Conflict early warning is the practice of collecting signs of rising tension, checking them, and getting them to people who can act before violence breaks out. A warning system has four parts: people who notice, a way to report, someone who checks and judges, and someone who responds. If any one of these fails, the system fails. As one researcher writes, success starts with a reliable analysis of local conflict dynamics.
Key points:
- It is about prevention, and it only works if someone acts.
- Local knowledge usually beats national statistics for predicting local violence.
- Phones and apps help, and they can also be used to organise violence.
- A report is not a fact until someone has checked it.
What does an early warning system do?
Think of a local example. A shopkeeper sees leaflets telling a community to leave. A trained monitor reports it. An analyst checks it against other reports, and calls a local chief. A peace committee meets the same day, and the police visit. If that happens quickly, the leaflets may come to nothing.
What are its four parts?
Monitoring means trained people, and the public, watching for signs. Reporting means a simple, safe way to send what they see. Analysis means checking the report, comparing it with others, and judging the risk. Response means someone with the power and will to act, in time.
Why does local matter so much?
Peace Insight notes that early warning used to be the preserve of large multilateral efforts and statistical modelling, and that it has taken a local turn. Country-level data struggles to predict local violence, and large bodies often struggle to act quickly. Local people know who is angry, and why, and they have the most to lose.
How do mobile phones help?
Violence is often the outcome of a lack of communication, as that article argues: too little information about the intentions of another group can feed fear. Phones let people report and check rumours quickly. The same tools can be used to organise violence, so peacebuilders have to plan for that.
What is the difference between early warning and intelligence?
Security intelligence serves the State, is usually secret, and focuses on threats. Community early warning aims to protect people, and works best when it's open, shared and trusted by the communities involved. Mixing the two can put monitors at risk, so be clear about who sees what.
Why do systems fail?
Most fail at the last step. Warnings arrive, and nobody acts. Other common problems are poor data, false alarms that wear out trust, funding that ends after an election, and monitors who feel unsafe. CEWARN itself lists declining and constrained funding among its priorities.
What makes a warning useful?
A useful warning says what is happening, where, when, who is involved, how reliable the source is and what might happen next. "Tension is high in the county" is too vague. "Youths from two villages are gathering at the market after a land dispute" lets someone act.
How do you avoid false alarms?
Check before you escalate, using a second source. Mark each report as unverified, and say so when you pass it on. Rumours are themselves a sign worth noting, even if they're false, because fear can cause violence. Track what happened after each warning, and learn from it.
How do you protect monitors and sources?
Keep names away from reports that circulate. Use codes for places where it's safer. Train people on when to step back. Nobody should be asked to go near danger to get a report.
Who runs early warning in Kenya?
Several bodies do, at different levels. Early warning systems in Kenya sets them out, including the police, NCIC, civil society and regional bodies.
How can a county start?
Start small. Agree what you'll watch, who reports, who checks and who acts. Use phones and tools people already have. Hold a practice run. Then widen it. A simple system that gets answers will do more good than a sophisticated one that nobody trusts.
How does Kenyan policy define it?
The national peace policy calls early warning the act of alerting a competent authority about the threat of new or renewed conflict, sufficiently in advance for preventive action. It describes early warning mechanisms as processes and structures that collect and analyse data on conflict indicators and give advance warning of risks and vulnerabilities.
Two words matter there. "Competent" means someone who can act. "Advance" means in time.
How is it different from hotspot mapping?
A hotspot map shows where risk is high over months. Early warning tells you what's happening this week in that place. You need both: the map tells you where to look, and the warning tells you when to move.
Why do warnings need to be local?
Conflict in Kenya is increasingly local. The AFSC study says that election violence now shows up at the level of wards and informal settlements. A national dashboard won't see a quarrel at a borehole in time.
What does a good day look like?
A monitor texts about strangers recruiting boys at a market. An analyst calls two contacts by noon. By evening, the chief, a youth leader and the local police commander have met. The recruiters leave. Nothing happens, and that's the point. Keep a record of it, because success looks like nothing.
Read next
- Conflict early warning system for Kenya: the TI pack page for this work
- Conflict early warning in Kenya: all the pages in this section
- Conflict early warning indicators
- From warning to response
- Early warning systems in Kenya
- Takwimu Intelligence (TI) home page
- National cohesion in Kenya: the guide to this library
- The TI platform and how we check our numbers
Where TI fits
Takwimu Intelligence (TI) groups reports of one event, keeps each unverified until a person confirms it, and shows what was done, so a warning leads to a record of response.
To see it on your own work, book a walk-through.