Livelihood restoration: the part of a RAP that outlives the payment

Compensation is a transaction. Livelihood restoration is an outcome. The standards require the second, and a project that has completed every transaction can still have failed the requirement — which is the single most consequential thing to understand about this part of a RAP.

For economically displaced people, PS-5 requires that livelihoods be restored, measured against pre-project conditions.[1] ESS-5 sets the same test.[3] Neither is satisfied by evidence that money was paid.

Why payment is not restoration

A farming household that receives full replacement value for its land, in cash, in a single payment, has been compensated. Whether its livelihood has been restored depends on what happens next — whether equivalent land was available to buy, whether the household had the capacity to manage a large lump sum, whether the new plot has water, whether the new location has a market.

The failure pattern is well documented and consistent: cash compensation is received, spent on immediate needs and obligations within months, and the household ends up landless with no income stream. This is not a failure of the household. It is a foreseeable consequence of converting a productive asset into a lump sum in an environment with no replacement asset to buy.

Compensation restores a balance sheet. Livelihood restoration has to restore an income.

When an LRP is a separate document

Where a project causes economic displacement without physical displacement — a transmission line taking a strip across farmland, a restriction cutting off access to a grazing area — the instrument is usually a standalone Livelihood Restoration Plan rather than a full RAP. Where both occur, livelihood restoration is a chapter within the RAP.

The distinction is administrative. The substantive requirement is identical, and the mistake to avoid is treating economic-only displacement as the lighter case. A household that keeps its home and loses its field has lost its income and kept its costs.

What an LRP has to do

The IFC's handbook module on livelihood restoration is organised around a sequence that is worth following, because each step depends on the one before it:[2]

1. Establish the baseline. Income sources, their relative contribution, seasonality, and who within the household controls each. Without this there is no benchmark to restore to.
2. Segment affected households by livelihood type. A commercial farmer, a subsistence farmer, a market trader and a wage labourer face different losses and need different measures. One programme offered to all four restores perhaps one.
3. Design measures against each segment's actual loss. Land-based restoration for land-based livelihoods; enterprise support for traders; employment linkage for wage earners.
4. Set measurable targets tied to the baseline, not to activity counts.
5. Monitor against them at defined intervals, with the ability to change course.
6. Complete only when the targets are met, not when the budget is spent.

The measurement trap

Livelihood restoration programmes are frequently reported in outputs: households trained, demonstration plots established, business grants disbursed. These are convenient because they are countable and they arrive early. They also do not answer the question the standard asks.

Training 400 households is an output. Whether household income has returned to baseline is the outcome. A programme can achieve every output target and restore nothing, and a completion audit is looking at the second column.

Vulnerable households

Both standards require particular attention to those least able to absorb the shock — elderly-headed households, households headed by women with no other income, people with disabilities, the landless.[1] The common failure is not omission from the plan. It is that the plan names a vulnerable category, defines no criteria, and produces no evidence that every household was screened.

A vulnerability row that cannot be traced to a screening record proves nothing at audit. The screening has to happen at census, against written criteria, with the result recorded per household — including the negatives.

Transitional support that can extend

Transitional allowances are typically set for a fixed period. If livelihoods have not recovered when that period ends, the fixed design becomes the problem: the standard's test is restoration, not the expiry of a schedule.

A well-built LRP makes extension a monitoring-triggered decision rather than a budget exception — with the trigger, the evidence and the decision all recorded. That record is also what allows a project to defend the opposite decision, that support ended because the targets were genuinely met.

Closing out

Livelihood restoration is the longest-running commitment in a RAP, frequently extending years past the last compensation payment and past the point where project staffing has wound down. Handover of the monitoring obligation — to whom, with what records, against what targets — is worth designing at the start, because the alternative is a completion audit arriving to find the programme was closed by attrition rather than by evidence.[4]