Escalation: what happens when a project's own grievance mechanism cannot resolve a complaint

1. The tiers and what distinguishes them

The structure is conventionally described as escalating seniority: the project mechanism first, then courts, then the lender's accountability mechanism. The more useful description is escalating independence from the party complained about. A project mechanism is administered by the project — the tier PS-5 and its guidance require every affected-community project to maintain. [1][2] A court is independent of it but operates on national law, which may not recognise the standard the complaint invokes. An accountability mechanism is independent of the project and applies the lender's own standards.

This reframing explains the pattern of use. Complainants escalate not because they want a more senior decision-maker but because the previous forum could not or would not decide against the project. Where a project mechanism has no authority to award anything the project has not already offered, it cannot resolve a dispute about whether the offer was adequate, and the complainant's only route is outward.

The World Bank's Inspection Panel and IFC's Compliance Advisor/Ombudsman are the reference examples. [9][6] Both accept complaints from people who believe they have been harmed by a financed project, both require that the project's own channels have been tried, and neither is a court.

2. What escalation signals

A complaint reaching an independent mechanism should be read by the project as diagnostic information about its own mechanism. Reaching that point requires a complainant to know the mechanism exists, understand eligibility, produce a written submission in an international language, and persist through a process taking months — a set of capabilities most affected households do not have.

It follows that the complaints which arrive are not a sample of grievances. They are the small subset in which someone had unusual capability, usually through a supporting civil society organisation, and unusual determination. Behind each is an unknown number of people with the same grievance and less capacity to pursue it.

Projects that treat an escalated complaint as an isolated dispute with a particular household therefore mis-read it. The appropriate response includes asking why the project mechanism did not resolve it, and how many similar situations exist among households that did not escalate — a question the project can answer from its own records and rarely does.

3. What the mechanisms can deliver

Independent mechanisms typically offer two functions: a dispute resolution or mediation function, and a compliance review function. [7] They differ fundamentally and complainants frequently do not understand the choice they are being offered.

Dispute resolution can produce a negotiated outcome including tangible remedy, but requires both parties to participate voluntarily and produces no finding about whether the standards were breached. Compliance review investigates whether the institution complied with its own policies and produces a finding — but the finding is about the institution's conduct, and the mechanism generally cannot order compensation to the complainant.

This is the structural limitation at the centre of the accountability literature. A complainant may obtain a finding that the institution failed to comply, and no entitlement to anything as a result. The management response and any corrective action plan are the route by which a finding might become remedy, and as examined elsewhere in this series, corrective action plans reliably favour procedural fixes over individual restoration. [8]

The gap is visible against the framework the UN Guiding Principles on Business and Human Rights set out, in which access to effective remedy is a distinct pillar rather than a by-product of compliance review. [10] Judged by the criteria that framework applies to non-judicial grievance mechanisms — legitimacy, accessibility, predictability, equitability, transparency, rights-compatibility — the weakness of the project tier is not that it lacks seniority but that it is frequently neither predictable nor equitable, and the tiers above it were never designed to supply an individual award.

4. Designing a project mechanism that resolves

If escalation indicates failure below, the design question is what a project-level mechanism needs in order to resolve genuine disputes. The first requirement is authority: a mechanism that can only explain the project's decision cannot resolve a complaint that the decision was wrong. Some capacity to vary an outcome — recheck a measurement, revise a valuation, recognise an omitted asset — is what distinguishes a mechanism from an explanation desk.

The second is that it must not be judged by the party whose decision is complained of. Where the valuation team reviews complaints about valuation, the complainant's assessment that the process is not independent is correct, and no amount of procedural documentation changes that.

IFC's handbook and the World Bank's disclosure guidance both treat timeliness against the complainant's own circumstances as a design requirement, not an aspiration. [3][4] The third is timeliness matched to the complainant's situation rather than the project's reporting cycle. A complaint about compensation for a field is time-critical against a planting season; a resolution arriving after the season has passed is not a resolution. Mechanisms that report average closure times in months are describing a process that cannot help with the disputes that matter most.

5. Conclusion

The tiered structure is coherent and its weakest link is the tier closest to the affected person. [5] Project mechanisms are the only forum most complainants will ever reach, and they are frequently designed without the authority to decide anything, which converts them into a channel for receiving complaints rather than resolving them.

The independent mechanisms above them are important, hard to reach, and constrained in what they can award. Treating them as the safety net for a weak project mechanism misunderstands both: they are the last resort for the very few, and their existence does not discharge the obligation to provide a working remedy locally.