Nigeria’s justice sector has not suffered from neglect by international partners. Since the 1999 return to civilian rule it has been one of the most intensively aided domains in African governance reform. British programmes, World Bank financing, EU multi-component consortia, UNODC operational placements, judicial-college partnerships, Commonwealth networks and Eurojust working arrangements have successively layered money, expertise, curricula, coordination structures and theories of change onto courts, police, prosecution, legal aid and traditional authorities. The cumulative volume is su. The British Council’s combined justice and education portfolio alone claimed to reach more than thirty million people; Justice for All brandished a price tag of roughly £47.8 million; EU-SDGN II is a €39 million, five-year, sixteen-organisation effort. Yet the central empirical and concurrent pattern that emerges from a quarter-century of documentation just falls short of cumulative institutional capability. It is a repeating cycle of high-activity pilots, impressive output tallies, and comparatively thin residue of durable, self-reproducing practice once external scaffolding is removed.
The pattern is not in and of itself indicative of uniform failure. Certain small, concrete, locally authored interventions have survived and spread whilst Larger, more architecturally ambitious consortia have generated volumes of activity whose connection to lasting behavioural or organisational change remains weaker. Scale has therefore functioned less as a reliable multiplier of impact and more as a signal of donor confidence that can, paradoxically, dilute the very local authorship, trialability and visible early results that diffusion theory identifies as the mechanisms of survival. The forensic task is therefore to trace each major trajectory: its design logic, measurable outputs, institutional fate, and contribution or subtraction from system-level progress, before asking what native or hybrid infrastructure would convert the accumulated experience into a self-reinforcing capacity to learn.
I. The 1999 Inflection and the Invention of “Justice as System”
Democratic transition reopened the courts and simultaneously redefined justice as an ecosystem threaded through security, anti-corruption, public-sector reform and economic growth. The first major expression of this logic was Britain’s Security, Justice and Growth (SJG) programme (roughly 2002–2010, budget in the order of £30 million). It treated commercial dispute resolution, land administration and contract enforcement as development issues in their own right, worked simultaneously on federal institutions, awaiting-trial populations and legislative frameworks, and attempted inter-agency coordination rather than single-institution strengthening.
Merits. SJG established the conceptual shift from “strengthen institution X” to “alter relationships among institutions.” Contemporary UK parliamentary evidence recorded contributions to improved contract enforcement and property registration that fed into Nigeria’s 2006 Doing Business rankings. It normalised the idea that justice outcomes are produced by chains (police, prosecution, courts, corrections and citizens) rather than by courts alone.
Demerits and consequential outcomes. The programme’s own documentation and subsequent evaluations leave limited public trace of which specific process changes or coordination structures survived funding closure. Institutional memory of SJG resides largely in donor archives rather than in a living Nigerian repository. The inter-agency architecture it promoted did not automatically generate continuing joint diagnostics or shared performance data. Contribution to progress was therefore foundational but non-cumulative: in effect it opened the systems lens without infusing the integral feedback mechanisms that would have allowed later programmes to interact and build directly on measured results rather than obstinately re-diagnosing identical problem vectors.
II. Justice for All: Coordination at Scale and the First Clear Diffusion Cases
J4A (UK-funded, British Council-delivered, approximately 2010–2017, £47.8 million) intensified the coordination logic. It created justice-sector reform/coordination structures at federal and state levels, worked on case management, lower courts, customary and Sharia courts, mediation and access to rights, and involved the National Judicial Institute in train-the-trainer activity. British Council accounts placed the combined justice–education portfolio at a reach exceeding thirty million people.
Merits and forensic traces of survival. Two cases stand out as textbook diffusion. In Jigawa, traditional rulers received training in human rights, alternative dispute resolution and gender issues. The state’s own Justice Sector and Law Reform Commission then appropriated the materials, adapted delivery, and trained several hundred additional rulers at its own initiative and expense. In Lagos, court-connected citizen mediation centres outlasted the originating project and settled into ordinary state practice. Both interventions sat high on the participation gradient: they worked with existing local authority figures and community dispute-resolution habits rather than displacing them. Compatibility, trialability and visible early results were present; foreign authorship was low; local institutional appetite was high. These are the clearest documented instances in the record of reinvention rather than mere replication.
Demerits and limits. The same programme generated large volumes of training, coordination meetings and materials whose post-closure fate is unevenly documented. Coordination structures did not automatically produce continuing shared diagnostics across police, prosecution, courts and legal aid. The programme’s scale purchased reach and visibility without necessarily centering the sustained, specific attention of the people inside receiving institutions who ultimately decide what practice is worth adoption into custom after funding is suspended. Contribution to progress is therefore dual: it produced genuine, durable local innovations and simultaneously illustrated how large multi-component activity can leave thin institutional residue.
III. World Bank Development Finance: Physical Embedding versus Temporary Cadres
The World Bank’s Access to Justice for the Poor project in Kaduna (approximately 2012–2015) operated on a different theory: financing, workflow re-engineering and institution-building rather than foreign-managed technical assistance. It established roughly fifteen legal-aid centres, trained staff and paralegals drawn from the communities served, and used outreach for legal literacy. Programme accounts record direct service to several hundred low-income clients.
Merits. Physical embedding and local staffing created an infrastructure with a plausible reason to survive funder exit. Paralegals living where the centres operated reduced the classic temporary-cadre problem. The intervention by implication was rather more collaborative than consultative. Relative to the multi-organisation consortia model, its modest scale was an advantage: it did not generate the coordination overhead that later programmes would face.
Demerits and outcomes. Public documentation of post-2015 continuity, expansion beyond Kaduna, or systematic comparison with other legal-aid models remains limited. The project did not itself create a national mechanism for transporting the model. Contribution to progress is therefore real but geographically bounded: illustrating that financing infrastructure can produce more durable local capacity than pure training cycles, yet not solve the diffusion problem that recurs across the larger history.
IV. The Consortium Era: EU-SDGN and the Coordination Costs of Ecosystem Ambition
EU Support to Democratic Governance in Nigeria (EU-SDGN), now in its second phase (€39 million, five years, sixteen implementing organisations, six components spanning electoral administration, National Assembly and judiciary, parties, media, marginalised groups and civil society), represents the fullest expression of ecosystem governance. The judicial component sits inside a wider democratic architecture on the premise that electoral justice cannot be understood in isolation from the electoral commission, legislature, parties and media.
Merits. The analytical theory is sophisticated and almost certainly correct: justice outcomes are relational. Later programme work produced a detailed study of 2023 election-petition litigation that captured judicial experience from tribunals through the Supreme Court, analysed patterns, and fed findings back into public and reform conversation. This is closer to institutional memory than most training interventions are offering/have offered. The programme’s own steering processes have begun to deliberate on strategic alignment across partners.
Demerits and the broken middle link. An independent mid-term assessment drawing on several hundred stakeholder consultations and field visits found outputs delivered at volume while the connection to durable behavioural or institutional change remained thinner than activity levels suggested. Insufficient attention was paid to the incentives and practical opportunities that determine whether actors reenact the changes learnt when the workshop ends. Coordination costs, reporting overhead and diffusion of accountability across sixteen organisations make it difficult, once funding ends, to know which parallel streams actually altered practice and which merely occurred without effect. Contribution to progress is therefore mixed: the programme has generated valuable evidence products and kept the ecosystem lens alive, yet it exemplifies how architectural ambition can outrun the local authorship and feedback loops required for survival.
V. Operational Specificity: RoLAC and Low-Complexity Behavioural Change
The EU Rule of Law and Anti-Corruption programme (Phase I 2017–2023, British Council delivery, federal level plus five focal states) moved closer to daily practice: sentencing guidelines, plea bargaining, Administration of Criminal Justice Law implementation, and reforms to how police take statements. In Edo State a facility was established to support electronic recording of suspect statements: a narrow, physical intervention aimed at a precise point of divergence between law and practice. Programme documentation frames improved compliance in behavioural rather than just procedural terms.
Merits. The intervention possessed the attributes which diffusion theory foretells will favour adoption: it asked officers to perform already legally required tasks (low perceived novelty and risk), required no elaborate new skill (only a room and a recorder), and made compliance or its absence immediately visible to supervisors. Low complexity and high observability are harnessed as advantages that large multi-component programmes often sacrifice.
Demerits and open questions. Whether Edo’s model reached other states, and whether it survived programme closure, is not yet robustly documented in the public record. A strong candidate for adoption becomes evidence of diffusion only once tracked past the closing date. Contribution is therefore promising but incomplete: it shows that reform at the level of daily operational practice is capable of changing behaviour when attributes favour adoption, but it still required subsequent and substantial local appetite and financing to travel.
VI. Transnational and Professional Layers: Relationships over Manuals
Two further strands sit outside the classic donor-project model. The 2023 Nigeria–Eurojust Working Arrangement (Nigeria, the first sub-Saharan African country to sign) enables a formal contact point, faster execution of cooperation requests and strategic information exchange on organised crime.
The earlier UNODC/Siracusa Liaison Magistrate Initiative placed Nigerian prosecutors in Italy and Spain from 2018 to improve real-time communication on trafficking and migrant-smuggling cases. Separately, the A4ID ROLE UK partnership between the Judicial College of England and Wales and Nigerian state ministries of justice trained the great majority of Niger State’s eighty-two magistrates over five days, built train-the-trainers from the outset, and adapted into two-way webinars when COVID interrupted physical exchange. Following one exchange, Kaduna courts were directed to resume sitting.
Merits. These interventions treat knowledge as embedded in people and relationships rather than documents. A liaison prosecutor carries procedural familiarity, trust and an accumulated sense of how another system behaves, none of which could be sufficiently catered by merely being written down. Train-the-trainers designs institutionalisation as a day-one objective rather than a sustainability paragraph. Adaptive partnership (absorbing a shock and continuing) demonstrates the standard a genuine learning relationship requires. Commonwealth Latimer House Principles (endorsed in Abuja 2003) and participation in the International Association of Judges further locate Nigeria inside older professional epistemic networks that predate project cycles.
Demerits. Conventional programme archiving is poorly equipped to capture relational knowledge. Continuity depends on individuals and personal networks that can dissipate with personnel change. Contribution is high-value but fragile unless deliberately converted into institutional memory.
VII. Comparative Synthesis: What the Record Reflects
When the cases are laid end to end, five models of reform become visible:
1. Institution strengthening (early technical assistance).
2. Sector coordination (SJG → J4A).
3. Ecosystem governance (EU-SDGN consortia).
4. Networked expertise and operational placement (ROLE UK, liaison prosecutors, Eurojust).
5. The still-missing institutional-learning function.
Durable outcomes cluster disproportionately among the smallest and most concretely embedded interventions: Jigawa’s self-funded continuation of traditional-ruler training, Lagos mediation centres, Kaduna paralegal centres, Edo’s statement-taking facility, Niger State’s train-the-trainers design. Each possessed compatibility with existing values, trialability, visible results and local authorship. Large multi-component programmes generated higher activity volumes and more confident systemic theories, yet repeatedly displayed a gap between outputs and lasting change, higher coordination costs, and thinner post-funding residue.
The inhibition taxonomy that emerges is precise and is neither generic nor incidental:
• Knowledge failure (innovation never reaches relevant actors).
• Translation failure (knowledge exists but is not converted into daily routine).
• Motivation/incentive failure (awareness without reason to change, explicitly flagged by EU-SDGN evaluation).
• Coordination failure (police–prosecution–courts–legal aid remains siloed).
• Capability failure (personnel, equipment or organisational capacity absent).
• Financing failure (pilot money easy; institutionalisation and scaling money scarce).
• Institutional-memory failure (knowledge lives in individuals or temporary structures).
• Diffusion failure (successful local models remain isolated).
• Evaluation failure (outputs measured; behavioural and institutional outcomes under-measured).
None of these aforementioned are exotic. What can be deduced to be missing is in fact the institutional habit of recording these interventions in-depthly and case by case such that it all accrues into a vast breath of knowledge base and scaffold upon which whatever next programme that emerges can be suspended upon to continue to build from and on. Instead the system fails to accumulate insight, choosing instead to relearn similar lessons on a fixed cycle, a process that institutionalizes redundancy and enforces a state of epistemic squalor.
VIII. The Scale Illusion and the Real Deficit
Scale has been treated as the reward for a good theory of change. Albeit history suggests the opposite relationship often holds in practice.
The mechanisms that actually produce diffusion (a state commission’s own appetite to continue, a facility staffed by local personnel, a design that assumes the foreign partner’s eventual departure) are grassroots by nature. They are not reliably manufactured by adding organisations, components or funding lines. Scale and grassroots institutionalisation may compete for the same scarce resource: the sustained attention of the people inside the receiving institution who must decide, after funding ends, whether the system is worth keeping.
The real deficit is therefore narrower than a general shortage of capacity or political will. Nigeria possesses an unusually rich stock of reform experience spanning integrated sector reform, development finance, ecosystem governance, criminal-justice behaviour change, transnational operational cooperation and professional-network participation. What it lacks is any single, living mechanism that allows anyone of note inside the justice sector to answer, reliably: which reforms have been tried since 1999, which worked under what conditions, why the failures broke down, which were reproduced, and which Nigerian innovations have travelled outward. That absence is the medium which cultivates the parasite that is responsible for the consistency in diagnosis recurring across successive programmes whilst also offering it subterfuge.
IX. Toward a Robust Native or Hybrid Learning Architecture
The remedy is not another large consortium or a new standalone bureaucracy that risks becoming a further silo. Diagnosis must precede design. There are fragments already in existence: the National Judicial Institute’s professional-development mandate and partnership infrastructure; National Judicial Council policy language on collaborative improvement and justice-sector coordination structures; state-level law-reform commissions that have, in isolated cases, demonstrated ownership; professional associations and Commonwealth/IAJ networks. The gap is connective tissue and operationalisation as a continuous learning system.
A hybrid architecture that synthesises the historical evidence rather than importing another ideal would therefore prioritise eight interlocking functions, distributed rather than concentrated:
1. System intelligence: continuous mapping of who is working on what, where, with whom, under what model, with what evidence.
2. Institutional memory: structured, searchable repository of programmes, evaluations, manuals, failures, costs and implementation experience (a National Justice Innovation Register).
3. Experimentation: safe, small-scale testing of alternatives before national recommendation.
4. Independent evaluation focused on behavioural and institutional outcomes, not merely outputs.
5. Translation: conversion of evidence into judicial guidance, operational protocols, training materials and practice directions.
6. Diffusion mechanisms: peer learning, state-to-state exchange, train-the-trainers, and deliberate replication support.
7. Financing for the pilot-to-institutionalisation transition: the point at which most successful experiments currently die.
8. Feedback loops that return implementation experience from judges, administrators, police, prosecutors, litigants and civil society into the next design cycle.
Organisationally this need not be a new central body. A networked model (small independent research and intermediary capacity formally linked to NJI/NJC, selected state institutions, universities and professional bodies) could perform the connective and documentation functions while ownership of experimentation and institutionalisation remains with the justice institutions themselves. Documentation becomes infrastructure rather than paperwork: every significant intervention captured against a standard matrix of problem, theory, context, actors, adaptation, resistance, behavioural change, outcome, cost, dependencies, replicability and current status. A deliberate “What Didn’t Scale - and Why” archive would make failure diagnostically useful rather than something to be buried.
The sequence must remain empirical: first complete the partnership-by-partnership and intervention-genealogy matrices for 1999–2026; second, map existing Nigerian institutions against the eight functions to locate precise gaps, duplications and blockages; only then design the minimal connective tissue required. Scaling successful practices becomes a later consequence of a system that can distinguish what is worth scaling from what should be abandoned.
The developmental transition the history points toward is therefore not more programmes or larger consortia. It is the conversion of a rich but fragmented history of projects into a national capacity to learn from them: specifically, locally, and by institutions that have decided, for their own reasons, that the knowledge is worth keeping after the light from donor outposts go out. That is the infrastructure scale has so far failed to buy, and the one the trail of records now demands.
The core problem is the lack of an efficient knowledge-management system geared towards optimisation and improvement, a function one would imagine belongs to a properly functioning institution. But alas, our institutions are all ailing. Whose responsibility is it, then, to fix the body responsible for fixing? And whose responsibility would it be should that function choose to be outsourced?
Whose responsibility is it to wrangle institutional memory into accord?
If you have patiently read this to the end and want to hear more discussions on topics like this, register today for the A2J parley hosted by PPDC.