A community rarely experiences one project. It experiences a river that three operators draw from, and a road that two mines and a quarry share. It experiences dust from a pit, a tailings beach and a haul road that belong to different companies. And it lives in a district whose population has doubled because of all of them at once. When the grievance arrives, it arrives at the nearest gate, and the company behind that gate answers, correctly, that most of the problem is not its doing. The community hears that nobody is responsible. It is right about that too. Responsibility divided among several parties, each with a good legal reason to decline it, is responsibility that nobody carries.
Cumulative impact is the condition in which each project is individually compliant and the community is collectively harmed. The IFC’s handbook on the subject, published in 2013, set out a six-step process for assessing impacts on valued components in combination with other projects and activities. Performance Standard 1 requires assessments to consider them. In practice cumulative assessment is done once, at permitting, by one company, for its own footprint, and then the region fills up. This article is about the grievances that cumulative impact produces, why the single-company response fails, and how a regional process can be built when no single company owns the problem.
Compliant projects, harmed community
Consider a scenario drawn from patterns across many mining regions. A valley has one river and, within fifteen years, three mines, each permitted separately. Each has a water licence within the river’s assessed capacity at the time it applied. Each monitors its own abstraction and discharge and reports compliance. Downstream, the villages that irrigated from the river in the dry season no longer can. The fish have gone, and the water that reaches them carries three companies’ permitted discharges combined. Every company’s monitoring report is clean. The village’s fields are dry.
The village’s grievance mechanism options are equally fragmented. It can complain to the nearest gate, which handles a third of the problem. It can complain to the regulator, which licensed all three. Or it can block the road that all three share, which is the only action that reaches everyone at once. That is why cumulative grievances so often skip the register and go straight to the road.
The same arithmetic runs through dust, noise, traffic, in-migration and pressure on services. A haul road that carries one mine’s trucks at a tolerable frequency carries three mines’ trucks at an intolerable one. A district clinic sized for one construction workforce is overwhelmed by two. A grievance mechanism at each gate registers complaints about that gate’s share, and the shares never add up to the community’s experience. The conflict data from Peru shows the pattern at national scale. Mining accounts for most of the country’s socio-environmental conflicts, and many of the hardest sit in corridors of multiple concessions.
Why the single-company answer fails
Each company’s instinct is to define its responsibility by its fence and its permit, and the instinct is legally sound and socially useless. Three responses recur, and each makes the grievance worse. The first is attribution: our dust monitors show our contribution is small. The community does not have three sets of lungs. The second is deflection: the traffic is mostly the quarry’s. The community did not license the quarry and cannot reach it. The third is the offer of a share: we will fund a third of the clinic upgrade if the others fund the rest. The others do not, and the clinic is not upgraded, and the offer becomes another unkept promise.
The failure is structural. The community’s grievance is regional, and the only actor with a regional mandate, the government, is often absent or has licensed the projects in the first place. Each company’s assessment, done at its own permitting, described a valley that no longer exists. And no company has an incentive to convene the others, because convening means accepting a share of a problem it could otherwise disown. The result is a grievance with no address, which is the kind that ends up in the road.
There is a further problem with sequence. The first project in a valley is assessed against an empty baseline and finds little to worry about. The third is assessed against a valley already under strain and is asked to carry the accumulated burden alone. Neither outcome is fair, and both are avoided only if the assessment is done for the valley rather than for the applicant.
Building the regional process
Cumulative impacts need a cumulative process, and someone has to start it. The IFC handbook’s steps give the technical shape. Scope the valued components with the affected people, establish the baseline including other projects, assess the combined effect, design management measures, and monitor jointly. The social shape is harder, because it requires the companies, the government and the communities to sit at one table over a problem none of them owns alone.
The starting point is usually one company, the largest or the most exposed, accepting that convening is cheaper than being blamed. It invites the other operators, the relevant authorities and the affected communities to a scoping conversation about the valley’s shared resources: the river, the road, the airshed, the town. An independent facilitator is close to essential, because no company will accept another company’s chairing and the communities will not accept any company’s. The first product is a shared baseline that everyone agrees describes the valley as it is. That is itself a major achievement, since each party has until then been working from its own. The second is a set of valued components the communities named, water for irrigation, safe roads for children, the clinic, that the process commits to protect. The third is a joint monitoring arrangement, run on the principles I set out in Participatory Environmental Monitoring That Communities Trust. The valley’s condition is then measured once and believed by all.
Sharing the response
The hardest negotiation in a regional process is who pays for what. Companies want contributions proportional to attribution, which invites endless argument about shares. Communities want the problem fixed and do not care whose trucks did it. A workable rule is to separate the two conversations. First, agree what needs to be done to protect the valued components. An abstraction cap on the river in the dry season. A shared haul road with a single traffic management plan. A jointly funded clinic upgrade. A regional influx plan with the municipality. Second, agree how to share the cost, using a simple and transparent formula, by production, by workforce, by licensed abstraction, rather than by contested attribution. A formula that is roughly fair and quickly agreed beats one that is precisely fair and never settled. In my experience the argument over attribution is where regional processes die, and the parties who most want precision are usually the ones who least want to pay.
The government’s role has to be negotiated too. Where the state has capacity, the regional process can feed a formal instrument: a basin plan, a regional development framework, a cumulative assessment required of the next applicant. Where it does not, the companies and communities can still build the arrangement between them, with the state inside it in whatever capacity it has. The structure can then be handed to public institutions when they are ready. The mapping of the state’s several roles that I describe in the coming cluster on government at the table starts from exactly this problem.
The grievance mechanism for a valley
Individual gate-level mechanisms cannot handle cumulative grievances, and communities know it, which is why cumulative complaints so rarely appear in any single register. A regional process needs a shared intake: a single point where a household can raise a complaint about the river or the road without first diagnosing which company caused it. The complaint is logged once, assessed jointly, and assigned to the responsible operator or to the joint arrangement where responsibility is shared. The complainant deals with one process rather than three, and the companies gain a view of the valley’s grievances that no single register provides. The intake can be hosted by the joint council, by the municipality or by an independent office; what matters is that it belongs to none of the operators. The multi-party mediation skills this requires are the ones I set out in Mediating Multi-Stakeholder Conflicts in Mining. A valley with three companies and a dozen communities is the standard case.
Map the valley before the valley maps you
The Cumulative Impact Response Roadmap sets out the build of a regional process in five phases and 15 steps. The phases run from recognising the condition and deciding to convene, through scoping with communities and a joint baseline, to the response, the cost-sharing rule and the shared governance arrangement. Score each step In place, Partial or Absent, and read the result phase by phase. Absents in the first phase mean each company is still defending its fence while the valley’s grievance grows. Absents in the fourth mean the process will produce a shared diagnosis and no shared action, which is worse than none. Download the Cumulative Impact Response Roadmap
Nobody’s problem is everybody’s risk
A cumulative grievance is the most dangerous kind, because it has no owner and therefore no exit. Each company can prove it is compliant and none can make the river run. The community’s livelihoods are lost to the sum of the projects and each project’s investment is exposed to the community’s response to the sum. The only protection for either is a process at the scale of the harm.
The Social Accord Architecture is designed to scale to that level. Its Diagnostic Mapping maps the whole valley’s actors and shared resources, not one project’s footprint. Its Blueprinting produces an accord in which several companies, the authorities and the communities hold commitments over the same river and road. They share a joint council, a single commitments register and one Feedback Loop. Its mediation pathway is where the disputes among the companies, as much as between the companies and the communities, are worked through. Dual accountability holds every operator to its share and the communities to using the regional process rather than blocking the nearest gate. Where a valley has already reached the point of a grievance with no address, the first meeting usually needs an independent mediator. Someone who can convene the operators without any of them appearing to lead.
One thing to do this quarter. List every project, road and abstraction within the watershed your community shares, and ask whether any of them has ever been in the same room. If the answer is no, the valley’s grievance is being written now, one compliant report at a time. To design a regional cumulative impact process or to convene the operators and communities of a shared valley, write to me at thomas@thomasgaultier.com.