Why an ex-ante evaluation must precede state police
Before Nigeria builds an irreversible security institution, one discipline can tell us — in advance — whether its underlying logic even holds, what it will truly cost, and whether anything cheaper would work better.
The question is no longer academic. The Presidency has disclosed that reasonable progress has been made towards establishing state police, with a constitutional amendment expected soon. The Chief of Staff to the President, Femi Gbajabiamila, says the debate has shifted from whether state police should be established to determining the most effective legal and institutional framework for its operation. The Senate has assured Nigerians it would complete the amendment before the end of 2026.
Momentum has become consensus. But the consensus skipped a stage. The debate moved from whether to how without passing through the question that belongs between them: before we commit, what does a rigorous appraisal predict this will achieve? That question is the entire purpose of ex-ante evaluation — and it is the missing first step of this reform.
–A tempting shortcut–
The frustration driving the reform is real and earned. Kidnappings, banditry, terrorism and violent crime have understandably pushed many Nigerians to demand new approaches to protecting lives and property. Recent killings in Jos, a terrorist attack in Kaduna and a spate of abductions have exposed the fragility of Nigeria’s internal security architecture from the North-West to the South-East.
When communities are burying their dead, the appeal of dramatic action is overwhelming. But emotion tells us we must act; it cannot tell us whether a given action will work.
The signs are that Nigeria is skipping straight to design: the new Inspector-General of Police, Tunji Disu, inaugurated an implementation committee within two weeks of taking office, and that committee, chaired by Professor Olu Ogunsakin, reportedly submitted its report within a short period without consulting Nigerians as widely as many expected. Note the word implementation. The committee was asked how to build state police — not first whether, on appraisal, it is the policy most likely to reduce harm per naira spent. That earlier question is the one ex-ante evaluation exists to answer.
—Three moments, one of them skipped—
Sound policy is evaluated at three moments. Ex-ante evaluation comes before commitment: it appraises a proposed policy’s likely costs, benefits, risks and feasibility, and weighs it against alternatives. A pilot tests the policy empirically in a limited setting. Ex-post evaluation judges results after full rollout.
Most of Nigeria’s debate has leapt to the third stage as an afterthought — we will see how it goes once it is nationwide. That is the most expensive and least reversible point at which to learn anything. The cheapest, earliest and most powerful filter is the one being bypassed: ex-ante appraisal, conducted before a single force is stood up or a single constitutional clause is finalised in operational terms.
—What an appraisal buys—
Its value lies precisely in when it happens — before money, political capital and institutional permanence are committed. At that moment, and only at that moment, a flaw can be corrected at the cost of a revised plan rather than a dismantled police force. Concretely, ex-ante evaluation delivers what no amount of debate can.
It forces an explicit theory of change. An appraisal would demand that proponents specify the causal chain: exactly how does creating a state police force translate into fewer kidnappings in Birnin Gwari or fewer killings in Bokkos? Through faster local response? Better grassroots intelligence? If the mechanism cannot be stated clearly and defended, that is a fatal warning sign — and far better discovered on paper than after deployment.
It establishes affordability before the bill arrives. Ex-ante cost-benefit analysis asks not only the start-up cost but the recurring one: can a state already struggling to pay salaries fund, train, equip and pension a police force indefinitely?
This is not hypothetical. By BudgIT’s latest State of States report, covering the 2024 fiscal year, 21 of Nigeria’s 36 states drew more than 70 per cent of their revenue from federal transfers, and for 31 states those transfers covered over 80 per cent of recurrent spending — even as many serviced heavy debts. Mexico’s local forces, chronically under-resourced, collapsed under exactly this pressure, leaving officers exposed — killings of law enforcement there fall overwhelmingly on local officers. An ex-ante appraisal flags an unaffordable design before it becomes an armed liability.
It compares alternatives rather than assuming the answer. Appraisal sets the proposed policy against other options for the same objective: properly funding and reforming the federal Nigeria Police Force, structured community policing, or constitutional decentralisation paired with strong federal coordination. The “how” debate quietly forecloses these by treating state police as settled. Ex-ante evaluation reopens the comparison, so the country adopts the most cost-effective route to safety, not merely the most politically visible.
It surfaces and prices the risks in advance. The known hazards — politicisation by governors, use against opponents, crime simply migrating across state lines — can be modelled before launch, with mitigations designed in rather than retrofitted after abuse occurs. President Tinubu has already urged that the amendment contain safeguards to prevent possible abuse; ex-ante risk assessment is how you discover which safeguards are needed and whether they are sufficient, instead of guessing.
And it defines success before launch, not after. Ex-ante evaluation sets the baseline crime and trust data and the metrics against which the reform will later be judged. This is what insulates the decision from politics: when success is specified in advance, the reform cannot be quietly redefined as a triumph after the fact to suit an electoral calendar.
—WHAT AN EX-ANTE APPRAISAL MUST ANSWER—
– Theory of change: How, precisely, does a state force translate into fewer kidnappings and killings?
– Lifecycle cost: Can the state fund, equip and pension the force indefinitely, not just launch it?
– Alternatives: Is reforming the federal police, or structured community policing, cheaper and more effective?
– Risk register: Politicisation, abuse, crime displacement — and a designed mitigation for each.
– Success metrics: Baseline crime and trust data, and the targets the reform will be judged against.
—Not mere guesswork—
The objection that we cannot know in advance misunderstands the method. Ex-ante evaluation forecasts by drawing on the documented experience of comparable systems — and that record is rich, precisely because decentralised policing has been tried, and has produced opposite results in different hands.
Mexico is the cautionary extreme. Its policing is exceptionally decentralised — local forces are spread across some 2,500 municipalities — yet that fragmentation left officers isolated and exposed. The eventual reversal — Mando Único, a reform InsightCrime says was designed to replace roughly 1,800 municipal units with 32 centralised state departments — cured little, because corruption proved just as endemic at state level, with centrally controlled officials still accused of collusion with organised crime; criminal groups have killed hundreds of local police, and analysts such as Daniel Sabet conclude the deeper problem is one of governance, not structure. The International Crisis Group has judged that militarised, centralised policing is poorly suited to crime that is, by nature, local.
Brazil shows the oversight hazard: under Article 144 of its 1988 constitution, each state’s civil and militarised police answer to the governor through a state Secretariat for Public Security, the militarised forces doubling as army reserves.
India has run decentralised policing for decades — under its constitution police and public order are state subjects, leaving operational command with state governments — with quality varying enormously.
Yet the model plainly can work: Australia’s six states each run their own police, as do Canada’s provinces and Germany’s Länder, within professionalised systems. Mexican survey research by Esparza and Ugues isolates the real variable — professionalisation and institutional quality, not structure itself, are what mitigate the corrosive effect of crime and corruption on public trust. Even the intuition that more local police means less crime fails the strongest test: a 2016 systematic review of 62 American studies, published in the Journal of Experimental Criminology, found police-force size had little measurable effect on crime.
Structure is not destiny; governance is.
This is exactly the material an ex-ante appraisal would weigh — to forecast, for Nigeria’s specific governance conditions, which of these outcomes the reform is likely to reproduce. And whether Nigeria has that governance is a question to appraise before committing, not discover after.
—Nigeria’s own experiment—
Nor need Nigeria look only abroad. Since 2020 the six South-West states have run Amotekun, a regional security network thousands of officers strong, while other states operate vigilante and community-policing outfits of their own. This is, in effect, a sub-national policing experiment already under way — and a revealing one.
Residents and officials credit Amotekun with blunting rural banditry, highway kidnapping and farmer-herder clashes in places; yet a 2025 evaluation in the East African Journal of Law, Policy and Globalization, while calling it a milestone in decentralised security, found its reach hobbled by legal grey areas, thin and uneven funding, and friction with federal agencies. That is precisely the mixed, contingent picture the international record predicts.
It is also the most relevant evidence Nigeria possesses — yet no government has commissioned the kind of systematic, published evaluation that would establish which design choices worked, at what cost, and why. Before scaling to 36 full police forces, that existing experience is the first thing an appraisal should mine. To legislate past it is to discard the one natural experiment Nigeria has already paid for.
—From paper to pilot—
Ex-ante evaluation does not replace a pilot — it justifies and shapes one. Appraisal forecasts likely outcomes and screens out unworkable designs on paper; a pilot then tests empirically what appraisal can only predict.
Nigeria need not stand up 36 forces at once. A favourable appraisal would point to a handful of representative states operating state police under defined conditions, measured against matched states that have not yet adopted it — so the country can see whether crime actually falls, whether forces stay accountable under gubernatorial pressure, and whether states can sustain the cost, before anything becomes permanent.
—Evidence is action—
The case for moving now is not frivolous, and it deserves to be put at its strongest. The federal police have manifestly failed to secure large swaths of the country; insecurity is killing Nigerians today; and there is broad public and political demand to act, voiced by figures such as Oyo State Governor Seyi Makinde in his call for expedited reform. Waiting is itself a decision, and its cost is measured in lives.
But ex-ante evaluation is fast where it counts. A rigorous appraisal can be completed in weeks to a few months — trivial against a permanent institution — and it can run in parallel with the constitutional drafting already under way. The choice is not between acting and waiting; it is between acting blindly and acting wisely. The effectiveness of decentralised policing depends not merely on structure but on the quality of governance that supports it. Even the Presidency concedes the reform cannot be implemented hastily. Ex-ante evaluation is what “not hastily” looks like in practice.
—A better sequence–
Let the constitutional amendment proceed. But before the enabling law authorises a single force, commission an independent ex-ante evaluation: a published appraisal of the policy’s theory of change, full lifecycle costs, fiscal sustainability for participating states, comparison against alternative security reforms, and a risk assessment with designed mitigations.
Let it set baselines and define success in advance. Let its findings shape a phased, evaluated pilot rather than a simultaneous national launch. And let the country commit, publicly and beforehand, to following the evidence rather than the mood.
The world has run this experiment many times, with results ranging from Australian competence to Mexican catastrophe. The difference lay in the details — and ex-ante evaluation is the discipline that surfaces those details while they can still be acted on cheaply. Nigerians demanding safer communities deserve a solution appraised and shown likely to deliver safety, not one that merely feels like decisive government.
Before we institutionalise state police across the federation, let us first prove, on the evidence and in advance, that it is worth institutionalising at all.
■ Dr Lawal Zakari, mni
Former Director, National Monitoring & Evaluation Department (2011–2023); Chief Executive Officer, Tazaar Management Consultants Ltd.
