Disbursement conditionality: what happens when compensation sits on the critical path for money

1. Why the financial link exists

The requirement that people be compensated before they are displaced is stated plainly in every major standard and is breached routinely. [1][5] The reason is structural rather than ideological: displacement is on the critical path for construction, compensation is administratively slow, and the party that suffers from a delayed payment is not the party deciding whether to proceed. Left to its own incentives, a project will move earth and settle claims afterwards.

Reporting does not correct this, because the breach is discovered after it has occurred and the remedy is a finding. Supervision does not correct it, because missions are periodic and the decision to proceed is daily. What corrects it is making the money conditional, so that the cost of proceeding early is borne by the party with the authority to proceed.

This is the underlying logic of disbursement conditionality, and it explains why the instrument is disproportionately effective relative to its administrative weight. It does not persuade anyone of anything. It changes who bears the cost of a breach.

2. How the condition is usually framed

The typical formulation withholds disbursement for civil works on a defined section until the borrower demonstrates that affected people on that section have received their compensation and any assistance due before possession. [1][4] The section-level framing is essential: a project-wide condition would be unworkable on a corridor, and a condition without geographic granularity cannot be tested against the works actually about to begin.

Demonstration usually requires a certified statement supported by underlying records, and sometimes independent verification. The stronger versions require the verifier to be someone other than the party being paid — an independent monitor or the lender's own specialist [2][3] — precisely because self-certification under schedule pressure is unreliable in a predictable direction.

Sophisticated structures also address the hard residue. Households that cannot be paid — unlocated owners, disputed titles, estates in succession, documented refusals — are handled through escrow or deposit arrangements with defined evidentiary requirements, so that a small number of genuinely intractable cases cannot hold an entire section hostage, while also not simply disappearing from the count.

3. What conditionality distorts

Any indicator tied to money is an indicator under pressure, and compensation completion is no exception. Where the condition is expressed as a percentage of households paid, the pressure is to reach the percentage, and the cheapest route is not always the honest one.

The observable distortions are consistent across settings. Households under-record their losses because a smaller claim settles faster. Complex cases are reclassified as out of scope rather than resolved. Signatures are obtained on agreements the signatory has not understood, because a signed agreement counts as settled. Payments are recorded as made when a cheque has been issued rather than when funds have been received, which is a materially different fact for a household without a bank account.

The pattern is consistent with what accountability mechanisms find on review of completed disbursement conditions. [6] None of this requires anyone to act in bad faith. It requires only that the metric be easier to satisfy than the underlying objective, and that people be under schedule pressure — which describes every large project. Conditionality drafted against a proxy therefore produces movement in the proxy, and the household experience it was designed to protect can deteriorate while the reported figure improves.

4. Choosing the milestone

The design question is what to condition on, and the answer is the fact closest to the household's actual position. Funds received and available for use is closer than payment authorised. Replacement housing occupied is closer than replacement housing constructed. A grievance resolved to the complainant's recorded satisfaction is closer than a grievance closed by the project.

Each of these is harder to evidence than its proxy, which is why proxies are chosen. The trade-off is real and should be made consciously: a project that conditions on the easy version is buying schedule certainty at the cost of the protection the condition was meant to provide.

A second design principle is that the condition should be capable of being satisfied. A condition that cannot be met — because it requires an outcome outside the borrower's control, or perfection in a population where perfection is unattainable — will be waived, and a waived condition teaches everyone that conditions are negotiable. It is better to condition on something demanding and achievable than on something absolute and therefore ceremonial.

5. Conclusion

Disbursement conditionality is the most powerful safeguard instrument in routine use, and its power is exactly proportional to the honesty of the milestone it selects. Attached to funds received by an identified household on an identified section, it does what no amount of supervision achieves. Attached to a percentage in a progress report, it produces a well-managed percentage.

The instrument therefore concentrates a great deal of consequence into a drafting decision taken early, by people negotiating a finance agreement rather than administering a resettlement programme. That is an argument for involving the people who will have to satisfy the condition in the drafting of it — which is rare, and cheap.