A Three-Year Lease on a Forest
In 1996 Madagascar passed a law that lets a village take over the management of the forest on its doorstep — for three years first, then ten more if the state decides it has gone well. Something over a thousand of those contracts have been signed since, almost all of them out of sight, and the one everybody drives past sits thirteen kilometres south of Ambalavao.
The ticket table at Anja is a wooden desk under a roof, thirteen kilometres south of Ambalavao on the RN7, and the man behind it writes your name in a ledger. Behind him the ground rises into a field of granite blocks the size of houses, fallen off the cliffs above, with forest growing in the gaps between them. Several hundred ring-tailed lemurs live in there. They come down to the path in the middle of the morning and sit in the sun with their arms out, and the guide who walks you in grew up in the village at the gate.
There is no Madagascar National Parks sign at Anja, no state ranger, no barrier. The thirty hectares are managed by an association of villagers under a contract with the forest administration, and the money from the desk stays here. That arrangement is not an experiment or a pilot. It is a national legal instrument that has been available since 1996, and Anja is the version of it that works well enough to be visible from a car.
What the 1996 law hands over
Loi 96-025, passed on 30 September 1996, is known by the shorthand GELOSE — gestion locale sécurisée, secured local management. It allows the state to transfer the management of renewable natural resources on its land to a communauté de base: a village association, called a VOI in Malagasy, formed for the purpose. The contract has three parties — the forest administration, the commune and the association — and it is negotiated rather than granted. What moves is the right to manage. Ownership does not move at all. One correction to the shorthand, because it matters to anyone who goes looking for the paperwork: full GELOSE was slow and expensive to set up, requiring a trained neutral mediator, and in 2001 a decree created a stripped-down forest-only version, gestion contractualisée des forêts, that dropped the mediator and could be signed by the forest service directly. Most of the transfers in the country are GCF contracts rather than GELOSE ones; what follows is common to both.
The term is the part worth reading twice. A transfer runs for three years. At the end of it the administration is supposed to evaluate what the community actually did, and if the answer is satisfactory the contract is renewed for ten. That structure is deliberate: it makes the village prove itself on a short leash before it gets anything durable. It also assumes a forest service with the staff and the fuel to make the evaluation visit, and in a great many cases across Madagascar that visit did not happen, which left contracts in a state that was neither renewed nor cancelled.
The dina, and who enforces it
Rules are useless without something behind them, and what stands behind a transfer is the dina. A dina is an old Malagasy institution — a set of rules a community writes for itself and binds itself to — and under the transfer system it is written down, attached to the management plan, and taken to the local court for approval. Once the tribunal has approved it, the dina is enforceable in the state's own courts, with penalties the community sets and the community collects, usually in cash or cattle or labour. The logic is sound and cheap: a forest agent based in a regional capital cannot watch a hillside, but the people who farm below it can, and they know who went up there with an axe on Tuesday. What a dina handles far less well is the outsider. Its authority is social before it is legal, and social authority does not reach a man who has driven in from another district with a lorry and no relatives in the village. Most of the serious cutting in Madagascar has been done by exactly that man.
A dina works on the people who signed it. It is weakest against the person a forest most needs protecting from — the one who arrives from somewhere else, with a vehicle, and leaves the same day.
Why Anja worked
Anja had three advantages that most transfers do not. It is small enough to see: thirty hectares can be walked in a morning and watched from the houses. Its resource is a lemur population people will pay to look at, rather than timber somebody will pay to take away, which turns protection and income into the same activity instead of opposing ones. And it is on the RN7, the longest paved road in the country and the one every southern tour drives, so the customers were already going past the gate. Association Anja Miray was formed at the end of the 1990s, and the reserve's income has since put up school buildings, water points and a fund the association administers itself.
What Anja does not do is prove the model. Most transfer sites are inland, far larger, reached by track, and hold products with a domestic market — firewood, charcoal, construction timber, honey, medicinal plants — and no visitor to sell a ticket to. Their income has to come from managing and selling the resource itself, which is slower, needs a functioning market for what they produce, and needs a state that will act when the rules are broken by someone the dina cannot touch. Anja is the exhibit because it is legible, not because it is typical.
What the transfers were being asked to deliver
The mechanism got a much bigger job in September 2003, when President Marc Ravalomanana told the Fifth IUCN World Parks Congress in Durban that Madagascar would triple its protected land — from about 1.7 million hectares to six million. Meeting that target on the old model was impossible, because the old model was the strict national park and there was no unclaimed land left to draw one around. So the country legislated new categories of protected area that people could live and work inside, and community management became the delivery mechanism for a commitment made in a conference hall in another country. Alongside it, Madagascar National Parks earmarks half of its entry fee revenue for micro-projects in the villages around each park — a school, a dam, a health post. Real money, but it arrives as a project agreed by a committee, which is a different thing from a village taking the gate receipts.
The weakness in all of it is the same weakness, and it has been demonstrated twice. After the political crisis of 2009 the visitors stopped coming and illegal rosewood logging in the north-east ran at a scale nobody had seen; between 2020 and 2022 the gates simply closed. A community site whose whole income is a ticket desk has nothing at all in those years, and the contract it signed does not care. None of which is an argument against buying the ticket. It is an argument for understanding what the ticket is: not a donation to a conservation project, but the revenue side of a contract that comes up for review, held by the people standing in front of you.