Thomas Gaultier
Menu

Social Impact Assessment Done Right

Thomas Gaultier 9 min read

Most social impact assessments fail the moment they are approved. The document goes to the regulator, the permit clears, and the binder goes on a shelf. Two years later the same project is negotiating with the same community over an impact the SIA named on page 140, because nobody built a way to track it, act on it, or revisit it. The assessment was treated as a hurdle. It was never treated as a management tool.

That is the gap this article is about. A social impact assessment done right is not a report. It is the start of a process that runs the length of the project, protects the livelihoods of the people nearby, and protects the investment of the people who financed it. Both, at once. When the SIA is written to be revisited, it becomes the shared reference that keeps a company and a community working from the same facts. When it is written to be filed, it becomes evidence, later, of what everyone knew and nobody managed.

The report is the smallest part of the job

The International Association for Impact Assessment sets out social impact assessment as four phases, not one deliverable. The 2015 guidance, led by Frank Vanclay with Ana Maria Esteves, Ilse Aucamp, and Daniel Franks, describes them plainly. Understand the issues. Predict and analyse the likely impacts. Develop and implement strategies to address them. Design and implement monitoring. Three of those four phases happen after the assessment report is written. The report sits at the end of phase two. If your SIA work stops when the document is signed, you have done half of one phase out of four.

This matters because the impacts that damage projects are rarely the ones nobody saw. They are the ones that were named and then left unmanaged. A resettlement schedule slips. An in-migration wave arrives faster than the water supply can carry it. A compensation process that looked fair on paper starts producing grievances because the local price of land moved. None of that is a surprise if you read the assessment. All of it becomes a conflict if the assessment has no life after approval.

Treat the report as the reference point, not the finish line. The value is in what you do with it over the next decade.

Impacts are things people experience, not things you count

Frank Vanclay drew a distinction in 2002 that most assessments still blur. A social change process, an increase in population, an influx of temporary workers, a shift in land use, is not in itself a social impact. It becomes an impact only when someone experiences it: as crowding, as a loss of safety, as a rise in the cost of housing they can no longer afford. Counting the change is easy. Understanding the experience is the actual work.

The practical failure follows directly. Many assessments measure what is convenient, population figures, employment numbers, kilometres of road, and call it a social baseline. Those are indicators of change. They are not the impact. A project can generate 400 jobs and still fracture a community if the jobs go to outsiders while local expectations were set higher. The number looks like success. The experience is betrayal.

An SIA done right keeps its eye on lived experience. It asks what will change, then it asks who will feel that change, how they will feel it, and whether they read it as harm or gain. That is why the assessment cannot be built from a desk. It has to be built from the community, through structured conversation with the people whose lives the project will alter. The root causes of most mining community conflicts trace back to impacts that were felt long before they were formally recorded, because the recording measured the wrong thing.

Predict, plan, monitor, adapt

The through-line of a working SIA is a loop, not a line. You predict what the project will do to people. You plan how to avoid, reduce, or offset the harm and how to build on the benefit. You monitor what actually happens against what you predicted. Then you adapt, because the prediction will be wrong somewhere, and the point of monitoring is to catch that early enough to act.

IFC Performance Standard 1 builds exactly this into its requirement for an Environmental and Social Management System. The standard is explicit that risk management is not a one-time study. It is a system that runs across the project, adjusts the management program as issues surface, and scales monitoring to the level of risk. The assessment feeds the system. The system is what actually protects anyone.

Consider a scenario drawn from patterns common to health programs on large sites, of the kind documented in the impact-assessment literature. A project identifies elevated disease transmission as a risk to workers and to surrounding villages during the assessment phase. Rather than noting it and moving on, the team sets a baseline, agrees targets with the local health authority, and tracks case rates month by month. When cases climb despite the first round of measures, the data triggers a change in approach, not a year-end excuse. That is monitoring built to drive decisions. It is the difference between a plan that manages a real problem and a plan that documents one.

The loop is also what lets a company show its financiers and its regulators that impacts are under control, with evidence rather than assertion. That evidence is an asset. It is what stands between a project and the true cost of a conflict that could have been caught early.

Build the monitoring plan into the assessment, not after it

The single most reliable marker of an SIA done right is a Social Impact Monitoring Plan that exists before the impacts do. A usable plan is a simple table, and its columns tell you whether it will work. For each significant impact, name the desired outcome, the method you will use to monitor it, who is responsible, how often it is measured, and the indicator that tells you whether you are on track.

Look closely at the responsibility column, because it is where most plans quietly fail. If every line says “the company,” the plan is a compliance exercise. In a plan that works, responsibility is often shared: the company and the district health service, the company and an independent monitor, the company and a farmer association. Shared responsibility is not a nicety. It is how the community gains a real hand on the instruments that measure its own wellbeing, and how the data earns enough trust on both sides to be worth acting on. A number the community helped produce is a number the community will believe.

Frequency matters as much as method. An impact measured once a year can only be corrected once a year, by which point a grievance has had eleven months to harden. Match the cadence to the speed of the impact. Dust and traffic move fast and need near-continuous eyes, and real-time monitors that feed a public dashboard let residents watch the same readings the site does. A change in agricultural livelihoods moves with the season and can be surveyed accordingly. The plan should read like an operating schedule, because that is what it is.

The indicators themselves should mean something to the people being affected, not only to head office. Where the affected group has its own priorities, the monitoring has to reach them. The Indigenous Navigator is one example. It is a tool set built by a consortium including the ILO and the Danish Institute for Human Rights, made so communities can measure the outcomes that matter to them rather than accept a company’s chosen metrics. An assessment that lets the community help set the indicators is measuring the right thing by construction. One that hands down a fixed dashboard is measuring what is easy to defend.

What separates a real SIA from a document nobody revisits

Put two assessments side by side and the difference is visible within minutes. One was written to persuade a regulator. The other was written to run a relationship.

The persuasive assessment is comprehensive, balanced, and inert. It lists impacts without ranking them by who gets hurt. It proposes mitigations without naming who does them or how you would know they worked. It has no baseline you could measure change against, because a baseline invites the question of whether anything improved. It reads as complete. It is, in Vanclay’s terms, a social overview with no future in it.

The working assessment is narrower and sharper. It commits to specific outcomes. It carries a monitoring plan with named owners and honest indicators. It says out loud where the prediction is uncertain and what would trigger a change of course. It treats the community as a party that will hold the document to account, not an audience to be managed past the permit. That last point is the real tell. A document written to be revisited assumes someone will revisit it, and someone with standing to object. A document written to be filed assumes no one will.

The cost of getting this wrong is not abstract. Franks and colleagues showed in 2014 that company-community conflict translates directly into business costs, and that companies routinely fail to record those costs until they are large. An assessment that names an impact and then abandons it is a slow route to exactly that expense. An assessment wired to predict, plan, monitor, and adapt is the cheapest insurance a project can buy.

Run your assessment through the checklist before you sign off on it

The SIA Follow-Through Checklist is a 16-point checklist across five sections, built to test whether your social impact assessment can actually manage impacts rather than just describe them. The five sections walk the full arc: scoping and baseline integrity, impact identification tied to lived experience, the management and mitigation commitments, the monitoring plan and its ownership, and the adaptive loop that keeps the assessment alive after approval. You score each checkpoint In place, Partial, or Absent. Anything scored Partial or Absent is where your assessment will leak, and the clustering tells you which phase you skipped. Use it as a pre-sign-off gate on your own SIA, or as a diligence lens on one a contractor delivered, before the permit clears and the binder heads for the shelf. Download it, run it against your live assessment, and act on the gaps while you still can.

Download the SIA Follow-Through Checklist

When impacts turn into disputes, structure is what holds

Even a well-run assessment will surface impacts that people disagree about. A compensation figure one side calls fair and the other calls an insult. A monitoring result each party reads differently. This is not a failure of the SIA. It is the point at which the assessment stops being a technical exercise and becomes a negotiation, and it needs a process built for that.

The reactive default is to let the disagreement escalate until lawyers or a blockade force the issue. The better path is to route it through independent, structured facilitation before it hardens: a mediator who works from the same assessment both sides already accept, and who helps them turn a contested finding into a workable agreement. Mediation succeeds here precisely because the SIA gave everyone a shared set of facts to argue from. The Social Accord Architecture is the field-tested methodology I use to do this: a common language and a structured sequence that carries a company and a community from the assessment’s raw findings to durable, monitored commitments. It is built for dual accountability, trusted by both sides, which is why it holds when a compliance report would collapse. If you are staring at an assessment that named the right impacts but left you no way to manage the disputes they produced, that is the work. Reach me at thomas@thomasgaultier.com.

An SIA done right does not eliminate conflict. It gives you the facts, the plan, and the relationship to resolve conflict when it comes. Start with one move this week: pull your most recent assessment, find the three impacts most likely to become grievances, and check whether each one has a named owner, a live indicator, and a date for the next measurement. If it does not, you do not have a management tool yet. You have a document. Fix that before the community does it for you.

I am Thomas Gaultier. I mediate company-community disputes in mining, oil and gas, and energy projects, and I have resolved more than 2,000 claims across Sub-Saharan Africa and Europe. More about the work, or tell me what you are dealing with.

If the relationship around a project is under pressure, or heading there, let us talk.

A 30-minute call is enough for me to understand the situation and tell you honestly whether I can help. Active opposition, a social license problem, an FPIC process that needs an independent facilitator, a negotiation a community has to prepare for, or a conflict that is not public yet but will be. It makes no difference which side of the table you are on. No obligation, no generalist pitch.

Every first conversation is confidential.

Cover of Legacy Grievances and Historical Claims by Thomas Gaultier

New release, and the last one

Legacy Grievances and Historical Claims

The tenth and final volume completes the library. It is the field guide to harm that outlived the operation: transitional justice adapted to corporate disputes, successor liability when the asset changes hands, and the pathway behind more than two thousand resolved claims.