirmative action plan covers what belongs in it. The piece on what contractors actually track covers reporting.

Cleaning up self-ID data helps you measure. It does not by itself put more veterans in your pipeline. That part takes sourcing.

Best Military Resume runs a talent pool built for that gap. More than 1,000 new profiles are added every month. Over 65,000 resumes have been built on the platform. Candidates are searchable by field, location, and background. A hiring team can work a real list instead of waiting on applications.

Want access to that pool? Start at our employer hiring page. Tell us what roles you are filling.