Power to Decide, Not Power to Rule: Prosecutorial Discretion and Constitutional Restraint in Uganda

Abstract

Article 120 of Uganda’s Constitution entrusts the Director of Public Prosecutions with wide powers because criminal justice cannot be administered by formula. That institutional freedom is indispensable, but it is not dominion. Article 120(6) protects the prosecutor from external direction; article 120(5) governs the quality of the decision made once that protection has done its work. Public interest, the administration of justice and the need to prevent abuse of legal process are therefore not ceremonial considerations. Read with articles 20, 21, 28, 42 and 50, they require a decision that can be justified by reference to evidence, lawful purpose, even-handedness and fair procedure. The 2023 Decision to Charge Guidelines are a serious institutional advance. They adopt a realistic-prospect test, separate evidential sufficiency from public interest, require written reasons and insist on continuing review. Yet the Guidelines also contain their own accountability deficit: compliance is mandatory inside the Office of the Director of Public Prosecutions, but the instrument disclaims third-party rights and withholds written reviews from those most affected by them. The result is an uncomfortable constitutional arrangement in which the state records why coercive power moved but ordinarily denies the affected person access to the explanation. This research study argues for a more exacting, but properly restrained, law of prosecutorial accountability. Bakaleeke, Rutayisire, Wanyoto and Begira Avito are not treated as passing illustrations. Each is examined for its premise, method, constitutional consequence and practical effect. Their combined lesson is that Ugandan courts recognise review in principle, but have not yet resolved how a claimant can prove irrationality, improper purpose or failure to consider relevant matter when the decisive record remains with the DPP. Comparative authority is used to address that difficulty without inviting courts to retry criminal cases before trial. The proposed model is structured discretion: clear criteria, contemporaneous reasons, independent reconsideration, controlled access to the record, remedies fitted to the defect and institutional measures that value decisional integrity rather than raw throughput. Independence survives intact. What disappears is the assumption that independence supplies its own justification.