Emergency and disaster assistance: speed versus proof when both are mandatory
Every disaster produces two mandates that point in opposite directions: get money to survivors immediately, and account for every dollar afterwards. The second mandate arrives with an inspector general six months later and judges decisions made in the first seventy-two hours.
Can disaster assistance be both fast and verified?
Yes, if the two are separated. Emergency programmes must pay quickly and cannot prove who anyone is, and treating speed and assurance as a single dial produces either fraud or people sleeping in cars. Tiering by amount and by reversibility lets both be true at once.
- Emergency programmes relax identity requirements uniformly, which applies the relaxation equally to a $500 payment and a $30,000 one.
- Post-payment recoupment recovers little and imposes hardship on survivors who received assistance in good faith.
- The design needs a false-exclusion budget stated in advance, because the trade between exclusion and fraud is irreducible.
Part of Public sector identity
The conditions the control must survive
Design for the actual environment rather than for a tabletop. In the days after a major event:
- Applicants may have no identity documents, because the documents were in the house.
- Mobile networks are congested or down; power may be intermittent.
- People are displaced, so the address of record is wrong and mail is undeliverable.
- Staff are surged, including personnel unfamiliar with the programme.
- The applicant may be using a borrowed device or a shelter terminal.
A control that assumes a smartphone, connectivity and documents has assumed away the disaster.
Why uniform relaxation is the wrong instrument
Programmes relax requirements globally when an emergency is declared, because differentiating takes time nobody has. The consequence is that the same reduced assurance applies across the value range.
| Assistance type | Typical value | Relaxation applied | Appropriate? |
|---|---|---|---|
| Immediate needs payment | Hundreds | Full | Yes — speed dominates |
| Rental assistance | Thousands | Full | Questionable |
| Home repair assistance | Tens of thousands | Full | No |
| Business disaster loan | Hundreds of thousands | Partial | No |
The bottom two rows are where organised fraud concentrates, because the return per fraudulent application justifies effort. The top row is where the humanitarian urgency is, and it is also where the loss per incident is smallest.
Tiering, with the low tier deliberately weak
The proposal is not to strengthen everything. It is to keep the immediate tier as fast as it is now, and to add assurance where value concentrates.
- Tier 1 — immediate needs. No change. Pay fast, accept the loss rate, recoup nothing. The administrative cost of recouping small payments exceeds their value and the hardship it causes is disproportionate.
- Tier 2 — rental and temporary housing. Credential enrolled at application if possible; assisted path at the disaster recovery centre; payment proceeds either way with a documented basis.
- Tier 3 — repair, replacement and loans. Authorisation required, with an assisted path that is properly staffed, and a documented exception route for genuinely unreachable applicants.
The offline requirement
At a disaster recovery centre with a satellite uplink shared by forty staff, any verification requiring a round trip to a central service will time out.
Verification against a published key, performed locally with no network call, is not a nicety here. It is the difference between a control that works at the point of service and one that gets switched off on day two.
The false-exclusion budget
State it before the event, not during. What share of eligible applicants may a control delay beyond a defined threshold before the control should not be used?
Then measure it during the response and report it alongside the fraud figures. Programmes currently report detected fraud and recoupment and do not report how many eligible people were delayed or excluded by integrity measures, which makes the tradeoff invisible to everyone reviewing it afterwards.
Why recoupment is not the answer
Post-payment recoupment is the current backstop and it performs badly on both dimensions. Recovery rates against organised fraud are low, because the money and the identities are gone. Recovery against ordinary eligible recipients who were later determined ineligible is higher — and it means pursuing disaster survivors for money the programme gave them.
That is a bad outcome arrived at honestly, and it is an argument for getting the authorisation right at payment rather than for chasing it afterwards.
A tier model for emergency assistance
| Payment | Reversibility | Assurance |
|---|---|---|
| Immediate needs, small | Effectively none | Minimal — speed wins |
| Interim, moderate | Partial | Identity proofing, deferred |
| Major repair or replacement | Recoverable in principle | Full proofing before payment |
| Ongoing assistance | Recoverable | Proofing plus periodic re-verification |
Objections and honest limits
“Any deferral means people wait.” Not if the first tier pays immediately. Deferring assurance is different from deferring money, and conflating them is how programmes end up choosing between the two failure modes.
“Fraudsters will take the first tier.” Some will, and the first tier is bounded by design. The question is whether the bound is set deliberately or discovered afterwards.
Designing the tiers
- Pay the first tier immediately. Bounded, and accept the loss rate as a policy decision.
- Defer assurance, not money. The two can move independently.
- Raise the bar with amount and irreversibility. Both axes, not just amount.
- Re-verify for ongoing assistance. Where the cumulative exposure actually sits.
Terms used here
- Deferred proofing
- Paying first and establishing identity afterwards, bounded by amount.
- Reversibility
- Whether a payment can realistically be recovered, which should set the assurance bar alongside amount.
- Tier bound
- The deliberately chosen maximum exposure in the fastest, least-verified tier.
Frequently asked questions
Would this slow down immediate assistance? No. The design explicitly leaves the immediate tier unchanged, because speed dominates at that value and the loss per incident is small.
How do you authorise someone with no documents? The credential is enrolled at application, not derived from documents. Enrolment establishes continuity between the applicant and later transactions; it does not prove who they are.
What about applicants who cannot enrol at all? A documented exception route with a recorded basis. Exceptions should be counted and reported rather than hidden.
Is any of this required? No. Programme integrity requirements address eligibility and documentation. Authorisation architecture under degraded conditions is not prescribed.
Doesn't deferring verification invite fraud? It bounds it. The first tier is capped by design, and the alternative — verifying before any payment — excludes people who have just lost their documents.
Why does reversibility matter as well as amount? Because a recoverable payment tolerates a lower bar than an irreversible one of the same size.
Where is the real cumulative exposure? Ongoing assistance, which is why periodic re-verification belongs there rather than at the emergency counter.
Where this fits in Manav
Manav binds the authorising official to the exact release, award or record change, and produces a receipt another agency, an auditor or a counterparty can verify without access to the issuing system.
Sources and further reading
- Federal disaster assistance programme requirements and eligibility rules.
- U.S. Government Accountability Office reports
- 22 CFR Parts 120-130 — International Traffic in Arms Regulations
- GAO-15-593SP — A Framework for Managing Fraud Risks in Federal Programs