445 Casualties In Fourteen Days, And Where They Happened
Abuja · Thursday 20 August 2026
Between 26 July and 8 August, Nigeria recorded 106 violent conflict incidents. They produced 445 casualties and 256 kidnapping victims. Fourteen days.
The figures come from a conflict tracking report carried by Nairametrics this week, and they are worth sitting with before the next announcement arrives to cover them over.
The concentration is the story
National totals in Nigeria almost always hide more than they show, because the violence is not distributed anything like evenly. Break the same fortnight down by region and the picture sharpens considerably.
The North West's share of a national fortnight
26 July to 8 August 2026
Share of the national total attributable to the North West. Source: conflict tracking data for 26 July to 8 August 2026.
Thirty-eight per cent of the incidents. Sixty-one per cent of the casualties. Eighty-three per cent of everyone kidnapped in Nigeria in that fortnight.
The North West is not having a worse version of the national problem. It is having a different problem, and the national figure is mostly a description of it.
Banditry, specifically
Within the same period, banditry on its own accounted for 35 incidents, 246 casualties and 196 kidnapping victims. One category of violence, a third of the incidents, more than half the deaths and injuries, and roughly three quarters of the abductions.
| 26 July to 8 August 2026 | Incidents | Casualties | Kidnapped |
|---|---|---|---|
| All violent conflict, nationally | 106 | 445 | 256 |
| Of which banditry | 35 | 246 | 196 |
| Banditry as a share | 33% | 55% | 77% |
Figures as reported. Casualties combine deaths and injuries.
Why kidnapping tracks differently from killing
Notice that the kidnapping share is higher than the casualty share, which is higher than the incident share. That gradient is not noise. It describes a business.
Insurgency kills because killing is the objective. Banditry kidnaps because kidnapping produces revenue, and killing is what happens when the revenue does not arrive or when a demonstration is required. An armed group running abduction as a commercial operation has every reason to take many people and kill relatively few of them, which is roughly what the numbers show.
It also explains why the North West figures are so resistant to the kind of response that worked, partially, against Boko Haram. You cannot degrade a business by taking territory from it if it does not hold territory. It moves.
Channels Television on the persistence of the problem.
The rescues are real and they are not the trend
Security forces did have results in the same window. Thirty-three passengers were freed in Tsafe local government area in Zamfara during August, and there have been strikes against named commanders.
Those operations matter enormously to the thirty-three people involved and they should be reported as what they are. They are also, against 256 abductions in a fortnight, not evidence that the direction has changed. A rescue is a data point. A ratio is a trend.
You cannot degrade a business by taking territory from it if it does not hold territory. It moves.
Read this next to Wednesday's announcement
The presidency ordered a National Threat Assessment on Wednesday, with ninety days to produce a ranking of the principal threats facing Nigeria by relative urgency.
This fortnight is what that ranking will have to explain. If the finished assessment does not place North West banditry at or near the top, it will be describing a country other than the one the casualty figures describe.
What these numbers are, and what they are not
Conflict data in Nigeria is compiled from media reporting, security force statements and local sources, then cross-checked. That method has known limits and anyone using the figures should hold them.
It undercounts. Violence in places with no reporter, no phone signal and no road is violence that frequently does not enter the record at all, and those are precisely the places where the worst of it happens. When a total like 445 is wrong, it is almost always wrong downwards.
It also conflates. "Casualties" here covers deaths and injuries together, which is standard practice and which flattens a real distinction. And an "incident" is a unit of reporting, not a unit of harm: a raid on a village and a single roadside killing both count as one.
None of that makes the figures useless. It makes them a floor rather than a measurement, and it means the comparisons between regions are more reliable than the absolute totals, because the same method is being applied everywhere.
Two different wars, counted as one problem
Nigerian security is usually discussed as a single national failure. It is at least two distinct conflicts with different structures, and treating them as one is why responses keep getting misapplied.
| North East insurgency | North West banditry | |
|---|---|---|
| Objective | Territorial and ideological | Revenue |
| Structure | Organised, hierarchical, named | Fragmented, fluid, many small groups |
| Holds ground | Yes, historically | Rarely, and does not need to |
| Primary harm | Killing and displacement | Abduction for ransom |
| Negotiable | Very difficult | Constantly, and locally |
| Responds to | Territorial military pressure | Unclear, and that is the problem |
A simplification, but a more useful one than the single category of "insecurity".
An insurgency that holds ground can be pushed off it, and Nigeria has done that. A kidnapping economy has no ground to be pushed off. Its assets are motorcycles, weapons, forest cover and knowledge of the terrain, and every one of those relocates in an afternoon.
The economics that keep it running
Ransom is the engine, and the engine is efficient in a way that is uncomfortable to state.
Taking a person costs very little: fuel, a few men, an hour. The payment, whether it comes from a family selling land or from a community raising a collection, arrives in cash and requires no laundering. There is no supply chain to interdict, no market to disrupt, and no export route to close. Almost the entire proceeds are profit.
Meanwhile the cost of not paying is a dead relative, so people pay. Governments announce that ransoms are illegal, families pay anyway because the alternative is unbearable, and the announcement changes nothing except whether the payment is discussed openly.
That is why the North West figures have proved so resistant. There is no pressure point in the model that conventional security operations reach. The things that would actually degrade it, rural policing at a density Nigeria has never had, road security, and enough state presence that a village can call for help and get it, are expensive, slow and unglamorous.
What this fortnight should tell the committee
The National Threat Assessment ordered on Wednesday has ninety days to rank Nigeria's threats by relative urgency.
Eighty-three per cent of the country's kidnappings in a fortnight came from one region. If the ranking that emerges in November does not reflect that, the document will have failed at the only task that actually matters, which is telling the truth about where the harm is.
Reporting: Nairametrics, Human Rights Watch World Report 2026, and conflict tracking data covering 26 July to 8 August 2026.
