(This essay originally appeared as a post by Dennis Petri on the Five4Faith Substack on April 16. Image source: ADF International.)
Every few months, a new cycle of violence in Nigeria produces a familiar pattern in the international press. Attacks on churches, pastors killed, communities displaced. Then come the op-eds and the fact-checks. And reliably, two competing narratives emerge, each claiming to correct the other, and neither giving an accurate account of what the evidence actually shows. For anyone trying to follow the debate carefully, the result is genuine confusion: not because the question is unanswerable, but because the information environment is broken.
Here is what that looks like in practice. A reader follows the fact-checks in the New York Times, the BBC, or Trouw, sees Intersociety’s numbers dismantled, and reasonably concludes that the religious-targeting claim has been dealt with. That reader will almost never encounter the Observatory of Religious Freedom in Africa (ORFA)’s incident-level data on attacks against Christian communities, the joint International Institute for Religious Freedom (IIRF) and Open Doors report No Road Home on displaced Christians in Benue and Plateau, or the proxy estimation John Bainbridge and I published in Frontiers in Sociology. The debunk travels. The counter-evidence does not.
The result is an information environment in which a basic question remains unanswered: is this violence targeting Christians, or is it something else dressed up in religious language? That question deserves a serious answer. What it gets instead is a choice between an overblown framing that the data cannot support and a dismissive counter-narrative that the data also cannot support.
Both need to be corrected.
The critique that lands
The statistics most often cited by advocates of the strongest anti-Christian framing come from a single Nigerian civil society organization, Intersociety, whose reports have been amplified by politicians including Donald Trump and by a wide range of religious advocacy groups. Mainstream outlets including the New York Times, the BBC, and Dutch newspaper Trouw have pointed out, correctly, that Intersociety publishes striking numbers without transparent sourcing. When a statistic cannot be traced to a verifiable methodology, it cannot be used as evidence. That is not political bias; it is basic epistemic hygiene.
And the rhetoric that accompanies those numbers often goes further than any data can support. Calling what is happening in Nigeria genocide is a serious legal and political claim. It requires a specific evidentiary standard. The figures in circulation do not meet it. Critics who push back on that framing are right to do so.
But here is where careful analysis stops and something more convenient begins.
The leap that does not follow
From “Intersociety’s data is unreliable” and “the genocide framing is overblown,” much commentary draws a third conclusion: that there is no credible evidence of religiously targeted violence at all, and that the conflict is better understood as a dispute over land and water driven by climate change and demographic pressure.
This third claim does not follow from the first two. Debunking a bad source does not validate the opposite conclusion. And the land-and-water framing, while it has real empirical content, is being applied far beyond what its actual explanatory range can support. Rarely are the roots of religious conflict mono-causal.
It is true that Fulani herder communities and farming communities are locked in genuine competition over grazing routes and water sources, competition that climate change is making more acute. That is documented, serious, and worth taking into account. But the resource-competition model cannot account for the geographic consistency of which villages are attacked, which buildings are targeted, and which community leaders are killed. If this were purely a resource conflict, we would expect attacks distributed more evenly across farming communities regardless of their religious composition. What the data from the Observatory of Religious Freedom in Africa (ORFA) actually show is a pattern of direct targeting that tracks religious identity in ways a resource model cannot explain.
This is not about reading minds or reconstructing what perpetrators intended. It is about what the attack data show. When churches and church gatherings are attacked at rates that far exceed the presence of Christian communities in the population, and when that pattern holds consistently across regions and over time, the evidence of targeting is in the pattern itself.
The land-and-water framing circulates comfortably in academic and humanitarian circles accustomed to explaining conflict through material and structural factors, and journalists tend to draw from those same institutional sources. Religious freedom research bodies like ORFA and IIRF are known in specialist circles but rarely appear in the Rolodex of conflict correspondents. The result is that when Intersociety’s numbers are debunked, the debunking feels complete, and the more careful research pointing to a real pattern of religious targeting never enters the frame.
The structural problem underneath
What makes this debate so difficult to resolve is a structural problem in how humanitarian data is collected. Whether it’s the datasets from ACLED, UNHCR, the International Organization for Migration, and the Internal Displacement Monitoring, they do not contain a column for the religious identity of displaced persons. Three separate dynamics explain why.
The first is simple unawareness. Many humanitarian organizations have not historically treated religious identity as a structuring factor in conflict, and their data collection frameworks were built without it. In societies drenched in religion, this is gross neglect. The second is deliberate avoidance: tracking the religion of displaced persons is resisted out of concern that it could politicize operations or compromise neutrality across conflict lines. This is more understandable, but failing to acknowledge the religious variable results in interverntions that are not sufficienctly sensitive or properly targeted, resulting in literally millions of dollars wasted. These are not trivial objections, but the consequence is that a protection-relevant variable is systematically excluded from the record. The third is a practical limitation: in active conflict zones, basic registration is often incomplete or impossible regardless of what organizations intend to collect.
Each dynamic produces the same outcome. Religion disappears from the analyses that inform policy and media coverage. The absence of data in official sources is then routinely misread as evidence that the phenomenon does not exist (and creates a self-reinforcing spiral). It is not. It is evidence that the phenomenon was never measured. Because religion is not recorded, it disappears from the analyses. Because it is absent from the analyses, it becomes easier to argue it is not a relevant factor. Because it is treated as irrelevant, there is no pressure to start recording it. The absence of evidence manufactures the appearance of evidence of absence.
This is exactly what ORFA’s incident-level data is designed to escape. By tracking attacks on religious communities directly, ORFA makes the pattern visible that aggregate displacement figures erase.
The narrative needs to be corrected
This matters because narratives have consequences. When the dominant media framing reduces a documented pattern of violence to a climate-driven resource dispute, it shapes what questions get asked, what protection needs get addressed, and what policy responses get designed. Displaced Christians in Benue and Plateau states face specific vulnerabilities that a generic displacement framework does not capture. If the framing that informs humanitarian response does not acknowledge the religious dimension of the targeting, those vulnerabilities remain invisible to the systems meant to address them.
The specifics matter. Displaced Christians in Benue and Plateau face risks that a generic IDP framework treats as incidental. Their pastors and catechists have been singled out in killings and kidnappings, leaving congregations without the figures who would normally coordinate return and recovery. Their parish churches, which in rural Middle Belt communities function as the de facto civic center, have been burned, which means the community anchor is gone before the displacement response even begins. Their farmland has in many cases been occupied by armed herder groups, making physical return to ancestral villages impossible on any humanitarian timeline. And inside mixed IDP camps, Christian families have reported repeat attacks, abductions of girls, and pressure to convert, none of which appear in ACLED’s event categories or in UNHCR’s protection indicators because neither dataset disaggregates by religious identity. The joint IIRF, Open Doors, and World Evangelical Alliance report No Road Home documents each of these patterns with case-level detail.
The corrective is not to swing to the opposite extreme. Reinstating the genocide framing or rehabilitating Intersociety’s numbers would trade one distortion for another. The corrective is to insist on the space the evidence actually occupies: a conflict with real resource dimensions and a real, measurable pattern of religious targeting, documented by organizations whose methodology is transparent and whose data is publicly available.
Major media outlets that have worked to debunk the overblown framing have a responsibility to apply the same rigor to the counter-narrative they have implicitly endorsed. The land-and-water explanation has not been established by evidence; it has been asserted as the reasonable alternative to a discredited source. That is not the same thing. The data exists. The methodology is sound. The task now is to make that evidence impossible to ignore.
What the evidence actually shows
The genocide framing is not supported by the evidence and should be retired. Intersociety’s numbers are not a reliable foundation for any conclusion. Those points stand. To be concrete about what should be retired: the untraceable aggregate death tolls that Intersociety puts into circulation and that political figures then quote as fact, the description of the violence as a completed or ongoing genocide when the legal threshold for that term has not been met, and the broader civilizational-erasure framing that presents Nigeria as evidence of a planned extinction of Christianity. These framings move faster than the evidence behind them, and when they collapse under scrutiny, they take the credibility of the underlying pattern down with them.
But the claim that religion is simply not a factor, that what is happening in Nigeria is a climate-driven resource conflict that happens to produce images of burning churches, is equally unsupported. ORFA’s data shows direct targeting of Christian communities in patterns that exceed their demographic weight significantly. The joint report of IIRF, Open Doors, and the World Evangelical Alliance, No Road Home, documents the specific vulnerability of displaced Christians in Benue and Plateau states with granular detail that the resource-competition narrative cannot absorb.
The coverage is missing something: a willingness to sit in the space between an overblown claim and a dismissive counter-narrative, and to ask what the best available evidence actually shows. That space is where the truth is, and it is where this conversation needs to go.
For anyone asking what that looks like in practice, the steps are not abstract. A journalist covering Nigeria can pair any ACLED figure with the corresponding ORFA incident data before filing, which takes one query and prevents the religious pattern from being erased at the sourcing stage. A humanitarian CEO can add a single religious-identity field to intake and registration protocols, piloted with trusted local partners in Benue and Plateau, and commit to publishing the disaggregated results. An editor can make a standing decision that coverage of Middle Belt violence does not go to press without a call to a researcher who works on religious targeting, the same way it would not go to press without a call to someone who works on herder-farmer dynamics. At IIRF we provide the methodological backbone for this: the Violent Incidents Database and ORFA’s incident records are publicly available, the No Road Home report sets out protection guidance at case level, and our team offers briefings and methodological training for newsrooms and humanitarian agencies that want to build religious-identity analysis into their standard workflow. The information gap is fixable. It requires the people who hold the microphones and the intake forms to treat fixing it as part of their job.