The cut-off date: setting it, disclosing it, and defending it later

Eligibility in a resettlement action plan turns on a single date. People present in the project area before the cut-off date are eligible; people who arrive after it are not. Almost every eligibility dispute a project will face, and a large share of the grievances it will receive, resolve back to whether that date was properly set and properly disclosed.

It is a deceptively simple instrument, and it fails in predictable ways.

What the cut-off date is for

The purpose is narrow: to prevent an influx of people moving into the project area to claim compensation once the project becomes known.[1] That risk is real. Announcing a compensation programme in an area of high land pressure without a fixed eligibility boundary invites speculative construction and speculative planting, which inflates cost and — more damagingly — displaces genuine claimants in the queue.

Note what the cut-off date is not for. It is not a device for excluding people who were always there but were missed by the census. That distinction is the crux of most disputes, and it is the one projects most often get backwards.

A cut-off date bars people who arrived after it. It does not bar people who were present before it and were simply not recorded.

Setting it

The cut-off date is normally the completion of the census and asset inventory in a given area.[2] Tying it to census completion rather than an administrative decision matters, because the census is what produces the evidentiary record of who was present. A date fixed months before enumeration leaves a gap in which the project asserts eligibility boundaries it cannot evidence.

On linear projects — roads, transmission lines, pipelines — enumeration proceeds in sections over many months. Applying one project-wide date to sections surveyed a year apart is not defensible. The workable approach is a cut-off date per section, each tied to that section's enumeration, each disclosed locally. It is more administrative work and it is the only version that survives scrutiny.

Disclosing it

An undisclosed cut-off date is not enforceable in any meaningful sense. Both standards require that the date be documented and disseminated throughout the project area in a form affected people can actually act on.[1][3]

"In a form they can act on" is the operative test, and it rules out a surprising amount of common practice. A notice in an English-language national newspaper does not reach a subsistence farming community. A notice in a district office reaches people who visit district offices. Disclosure that works generally means several channels at once:

• public meetings in each affected community, in the local language;
• written notices at locations people actually pass;
• local radio, which in rural East Africa outperforms every print channel;
• briefing local leadership structures, who become the practical relay.

Each of these needs a record — date, location, attendance, language, materials used. That evidence pack is what makes the date enforceable later. Without it a project is asserting a boundary it cannot prove anyone was told about, and every exclusion made on the basis of that date becomes contestable.

When the date slips

Projects are delayed. A cut-off date fixed at census completion and then followed by a three-year funding gap creates a genuine problem: households have formed, children have built on family land, structures have been extended. Enforcing the original date strictly against three years of ordinary life produces exactly the injustice the standard exists to prevent.

The accepted response is an update census before implementation, revising the inventory while retaining the original date's function of excluding opportunistic entry. What distinguishes a legitimate update from a reopened free-for-all is documentation: recording natural household growth and ordinary construction as such, and distinguishing it from speculative activity. That determination has to be made contemporaneously, by people who know the area, and written down at the time.[4]

The record that has to exist

Two years after the fact, if a household claims they were present before the cut-off date and were wrongly excluded, the project needs to be able to answer. Answering requires:

• the declared date for that specific section, and the basis for it;
• the disclosure evidence pack showing how and when it was communicated there;
• the census record for that location, including who was enumerated and by whom;
• any subsequent determination about this household, with reasons and a date.

None of this can be reconstructed. It either exists, timestamped and attributable, or the project concedes the dispute — which in practice means paying claims it cannot verify, or excluding people it cannot prove were ineligible. Both outcomes are expensive, and only one of them is unjust.