Guinea, 2014. The messages are failing. You have been asked to fix them.
"You say you're doctors, but you've been here for eighteen months and you haven't treated a single patient."
Ebola is spreading through Guinea's rural villages. Roughly half of all patients are dying. Traditional burial practices — washing, dressing, and holding the body — are driving transmission. Stopping the epidemic requires community cooperation: families must report illness, name contacts, and change burial customs that carry deep cultural meaning.
CDC teams had been in Guinea for eighteen months. They had focused, appropriately, on infection control, contact tracing, and stopping transmission. What they had not done was treat patients. Communities noticed. Dr. Abdou Salam Gueye, a Senegalese epidemiologist with deep regional knowledge who CDC deployed to Guinea, reported what people in communities were saying: "You say you're doctors, but you've been here for eighteen months and you haven't treated a single patient."
This exercise has two connected acts. In Act I you will design Guinea's Ebola communication strategy — configuring the message and selecting its messenger. In Act II you will design the enforcement approach. Each decision updates the Community Trust score that runs across both acts. The trust level you carry into Act II changes what kinds of enforcement can succeed — making the connection between communication and authority genuinely operational rather than abstract.
Act I has two stations: (1) message configurator — you assemble a message by setting three independent design dimensions, then see how your combination maps against what worked; (2) messenger — you identify effective messenger profiles from a set of options and explain what makes each credible. Act II has two stations: (3) conditions assessment — you rate whether four conditions for effective mandates are met; (4) enforcement design — your choices are evaluated against your own assessment. Allow 15–20 minutes for the full exercise.
Design a replacement message. Set three dimensions, then see what your combination produces.
Each dimension addresses a different failure mode from the early messages. As you choose, identify the causal mechanism: how does each attribute translate — or fail to translate — into a specific community behavior? The question is not "Which sounds better?" but "Which produces the right action in this specific community context, given what you know about the trust deficit that already exists?"
The two early Guinea messages — "Don't eat bushmeat" and "Ebola cannot be cured" — were factually accurate. Both made things worse. For each design dimension below, choose the attribute your replacement message should have. Your live message preview updates as you choose.
The message is ready. Now identify which messengers will actually be heard — and why.
Terrie Hall and Christian Nwigwe were effective in very different settings, but both had something in common that institutional messengers lacked: the audience believed them. Nwigwe spoke no Spanish; Hall had no medical credentials. Ask what kind of credibility each messenger below carries — and whether it is the right kind for this specific context, given the trust deficit that already exists between communities and outside health workers.
You are deploying to rural Guinea communities. CDC teams have been present for eighteen months without treating patients. Community distrust of outside health workers is established. The message needs to reach families in villages where government authority is associated with surveillance, not support. Select two messenger profiles that together give the best chance of reaching these communities effectively. Explain your reasoning by selecting the credibility type that applies to each.
For each messenger below, decide: Include or Exclude from your strategy. Then, for those you include, identify the type of credibility they carry. Select exactly two messengers.
What your communication decisions built — and what carries into Act II.
Trust is not a sentiment — it is an operational condition. The score above reflects the communication foundation your response has built. In Act II, it sets the conditions into which your enforcement approach will land. A higher score means communities are more likely to report cases, name contacts, and engage honestly with restrictions. A lower score means enforcement must do more work to produce compliance, and is more likely to produce evasion instead.
Notice that the trust deficit in Guinea was not primarily a communication failure — it was a program design failure. CDC teams had been present for eighteen months conducting surveillance without delivering visible services. No message strategy, however well designed, can fully repair a gap that program design created. The lesson is not that communication is unimportant; it is that communication operates within conditions that program design sets. A communication strategy can build on a strong foundation or try to compensate for a weak one — but it cannot substitute for it.
Before designing enforcement, assess the conditions on the ground.
"The honey has to get there before the flies leave!"
The instructor materials identify four conditions under which mandatory restrictions can work. Your task here is not to apply a framework mechanically — it is to assess the actual situation in rural Guinea in mid-2015, based on what you know from the case. Your assessment in this station will be carried forward: in Station 2, your enforcement design choices will be evaluated against the conditions you identify here. An enforcement design inconsistent with your own conditions assessment is the most analytically productive thing the exercise can surface.
Ebola continues to spread in Guinea's rural villages. President Alpha Condé is under pressure to act. His initial instinct is to force communities to close — to prevent exposed contacts from traveling and seeding new clusters. It was a reasonable response from a leader watching his country suffer.
Rate each of the four conditions for effective mandatory restriction as it applies to this specific context. Use the case materials, not general principles.
Given your conditions assessment, how should the government respond when a village reports an Ebola case?
Each restriction model encodes an assumption about what produces compliance. The consistency check will compare your design choices against the conditions you assessed in Station 1. An enforcement design that is internally inconsistent with your own assessment of the conditions — for instance, choosing hard restriction after rating trust as "not met" — is the most analytically important output this exercise produces. It is not a mistake; it is the scenario that most deserves discussion.
Step 1 — Choose the movement restriction model
Your approach in practice
The contrast between Guinea's microcerclage and Sierra Leone's police enforcement is a natural experiment within the same epidemic. Same virus. Same regional context. Same formal authority of the state. Different instruments — and dramatically different outcomes.
Sierra Leone posted police outside the homes of named contacts throughout the epidemic. Contacts who relied on agriculture could not tend their fields; planting season passed, and families faced a year of food insecurity. Stigma attached to anyone identified as a contact. Bribes became the mechanism for leaving. Patients stopped naming contacts to protect their families from hardship. Trust in government collapsed. Sierra Leone's epidemic continued for many months longer than it might have with a different approach.
Guinea's microcerclage produced a different signal: nearby communities without any Ebola cases asked whether they too could undergo microcerclage. An enforcement mechanism had become a sought-after service.
What these two cases reveal about communication, enforcement, and trust.
Accuracy is necessary — not sufficient
A message that is true but produces the wrong behavior is a failed message. The test is not "Is this accurate?" but "Does this produce the right action?"
The messenger is the message
Credibility comes from proximity to the experience of the audience — not from institutional authority. Health systems that rely only on official spokespeople will fail to reach communities that do not trust officials.
Legal authority is a tool, not a strategy
Restrictions can interrupt transmission only when communities cooperate with the broader response. Authority without trust is an engine without traction.
Reciprocal obligation transforms enforcement into service
When restriction is paired with comprehensive support, state authority becomes a resource communities seek out rather than evade. The signal is voluntary demand for the intervention.
Trust is an operational asset, not a sentiment
Responders who deliver no visible services earn no social license to conduct surveillance. Cooperation with invisible interventions requires demonstrated commitment to community wellbeing.
Support must precede the crisis of trust
Once communities have experienced the state as an adversary, no mandate can rebuild trust quickly enough to change the trajectory of an outbreak. Community relationships must be built between emergencies, not during them.
Discussion questions
For individual reflection or small-group discussion. Graduate extensions appear in graduate mode.
- Both early Ebola messages in Guinea were factually accurate. Under what conditions can an accurate message undermine public health outcomes? What standard should replace accuracy as the primary test of a message?
Trace the causal mechanism: from "Ebola cannot be cured" to a specific community decision, to a specific epidemiological consequence. Then ask: what monitoring system would have detected this behavioral failure before it became embedded in the response?
- Microcerclage imposed travel restrictions and delivered services simultaneously. How did that combination change communities' relationship to the intervention? What does this tell us about the relationship between rights, obligations, and trust in public health law?
Map the rights and obligations on both sides of the microcerclage contract. The standard framing — the state imposes a restriction — misses that the state also incurred obligations. How does naming this as a mutual contract, rather than a unilateral exercise of authority, change the political and legal analysis of the intervention?
- Dr. Gueye reported that communities saw CDC teams as doctors who had never treated a patient. Both CDC's self-understanding and communities' interpretation were simultaneously true. What does this gap reveal about how trust is built and destroyed during outbreak response?
The gap cannot be corrected through argument — neither side's account is wrong. What program design changes would have closed this gap? At what point in the response is it too late to close it through program redesign alone?
- Your country's health minister wants to impose mandatory quarantine for household contacts of a novel respiratory virus. Based on the Guinea and Sierra Leone cases, what conditions would need to be met for mandatory quarantine to be effective? What legal and operational safeguards would you recommend?
Identify the operational conditions — not just legal ones — that determine whether the measure produces compliance or evasion. Which condition is hardest to establish in the timeframe of an acute outbreak? What does that imply for pre-outbreak investment?
- President Condé said: "the honey has to get there before the flies leave." What does this mean operationally — not as a communication strategy, but as an infrastructure investment? What would a ministry of health need to have built before this outbreak began?
Estimate the lead time. If an outbreak is declared today, how long would it take for comprehensive support services to reach a remote rural community in a low-income country? If the answer is weeks, what pre-outbreak investments would shorten it to days — and who is responsible for making them?
These two cases together make an argument about program design rather than communication strategy or legal authority. The common failure — in messaging and in enforcement — was treating the visible intervention as independent of the conditions surrounding it. "Ebola cannot be cured" failed not because it was inaccurate, but because it was deployed into a context where communities had no reason to believe cooperation would benefit them. Sierra Leone's enforcement failed not because legal authority was wrong, but because authority was exercised without the reciprocal obligations that make it legitimate in practice. The design question in both cases is the same: what conditions must surround this intervention for it to produce the behavior it requires?