The survey most mines run is not the survey they need
Sit in on a stakeholder engagement meeting at almost any South African operation and you will hear the same figure quoted: rand spent on community projects. Boreholes drilled. Bursaries awarded. Classrooms built. The number is real and the work is often good. It is also not what a social licence is made of.
A social licence is a measure of what people believe about you, not what you have spent on them. The two come apart more often than mine management expects. An operation can be well ahead of its Social and Labour Plan commitments and still be blockaded on a Monday morning, because the households nearest the haul road have concluded that the benefits go somewhere else.
A perception survey is the instrument that tells you the difference. Run badly, it produces a document that reassures the board and predicts nothing. Run properly, it tells you where consent is thin, which villages are drifting, and how much time you have.
Three design faults that show up again and again
Surveying only the structures you already talk to
The fastest way to produce a comfortable and useless result is to route the survey through the recognised community structures — the forum, the trust, the traditional council. These are the people already in the room. They are not a sample. They are the group most likely to report that engagement is working, because they are the engagement.
The households that organise a protest are almost never the households on that list. A survey that cannot reach them cannot warn you about them. Sampling has to be drawn from the settlement, at household level, with a defined geographic frame — typically the wards falling inside the operation’s identified zone of influence — and it has to include people who have never attended a meeting.
Asking about the mine instead of about the household
“Are you satisfied with the mine’s community investment?” is a question that invites a strategic answer. People who want more will say no regardless. People who fear losing what they have will say yes regardless.
Better questions start from the household’s own position and work outwards. Has your water supply improved or worsened in the past two years? Has anyone in this household applied for work at the operation, and what happened? Do you know where to report a grievance, and has anyone you know done so? These produce answers you can act on, and they can be tracked over time without the wording drifting.
Measuring once and calling it a baseline
A single survey is a photograph. Social licence is a trend. The value in this work sits almost entirely in the second and third rounds, when you can see which villages moved and in which direction. One-off studies commissioned in the month before a Section 11 application or a mining right renewal are visibly transactional, and communities read them that way.
Fix the instrument early, then repeat it on a fixed cycle — annually is usually enough, twice a year where an operation is in transition or approaching closure.
What a survey needs to survive scrutiny
Assume the result will eventually be read by a regulator, a lender’s environmental and social team, or a lawyer acting for the community. Design for that reader from the start.
- A written sampling frame that states how households were selected and what the margin of error is at ward level, not only for the sample as a whole.
- Enumerators who speak the languages actually spoken in the settlement, and who are not employees of the operation or of a contractor to it.
- Instrument questions kept identical between rounds, so that a change in the result means a change in the community rather than a change in the wording.
- Data handled under POPIA — consent recorded, personal information separated from responses, and a stated retention period.
- Disaggregation by ward, gender and age, because the aggregate almost always hides the group that matters.
- Findings reported back to the communities surveyed, in their own languages, before the report reaches the board.
That last point costs little and changes the standing of the whole exercise. A community that has seen its own results treats the next round as something it participates in rather than something done to it.
Where this connects to the rest of your compliance work
Perception data does not sit on its own. It reads alongside environmental monitoring, because the complaints that damage a social licence most reliably are dust, water and blasting. An operation that can show both its rehabilitation and biodiversity monitoring results and a community’s changing perception of them is in a far stronger position at a renewal hearing than one holding either alone.
It also feeds skills and procurement planning. If the survey shows that residents believe local hiring is closed to them, the answer is usually a visible, verifiable skills development pathway with published intake numbers, rather than another communications campaign.
What to do on Monday
You do not need a budget approval to start. Take your last community engagement report and check three things. Does it distinguish between what you spent and what people believe? Does it name the wards, or only the community? Could you repeat the same measurement next year and get a comparable number?
If the answer to any of those is no, you have identified your first piece of work, and it is a scoping conversation rather than a procurement exercise. IALE Institute designs and runs perception and social licence research across South Africa and the wider SADC region, from offices in Sandton, Kimberley and Kuruman.