Where a mine, a plantation and a village draw on the same watershed, the fight that follows is almost never really about geology. It is about rules — who is entitled to what, who decides, and whether the decision can be enforced. The most damaging resource conflicts trace back less to the projects themselves than to the vacuum where clear, enforceable tenure should have been.

The scholarship that grew up around the “resource curse” — the paradox that mineral wealth so often coincides with poverty, corruption and conflict — increasingly points at institutions rather than at the resources. Roughly three and a half billion people live in countries rich in oil, gas or minerals; whether that endowment becomes a school or a militia depends on how it is governed, not on the size of the deposit.

What transparency fixed, and what it did not

The flagship response has been transparency. The Extractive Industries Transparency Initiative (EITI), launched in 2002 and endorsed by the World Bank the following year, now runs in some fifty implementing countries, requiring governments and companies to disclose what is paid and received along the value chain, under the oversight of multi-stakeholder groups drawn from the state, industry and civil society. Related norms have spread with it: free, prior and informed consent for affected communities, and the disclosure of contracts and concessions.

These are real gains, and it is worth being precise about their limits, because overselling them is its own kind of failure. The evidence on whether transparency actually breaks the curse is genuinely mixed. Some studies associate EITI implementation with better short- and medium-term growth and modest improvements in accountability; others find little measurable effect on most governance indicators. The finding that recurs is a conditional one: disclosure converts into accountability only where civil society is strong enough to use it. Where it is weak, information can be published and simply ignored — evaluations in several resource-rich states describe precisely that outcome.

Transparency is a tool, not a cure. Information changes nothing where no one has the standing to act on it.

The governance gap

This is the gap that decides events on the frontier. A project arriving into clear tenure, enforceable contracts and a capable local counterpart can be negotiated, compensated and policed. The same project arriving into contested land, opaque concessions and a hollow administration becomes a conflict, almost regardless of how it behaves — and the costs fall on the smallholders and pastoralists least able to absorb them. The lesson is not that extraction is always ruin, nor always development; it is that the surrounding rules decide which one a community gets. An Institut Kandrave analysis of frontier disputes reaches a similar conclusion: much of what is blamed on a project is, on inspection, the fingerprint of the institutions that failed to govern it.

The policy implication is unglamorous but robust. Invest in the boring machinery — cadastres, courts, disclosure, genuine consent — before the trucks arrive, not after. It is far cheaper than the conflict it prevents, and it is the only version of resource development in which the people nearest the deposit are not the ones left paying for it.