Introduction
As soon as a project acquires land, interrupts access, displaces a dwelling or prevents the exercise of an economic activity, the question of involuntary resettlement arises. IFC Performance Standard 5 (paragraph 1) sets out the principle: avoid displacement where possible, minimise it when necessary, fully compensate affected persons and restore their livelihoods to a level at least equivalent to that before the project. AfDB Operational Safeguard 2 and World Bank Environmental and Social Standard 5 follow the same logic, with variations on the scope and categories of eligible persons.
The passage from principle to implementation is where projects stumble. A credible RAP is not a document drafted in the office and reviewed by a social expert; it is a sequence of stages, each mobilising a profession, a data source and a legitimacy vis-à-vis communities that cannot be simulated. When one of these stages is rushed, everything that follows depends on it and the mechanism cracks six months after the first payments.
This article presents the six stages which, correctly sequenced, produce a sustainable RAP. I know them from having led resettlement operations on road, hydropower and civil engineering sites in Central and West Africa. What follows is not an exhaustive manual; it is the map of breaking points: the moments when a poorly made decision costs the most, and the signals that allow them to be anticipated.
Stage 1: the census, a non-negotiable foundation
The census is the snapshot at a given date of all persons, assets and activities affected by the project. PS5 (paragraph 17) requires that it be conducted before publication of the cut-off date and that it physically cover all potential displaced persons, including occupants without recognised land title.
Three rules govern its quality. The cut-off date, officially set and communicated through channels appropriate to the populations concerned (radio, posting, announcements at village assemblies), freezes the eligible population. The coupling of cadastral survey and socio-economic survey guarantees that no tenant, customary usufructuary or economic right-holder is forgotten. Double entry, one in the field with GPS, photographs and signatures, the other centralised in an encrypted database, protects against subsequent disputes and falsifications.
The census produces a nominative register which becomes, from that moment, a sensitive document. Its security (restricted access, logging of consultations, off-site backup) is a point systematically examined in supervision. I have seen a project lose six months because a register had been copied to a personal USB stick by a data-entry operator; the incident had not compromised the data, but it had been enough to reopen the list of eligible persons under community pressure.
Stage 2: categorising affected persons
The frameworks distinguish, under different terminologies, three levels of rights. PS5 (paragraphs 17 to 19) provides the template.
Holders of a land title recognised by national law receive the most complete compensation (land, buildings, crops, income, moving costs). Occupants without title, but whose occupation predates the cut-off date, are entitled to compensation for assets and income; the land as such remains State property, but PS5 and ESS5 require that they be provided with a viable alternative (equivalent land or housing, plus support for asset reconstitution). Persons who settled after the cut-off date are not eligible for compensation, but their situation must be documented and, if vulnerable, be the subject of minimum assistance.
A poorly categorised RAP produces two symmetrical risks that appear simultaneously. The unjustified exclusion of vulnerable populations (widows, elderly persons, internal migrants) feeds complaints and, in contexts where rights-based NGOs monitor projects, attracts international attention. The opportunistic inclusion of late beneficiaries destroys the financial sustainability of the mechanism and opens the door to endless disputes.
Stage 3: valuation at full replacement cost
Asset valuation follows a guiding principle that PS5 (paragraph 9) formulates clearly: full replacement cost, that is the actual cost of acquiring an equivalent asset under local market conditions at the time of compensation. This concept is more demanding than market value, which may underestimate agricultural land, old productive trees or traditional buildings.
The choice between in-kind compensation and cash compensation is one of the most concrete tensions in a RAP. PS5 and ESS5 favour in-kind compensation for assets that constitute the bedrock of the household's economic life (housing, main cultivated land) and allow cash compensation for secondary assets, or at the explicit and documented request of the beneficiary.
The argument behind this hierarchy is not ideological. It derives from repeated observation that cash compensation paid in a lump sum, in contexts where financial inclusion is low, produces situations of reinforced vulnerability: squandering, intra-family tensions, capture by intermediaries. A RAP that pays everything in cash without support places itself at risk of reproach eighteen months after the end of payments, when the first signs of decapitalisation surface via the grievance mechanism.
Stage 4: the livelihood restoration plan
PS5 (paragraph 25) is explicit on this point: compensation alone is not sufficient; the project must ensure that affected persons can "restore or improve" their livelihoods within a reasonable timeframe. This timeframe is generally assessed at three to five years depending on the type of activity and initial vulnerability.
The plan is dimensioned by profile. An urban market gardener who loses his plot needs a substitute plot equivalent in agronomic quality and distance to market, and technical support over the recovery period. A fisherman whose fishing zone is fragmented by a bridge needs new access zones, equipment and a period of training in techniques applicable to the new sites. A small trader displaced from a central market to a secondary zone needs working capital, visibility for his new location and commercial support over several months.
When the plan merely lists generic training ("management awareness", "entrepreneurship support"), the lender detects it immediately. What distinguishes a credible plan from a decorative plan lies in the level of detail: who trains, for how many hours, with what individual follow-up, and especially with what success indicators. Without these indicators, restoration is not measurable and the RAP cannot be closed.
Stage 5: consultation and the grievance mechanism
PS5 (paragraph 10) requires "meaningful" consultation of affected persons, including on resettlement options and compensation arrangements. This requirement is not formal; it is a source of information without equivalent. Affected households know their territory better than any consultancy, and their observations make it possible to correct methodological biases that the socio-economic survey would not have identified alone.
Consultation must be adapted to the context: local language, visual materials for populations with low literacy, calendar respectful of economic constraints (no meeting at planting time), locations accessible on foot from the hamlets concerned. It produces a written record that is not only intended for the lender but for the participants themselves: distributed minutes, posting at the community centre, oral restitution at the next session.
The grievance mechanism, meanwhile, must function before, during and after displacement. It is probably the most closely observed health indicator of the RAP in supervision. A mechanism that receives no complaint in six months is not a perfect RAP; it is a RAP whose mechanism is inaccessible or perceived as illegitimate. I systematically examine the average processing time, the proportion of complaints resolved amicably and the rate of escalation to the next level, because these three indicators together tell the real state of the relationship with communities.
Stage 6: post-displacement monitoring, three to five years of discipline
A RAP does not close the day households have moved. PS5 (paragraph 22) requires monitoring over "several years" and requires the client to demonstrate that restoration objectives have been achieved before declaring resettlement complete. Three years is a common floor, five years the norm when the project has displaced agricultural activities or livelihoods dependent on a specific ecosystem.
Monitoring must provide for catch-up clauses. What happens if a household has not regained its previous income level by the initially planned deadline? Lenders examine this clause carefully, because it measures the sincerity of the restoration commitment. A clause that simply refers "to the appreciation of the steering committee" does not hold long; there must be a protocol defined in advance (third-party audit, supplementary compensation, targeted technical support).
Conclusion
The RAP is the deliverable where the distance between plan and reality is measured directly in people's lives. Its quality is judged neither by the thickness of the document nor by the sophistication of the mapping, but by the capacity of the mechanism to hold over time and to absorb the individual situations that the method had not anticipated.
The six stages presented here constitute the structural minimum. The rest, perhaps the essential, lies in the quality of the field work, the constancy of teams over the duration of the project site and the discipline of monitoring. No tool, no matrix, no indicator replaces this constancy.
A RAP that holds will be one that has considered affected persons as interlocutors, not as passive beneficiaries. This nuance runs through all documents, all decisions, and it reads in the tone of an audit report as it reads in a village assembly.
Comments
Be the first to react to this article.