Why Most SEZs Fail Before They Launch

The decision that determines the outcome is institutional, and it is made before the first tenant ever signs.

By Yaschin Mohabir · 6 July 2026 · 10 min read

A zone cannot move itself closer to a market or conjure one into being. What it can choose is how it is governed, and that is the choice that decides whether it works. Infrastructure and incentives are what zones compete on in public. The governance settlement fixed at inception, the one decision nobody goes back to question, is what separates the zones that compound from the ones that stall.

Lekki had everything on the checklist except an answer to one question

Nigeria's Lekki Free Zone had everything a zone is supposed to need. Chinese capital. A deepwater port on its doorstep. A government in Lagos determined to prove it could compete with Tema and Mombasa. Tax holidays generous enough to draw serious industrial tenants. It underperformed for years after its 2006 launch anyway, caught in a slow dispute between the Lagos State government, the federal government, and its Chinese developers over who actually controlled what.8 The infrastructure was real. The incentives were real. The governance was a fiction, agreed to by parties who had never resolved their underlying disagreement about authority. That is not the exception. It is the template. Across three decades of zone programs, the recurring cause of failure is a governance settlement written to satisfy a political negotiation and locked in before anyone has answered the prior question: what does this zone do for a specific tenant, and what does delivering that require from the institution running it? The map and the incentive package are rarely the thing that breaks. None of this is exotic knowledge. Every experienced hand in that dispute could have drawn the org chart that would h...

The fatal choice is made first and revisited never

They get the order wrong because the politics runs ahead of the analysis. The pressure to announce a zone, to sign the concession and break ground, arrives long before anyone has done the work of defining what the zone is for. So the governance model gets decided early, around the convenience of existing bureaucratic boundaries, and then calcifies around the politics of that moment. By the time the value case finally arrives, it has to be delivered through an institution that was never designed to deliver it. The first customs hold or stalled permit is the original sequencing error surfacing on schedule, not a teething problem. By then you are no longer designing the zone. You are litigating it. Invert the order and the failure rate moves. Define the industrial logic first: what value, for which tenant, requiring what from the institution. Then specify the institutional architecture that delivers it. Then, and only then, negotiate the governance structure around that requirement. The sequence is easy to state and hard to hold, because the pressure to announce a zone always arrives before the analytical work that would make it succeed. The empirical record supports the priority, w...

The zones that compounded share one trait, and it is not their tax rate

If incentives are not the differentiator, the successful zones should show what is. They do, and you can watch it operate. Gabon's Special Economic Zone, developed by ARISE, did not build sheds and offer a tax holiday. It stood up an integrated authority empowered to issue business licenses, clear customs, supply utilities, and enforce environmental compliance, all inside the perimeter and all accountable to one operational body rather than spread across ministries. When a Belgian wood processor moved veneer production into the zone, it cleared equipment through customs in days, against the months that Gabonese import procedures outside the zone routinely demanded.6 The structure made the proposition credible. Without it, an efficient processing environment would have been a brochure, not a fact. Shannon makes the same case from a different century. When it launched in 1959 as the world's first purpose-built free trade zone7, the Shannon Free Airport Development Company received a unified mandate over industrial promotion, infrastructure, and community development. It was not asked to coordinate with seven agencies; it was told to deliver outcomes and given the authority to do it....

The global minimum tax just repriced the entire tax-holiday playbook

If the case so far makes tax the weaker lever, the global minimum tax has just taken the strongest version of it off the table. For zones that still lead with a tax holiday, the ground has moved. Under the OECD's Pillar Two regime, in force across more than fifty jurisdictions and projected to lift global corporate tax revenue by 155 to 192 billion dollars a year,3 large multinational groups must now pay an effective minimum of fifteen percent on their profits in every country where they book them, whatever local holidays promise. Cut a tenant's rate below that floor with a holiday, and another government simply collects the difference as a top-up tax. The break never reaches the company. It is quietly transferred to the multinational's home treasury. The picture is not absolute, and the reason is specific. In January 2026 the Inclusive Framework agreed a side-by-side package that includes a substance-based safe harbor, giving more room to incentives tied to real activity, payroll, and physical assets on the ground.4 So a zone can still compete on substance. What it can no longer do is compete on a naked rate. The lever that survives Pillar Two is the one this paper has been descr...

Four parties are mispricing the same risk

The governance gap is not an abstraction. It lands as a specific, avoidable cost on four groups at the table, each of whom tends to discover it too late. For developers, governance designed to satisfy a negotiation rather than an operating requirement is a liability that surfaces in the first hard conversation with a real tenant. For investors, the due diligence question is whether the authority can actually deliver the operating environment it advertises, regardless of the headline tax rate. For policymakers, the standard sequence has to be inverted, because incentives carved before the institution is designed buy visibility today and underperformance for a decade. And for the operators who run these zones day to day, the uncomfortable truth is that they usually inherit the governance settlement rather than design it. Handed a fragmented mandate, the most an operator can do is document the friction, price it, and escalate it, which is a thin substitute for holding the authority in the first place.

Five questions that reveal whether a zone is real

I would make a governance audit a precondition for any zone receiving public support or a development license. Concretely, that means five questions answered before a concession is signed. Each one exposes whether a governance model is built around the tenant's operating reality or the developer's financing structure. None of them is academic. None of this requires a government to hand a private operator sovereign power. The workable line runs through three tiers. Operational authority, customs clearance, licensing turnaround, utility connection, can be delegated to the zone authority by statute and bounded by service-level terms. Quasi-judicial authority, the resolution of tenant disputes, needs explicit legal backing, which is why the durable zones stand up a dedicated court or arbitration centre rather than relying on an administrative promise. Core sovereign functions, taxation base, criminal law, national security, stay with the state. Durability across a change of government comes from putting the first two tiers in primary legislation, not in a concession a later cabinet can quietly rewrite. Audits of exactly this kind are now starting in the Gulf, and the early findings p...

A zone is an answer to one question about authority

Strip away the masterplan, the incentive schedule, and the groundbreaking ceremony, and a special economic zone reduces to a single claim: that inside this line, the institution in charge can deliver an operating environment the surrounding state cannot. Everything else is in service of that claim, and nothing else can rescue it if the claim is false. The infrastructure can be world-class and the holiday generous, and the zone will still stall if no one inside the perimeter can resolve a customs dispute by Friday. So the test is simple, and most zone documents fail it. If a foreign investor in your zone files a dispute against a government agency tomorrow, which institution resolves it, in how many days, and under which law? If you cannot answer without checking three documents and making two phone calls, you have not built a zone. You have built a geography with a logo.