Corruption risks in Uganda land acquisition: where the RAP control chain can break

1. Name the risk precisely

Corruption is not a synonym for every error, delay or unfair outcome. The Anti-Corruption Act, Cap. 116 creates offences [1] and the Inspectorate of Government investigates and reports on accountability matters. [4] The Leadership Code Act adds declaration and conflict-of-interest duties for officials, which is the provision most often engaged where a decision-maker holds an interest in affected land. [2] A responsible RAP analysis should therefore say whether it is describing a control weakness, a conflict of interest, an allegation, an investigation or a proven offence.

That precision matters for people and institutions. Unproved accusations can damage legitimate officials and claimants; vague language can also hide a real risk from the control system.

2. The concentrated decision points

The predictable risk points are eligibility inclusion, asset measurement, rate selection, exception approval, claimant identification, payment routing, procurement of consultants and closure certification. Each involves discretion, information asymmetry and a person who may not know how to challenge the decision. Procurement of consultants is governed separately under the PPDA framework, which is why a valuation conflict can be a procurement problem before it is ever an offence. [3]

The control response is separation of duties, versioned records, independent review, transparent exception reasons, conflict declarations, claimant-held copies and a grievance route that does not depend on the person whose decision is challenged. [5] IFC's handbook treats several of these — claimant-held documentation, independent verification, a functioning grievance channel — as ordinary features of a competent resettlement programme rather than as anti-fraud measures. [6]

3. Research gap

Uganda needs anonymised project-level research linking control design to detected anomalies: duplicate claimants, altered measurements, unexplained rate overrides, payment destinations, repeated grievances and procurement changes. The goal is not to publish a suspicion list; it is to learn which controls detect and prevent harm.