First-line inspection exists to produce a defensible decision quickly: accept or refer. It does not exist to determine whether a document is genuine or false. Confusing those two purposes is the most common reason first-line checks take too long and still end inconclusively.
The boundary
First line works with handheld instruments in seconds to a minute, on a document type the position knows, against reference material for that type. The output is a decision plus a record.
Second line has time, comparison material, and instruments that first line does not: higher magnification with image capture, infrared examination, retroreflective viewers, and access to verified specimen data. The output is a determination.
The boundary is not about competence. It is about what can be established with the available time and tools. An inspector who tries to reach a determination at the counter is working outside both.
Findings that warrant referral
- An unresolved substrate response. The material behaves unexpectedly under ultraviolet and reference material does not explain it.
- Print quality that does not match expectation at magnification, such as raster patterns where continuous line work belongs.
- An optically variable feature that does not behave correctly after you have tried the angle properly. See security features that depend on viewing angle.
- Data inconsistency between two representations, for example between the visual zone and the machine readable zone. See the MRZ.
- Evidence of alteration: a lifted laminate, thinned substrate, a photograph that sits on rather than in the document.
- A document type outside the scope of the position. Improvising on an unfamiliar type is worse than referring it.
- An unresolved impression with no specific finding. This is a legitimate reason, and it should be recorded as an impression rather than dressed up as a finding.
Findings that do not warrant referral
Equally important, because unnecessary referrals erode the credibility of the route.
An unfamiliar but genuine older model. A truncated name in the machine readable zone, which is by design. Wear consistent with age and use. A feature you could not see until you corrected the lighting geometry. A transliteration that looks wrong but follows convention.
Most of these come down to reference material. A position with current and previous models for the document types it sees refers considerably less often and more accurately. See building reference material.
What to hand over
A referral without a record is a referral that starts from zero. What second line needs:
- The specific observation, with the area of the document and the condition under which it was made.
- What was expected instead.
- Which checks were performed, including the ones that found nothing.
- Any images, with the condition noted alongside.
- Who and when.
That is the same structure as any usable finding, set out in recording findings. It takes under a minute and it is the difference between a referral that can be worked on and one that has to be redone.
Why trying to resolve it at the counter costs more
Three reasons, and the third is the one usually overlooked.
The time is not there, so the attempt is rushed and the conclusion unreliable. The instruments are not there, so the categories of check that would actually settle it cannot be performed. And repeated handling of a document that may be evidence degrades it, which matters if the case goes further.
There is a fourth, practical one: a queue. A position that spends fifteen minutes on one document creates pressure to wave the next five through, and that is a worse outcome than a single referral.
Making referral a normal outcome
If staff believe referring reflects badly on them, they will accept documents they are unsure about. That is a management issue rather than a training issue, and it undermines everything else in the process.
Say explicitly that referral is a correct outcome. Define who receives it. Review referrals periodically, not to judge the decisions but to improve the reference material and to see whether observations are becoming more specific. A position that never refers anything is not necessarily performing well. See training staff for ID document checks.
Frequently asked questions
Can I refer without a specific finding?
Yes. Record it as an unresolved impression and refer. That is legitimate and it is better than accepting a document you doubt.
Who should second line be?
A named function inside the organisation, or the relevant authority. What matters is that the route exists and staff know it.
How long should a first-line check take before escalating?
If the standard sequence does not resolve it, refer. Extending the attempt at the counter rarely helps.
Does referring mean I think it is a forgery?
No. It means the check could not be completed with the available time and instruments.
What if the person becomes difficult?
That is a procedural and safety matter for your organisation, and it should be covered separately from the document check itself.
Equipping a position
The better a first-line position is equipped, the fewer avoidable referrals it makes. For an overview see document authentication equipment and the 15x Magnifier DXQ range, or request a quote describing your setting.
