By Oyewole Sarumi
The announcement landed with the weight of a policy earthquake. The Federal Government of Nigeria, through Education Minister Dr Tunji Alausa, declared that the days of estimate-based planning were over. Henceforth, the nation would adopt Kaduna State’s household mapping model as the national framework for identifying out-of-school children—a data-driven approach that would replace broad projections with verified, household-level information. On its face, this sounds like an unqualified triumph for evidence-based policymaking. Who could possibly object to better data?
The problem, however, is that the headline is not the methodology. And in the fraught world of education statistics, the methodology is everything.
The central question before Nigeria today is not whether Kaduna’s figure of fewer than 200,000 out-of-school children is correct, or whether the older estimate of approximately 585,000 was wrong. The real forensic question is far more fundamental: What exactly did Kaduna count? Who was counted and who was excluded? What definition of “out-of-school” was used? How were children living outside conventional households captured? And crucially, can this methodology legitimately be scaled to the entire country without losing the most vulnerable children in the process?
Until these questions are answered publicly and transparently, it is premature to conclude that Nigeria has discovered its “true” out-of-school children figure—or that the previous national estimates were somehow invalid. This is not an argument against improving education data. It is an argument for doing so with the rigour, transparency and methodological sophistication that a crisis of this magnitude demands.
The Scale of the Crisis We Are Trying to Measure
Before we interrogate the methodology, we must first appreciate what is at stake. Nigeria currently holds the unenviable distinction of harbouring the world’s largest population of out-of-school children. According to UNESCO, the figure exceeds 20 million. UNICEF’s 2024 report puts the number at approximately 18.3 million, comprising 10.2 million children of primary school age and 8.1 million at the junior secondary level. Save the Children, in January 2026, placed the figure even higher, at around 28 million children and adolescents. One in every five out-of-school children globally is Nigerian, and Nigeria accounts for approximately 15 per cent of the world’s total.
The burden is overwhelmingly concentrated in the north. Between 73 and 80 per cent of Nigeria’s out-of-school children are in the northern states. In the North-West and North-East, only 30 per cent of school-age children are in formal classrooms. The Northern States Governors’ Forum itself admits that 80 per cent of the 18.3 million figure comes from the region. Three states—Kano, Katsina and Jigawa—account for 16 per cent of the national total, with Kano alone harbouring close to 900,000 children out of school.
At the heart of this crisis lies the Almajiri system. Estimates indicate there are between 7 and 10 million Almajiri children in Northern Nigeria, with some figures suggesting the number may be as high as 9.5 million. Almajiri children are estimated to make up 72 to 81 per cent of Nigeria’s out-of-school population. These are boys sent far from their home communities to study under traditional Qur’anic teachers, often forced to beg for survival. They constitute, in the words of Governor Abdullahi Sule of Nasarawa, “the single largest structural contributor” to Nigeria’s education crisis.
These are not abstract statistics. They represent millions of children whose futures are being foreclosed, generation after generation. And it is precisely because the stakes are so high that we cannot afford to get the numbers wrong—not in either direction. An inflated figure may attract more resources, but a deflated figure may create a dangerous complacency that leaves millions of excluded children invisible.
The First Forensic Distinction: Estimate Is Not the Same as Census
The Federal Government’s argument appears to rest on a fundamental distinction between estimated out-of-school children numbers and household-mapped out-of-school children numbers. Minister Alausa has been explicit: “Planning has largely been based on projections that could not accurately show where out-of-school children were located”. The new exercise, he argues, “changes that by providing verified information”.
This distinction is legitimate as far as it goes. But it is potentially misleading to imply that the first is automatically unreliable and the second automatically represents the “actual” number. Internationally, serious education statisticians do not generally treat the problem as a simplistic contest between “estimates” and “actual counts.”
UNESCO’s Institute for Statistics, for example, uses a combination of administrative enrolment data, population estimates, censuses and household surveys. UNESCO explicitly recognises that different sources measure different things: administrative data typically measure enrolment, while household surveys measure attendance. The organisation distinguishes between these different statistical concepts rather than assuming that one source alone necessarily produces the definitive answer. As UNESCO has documented, different sources can generate substantially different estimates, and the objective should be to harmonise definitions and understand not only the number of excluded children but also who they are and the barriers preventing their participation.
This is extremely important for Nigeria. Suppose administrative records suggest 585,000 children are not enrolled in Kaduna. A household survey identifies 300,000 children who did not attend school. A household enumeration finds 200,000 children physically present in mapped households. A separate survey identifies 150,000 children engaged in informal religious education but not formal schooling. Are these figures contradictory? Not necessarily. They may simply be measuring different populations.
The critical question is therefore: what population did Kaduna’s under-200,000 figure actually represent? Was it children who were physically found in mapped households? Children of school age who were registered at a household address? Children who were not enrolled in formal schools? Children who were not attending formal schools? Children who had never attended formal school? Children absent from school at the time of the survey? Children aged 6 to 15, or 5 to 17? Primary-school-age children only, or primary and junior-secondary-age children? Children receiving any form of Islamic or Qur’anic education? These definitions can produce radically different numbers. Without knowing the precise definitional framework, comparing 585,000 with under 200,000 is potentially comparing apples with oranges.
The Biggest Missing Piece: What Exactly Is the “Kaduna Model”?
This is, in my assessment, the most important unanswered question in the entire announcement. The Federal Government says it will adopt Kaduna’s “household mapping model.” Fine. But what does that mean operationally?
From the available information, we know that Kaduna’s mapping exercise covered 360,319 households in communities with high concentrations of out-of-school children, identifying 187,719 children who are currently out of school. We know that poverty was identified by over 91 per cent of affected households as the major barrier to enrolment.
We know that the exercise covered all 23 local government areas of the state and employed a digitised Education Management Information System. We know that Save the Children International has described it as a “landmark study that replaced guesswork with household level data”.
These are all positive developments. But they are not a methodology. A credible methodology should publicly disclose, at a minimum, the following: the precise definition of an out-of-school child; the age range covered; the geographical sampling frame; the number of enumeration areas visited; whether the exercise was a census or a sample survey; the response rate; how non-response was treated; how children absent from households were captured; how internally displaced children were counted; how nomadic populations were covered; how Almajiri children were identified; how children living in Qur’anic schools were enumerated; how street-connected children were captured; how children in informal settlements were enumerated; how children in boarding facilities were treated; how children in child-care institutions were counted; how children temporarily away from home were handled; how children in conflict-affected areas were enumerated; how duplicate records were eliminated; how the results were independently validated; whether the methodology was peer-reviewed; and whether independent statisticians participated.
Without these details, the claim that Kaduna has discovered the “actual” number remains unproven. It may have discovered a more accurate number. But it may also have discovered a narrower number. Those are not necessarily the same thing. The distinction is not pedantic; it is the difference between a credible statistical exercise and a potentially misleading political headline.
The Almajiri Question Cannot Be Sidestepped
This is where the methodological concerns become most acute. In northern Nigeria, a substantial number of children associated with the Almajiri system may not live in conventional family households in the way a standard household enumeration expects. Some children may live with Qur’anic teachers; sleep in religious school compounds; live in informal shelters; move between communities; spend extended periods away from their parents; live in overcrowded urban settlements; sleep in makeshift structures; spend significant time on the streets; move seasonally; or be attached to informal religious learning environments.
A household survey that begins with “Who lives in this household?” may not necessarily find every child who is part of that household’s extended family. The question becomes even more complicated when children are sent from their home communities to another state to study under an Islamic teacher.
Which household owns the child statistically? The biological family’s household? The teacher’s compound? The Qur’anic school? The temporary shelter? The community where the child currently resides? Or nowhere, if the child is living in a structure that is not classified as a conventional household?
This is a classic coverage problem. A methodology can be statistically rigorous and still produce an undercount if its sampling frame does not cover the population it is supposed to measure. The concern is not that the Kaduna model is necessarily flawed; the concern is that we do not know whether it adequately captures these populations because the methodology has not been made public.
Minister Alausa has acknowledged that preliminary findings from ongoing nationwide sampling suggest Nigeria’s out-of-school population may currently be below eight million. He has also challenged widely quoted figures by UNICEF and UNESCO, noting that while UNICEF estimated Kaduna had 1.8 million out-of-school children, the state-wide mapping exercise identified about 700,000. (It is worth noting the discrepancy even within the government’s own statements: the minister has cited both “fewer than 200,000” and “700,000” for Kaduna at different times. This inconsistency itself warrants explanation.)
If the new methodology systematically produces a substantially lower national out-of-school children figure, the political consequences will be significant. The government could then legitimately claim that “Nigeria does not have 20 million out-of-school children. The actual figure is significantly lower.”
That may turn out to be true. But it may also be statistically misleading if the reduction is achieved through a narrower definition of out-of-school children; exclusion of children in informal education; exclusion of children outside conventional households; exclusion of mobile populations; exclusion of children in religious compounds; exclusion of children in unenumerated settlements; counting enrolment rather than actual attendance; or counting nominal participation rather than meaningful participation.
The Dangerous Possibility: “Out of School” Could Be Redefined
This is perhaps the most politically sensitive issue of all. Nigeria has a complex education ecosystem. A child may be attending a conventional public school; attending a private school; attending a faith-based school; attending a Qur’anic school; attending an Almajiri education programme; receiving informal religious instruction; attending a non-formal education centre; attending school irregularly; registered but chronically absent; enrolled but not learning; or attending an unrecognised institution.
The statistical classification matters enormously. UNESCO’s Institute for Statistics makes an important distinction: administrative data generally classify participation according to formal enrolment, while household surveys commonly measure attendance. A child who did not attend formal primary or secondary education during the relevant reference period may be classified as out of school even if the child receives some other form of education.
Nigeria must therefore be extremely careful not to create a statistical illusion whereby “the child is receiving some form of instruction” is equated with “the child is not out of school.” That is not necessarily the international definition of formal school participation. If a child spends years learning Qur’anic texts but has no access to basic literacy, numeracy, science, civic education, digital literacy and other components of foundational education, the child may still be educationally excluded from the formal basic education system.
This does not diminish the value of Islamic education. It simply recognises that religious education and formal basic education are not necessarily interchangeable statistical categories. The appropriate solution is therefore not to erase such children from out-of-school children statistics. It is to count them accurately and classify them appropriately. A child receiving Qur’anic instruction but no formal education should be counted as out of school if the definition of “school” refers to formal basic education. If the definition is broadened to include any form of learning, then the entire concept of “out of school” becomes meaningless.
The “Household” Problem: Children Who Do Not Live in Households
A national out-of-school children methodology must go beyond conventional household enumeration. Nigeria has populations that may be missed by standard household surveys, including street-connected children; internally displaced children; nomadic children; children in refugee-like situations; children living in institutional settings; children in informal settlements; children living in religious compounds; children in correctional facilities; children separated from their parents; children who migrate seasonally; children whose parents are absent; and children in conflict-affected communities.
The Federal Government has acknowledged some of these challenges. A six-phase survey, which had its pilot run in July 2025, was designed to extend to roaming Almajiri children across Nigeria’s six geopolitical zones. The project was also intended to profile nomadic children, geolocate their communities, and determine where learning centres or teachers are needed. This is encouraging. But it also raises a question: if special enumeration is needed for nomadic and Almajiri populations, does that mean the standard household mapping model does not adequately capture them?
This is precisely why the Kaduna model must be subjected to a coverage audit before being declared a national model. The question should be: does the methodology find households, or does it find children? Those are not necessarily the same thing. A methodology can be excellent at mapping households and still fail to identify children who do not fit neatly into the conventional household structure.
Why International Practice Is Different
The international approach is generally more sophisticated than simply sending enumerators from household to household and adding up the children they find. UNESCO’s current methodology combines administrative enrolment records, population estimates, censuses, household surveys, Demographic and Health Survey data, Multiple Indicator Cluster Survey data, country-specific surveys, and statistical modelling. The objective is to produce coherent estimates while recognising the limitations of individual data sources.
The Global Initiative on Out-of-School Children, involving UNESCO’s Institute for Statistics and UNICEF, has similarly promoted methodologies that combine administrative and household data and investigate the characteristics and barriers affecting excluded children. This initiative has involved more than 70 countries worldwide in developing and applying innovative methodologies to assess and use data from administrative records and household surveys to better identify how many children are excluded from education.
This is a crucial lesson for Nigeria: the international standard is not “household mapping versus estimates.” It is administrative data plus population data plus household data plus census data plus specialised enumeration plus statistical reconciliation. That is a much stronger approach.
Countries with complex populations generally do not rely on one source. The methodological principle is broadly consistent: use multiple sources to identify gaps and then reconcile them. A country may use school administrative records to determine enrolment; population census data to determine the school-age population; household surveys to measure actual attendance; specialised surveys to identify vulnerable populations; local education registers to locate children; and community-level enumeration to reach marginalised groups.
The reason is simple. Administrative data tell you who is registered. Household data tell you who reports attending. Census data tell you how many children exist in the relevant age groups. Specialised enumeration tells you about populations that conventional household surveys may miss. Each answers a different question. Nigeria should therefore be cautious about adopting a single model as a national “truth machine.”
The 585,000 to Under 200,000 Reduction Requires a Forensic Audit
Let us examine the arithmetic carefully. Old figure: approximately 585,000. New figure: fewer than 200,000. Approximate reduction: more than 385,000 children. Percentage reduction: more than 65 per cent. That is an extraordinary statistical difference. It should therefore trigger a rigorous methodological reconciliation.
The government should publish a detailed comparison table showing, for both the old estimate and the new mapping, the age range used; the definition of out-of-school child; whether primary only or secondary included; whether formal education only or other forms included; whether Qur’anic education was included; whether Almajiri children were included; whether street children were included; whether internally displaced persons were included; whether nomadic children were included; whether children in informal shelters were included; whether children absent from home were included; whether children enrolled but not attending were included; the verification method used; and whether independent validation was conducted.
Until this table can be completed, comparing 585,000 with under 200,000 is potentially comparing apples with oranges. The difference may reflect a genuine improvement in data quality. Or it may reflect a narrowing of the definition of who counts as “out of school.” The public deserves to know which it is.
Is the Federal Government Trying to “Debunk” the 20 Million Figure?
This is where we need to be intellectually disciplined. There is currently insufficient evidence to conclude that the Federal Government is deliberately trying to manipulate the out-of-school children figures downward. That would be an allegation requiring evidence. However, there is a legitimate reason for public concern.
Minister Alausa has publicly challenged widely quoted figures by UNICEF and UNESCO. He has suggested that preliminary findings indicate Nigeria’s out-of-school population may be below eight million. He has also claimed that the Federal Government has returned over one million out-of-school children to classrooms within the last 30 months. These statements, combined with the dramatic reduction in Kaduna’s figures, naturally invite scrutiny.
The question should not be “Is the government trying to debunk the 20 million figure?” The better question is “Will the new methodology produce a more accurate figure, or merely a smaller figure?” Those are two completely different outcomes. A more accurate figure would be a genuine achievement. A merely smaller figure would be a statistical illusion—and a dangerous one at that.
The Political Economy of Statistics
There is another dimension that must be acknowledged. Statistics are not politically neutral. The number of out-of-school children affects budget allocations; donor funding; international rankings; Sustainable Development Goal reporting; education intervention programmes; school construction; teacher recruitment; social investment programmes; and state and federal priorities.
A lower out-of-school children figure could have political advantages. It could suggest that previous governments exaggerated the problem; that current reforms are working; that Nigeria’s education crisis is smaller than reported; and that government interventions are succeeding. But a higher figure could also attract more resources. Therefore, the integrity of the data depends heavily on methodological independence and transparency.
This is why the involvement of the National Bureau of Statistics is potentially important. But NBS involvement alone is not enough. The process should be independently scrutinised by multiple stakeholders, including the Federal Ministry of Education, state ministries of education, the Universal Basic Education Commission, UNICEF, UNESCO’s Institute for Statistics, the World Bank, independent Nigerian statisticians, universities and research institutions.
What the Federal Government Should Do Before Rolling Out the Model
My strongest recommendation is that Nigeria should not simply adopt Kaduna’s methodology. It should use Kaduna as one component of a national multi-source system. I would recommend a National Out-of-School Children Data Reconciliation Framework with five layers.
The first layer is household enumeration: mapping children living in conventional households. The second layer is school administrative records: comparing household-reported school attendance with actual school enrolment records. The third layer is population denominator: using reliable demographic data to establish the total school-age population.
The fourth layer is special population enumeration: conducting dedicated enumeration of Almajiri children, Qur’anic schools, nomadic communities, internally displaced persons camps, street-connected children, informal settlements, institutionalised children and conflict-affected communities. The fifth layer is independent validation: engaging the National Bureau of Statistics, the Federal Ministry of Education, state ministries of education, UBEC, UNICEF, UNESCO’s Institute for Statistics, the World Bank, independent Nigerian statisticians, universities and research institutions.
The final figure should then be produced through triangulation—not by simply choosing whichever number is politically or administratively convenient. This is precisely the approach that UNESCO and UNICEF have promoted through the Global Initiative on Out-of-School Children. Nigeria would do well to follow this international best practice rather than attempting to reinvent the wheel with a single-source methodology.
A Better Set of Questions for the Federal Government
Instead of asking “How many out-of-school children are there?” Nigeria should ask four questions simultaneously. First, how many children are not enrolled? Second, how many children are enrolled but not attending? Third, how many children attend non-formal or religious education but lack access to foundational formal education? Fourth, how many children are completely excluded from any meaningful education?
These four groups may overlap, but they should not be collapsed into one simplistic number. Nigeria’s education crisis is more complex than a single headline figure. The government should be able to disaggregate its data along multiple dimensions: age, gender, geographic location, enrolment status, attendance status, type of education received (if any), and barriers to participation.
The ultimate test of the Kaduna model should be this: if an Almajiri boy leaves his parents in Katsina, travels to Kaduna, sleeps in an informal Qur’anic school compound and receives Qur’anic instruction but no formal basic education, will the system find him? If the answer is yes, the methodology is promising. If the answer is no, then the methodology is incomplete.
Similarly, if a child lives in an informal settlement that is not recognised as a conventional household, will the child be counted? If a child is displaced by insecurity, will the child be counted? If a child is enrolled on paper but has not attended school for six months, will the child be considered “in school”? If a child attends a Qur’anic school but has no access to formal basic education, how is that child classified? These questions are more important than whether the final number is 20 million, 10 million, 5 million or 2 million.
My Forensic Verdict
My assessment is that the Federal Government’s initiative is potentially a very good reform, but the current public presentation is methodologically incomplete. The decision to improve out-of-school children data is strongly justified. The use of household mapping is methodologically useful. The adoption of household mapping as a national framework is potentially useful but requires rigorous validation. The claim that Kaduna’s under-200,000 figure is the “actual” number is not yet sufficiently demonstrated from the information publicly available.
The comparison with the previous 585,000 estimate requires methodological reconciliation before the two figures can be directly compared. The concern that vulnerable and mobile populations may be missed is legitimate and technically significant. The concern about Almajiri children being undercounted is a critical methodological issue that must be explicitly addressed. The suspicion that the exercise is designed to discredit the 20-million estimate is possible as a political interpretation but is not established by evidence currently available. The need for transparency is absolutely essential.
The Bottom Line
The Federal Government should be applauded for wanting to replace weak data with better data. But there is a fundamental principle in public policy: you cannot improve statistics merely by changing the method of counting. You improve statistics by making the method more comprehensive, more transparent, more independently verifiable and more consistent with internationally recognised definitions.
The real success of this initiative will not be measured by whether Nigeria’s out-of-school children figure falls from 20 million to 10 million—or from 585,000 to 200,000 in Kaduna. It will be measured by whether every child who should be in basic education but is not actually receiving it can be found, counted and reached.
And this is why I would urge the Federal Government to publish the complete Kaduna methodology before using it as the national template. Show Nigerians the questionnaire. Show us the sampling frame. Show us the age definition. Show us how Almajiri children were counted. Show us how children living in Qur’anic school compounds were counted. Show us how internally displaced persons, nomadic populations, street-connected children and children in informal settlements were covered. Show us how the 585,000 estimate was calculated and why the new methodology produced fewer than 200,000. Show us the independent validation.
Then Nigerians can make an informed judgment. Because the most dangerous outcome is not having a high number of out-of-school children. The most dangerous outcome is having a low number that gives us false confidence while millions of excluded children remain invisible. In the final analysis, Nigeria does not need a smaller out-of-school children number. Nigeria needs a truer out-of-school children number. And the difference between the two is methodology.
Before Nigeria adopts the Kaduna approach wholesale, the Federal Ministry of Education and the National Bureau of Statistics should release a technical methodology paper and subject it to independent peer review by the National Bureau of Statistics, the Universal Basic Education Commission, UNESCO’s Institute for Statistics, UNICEF, relevant Nigerian universities and independent statisticians. That would turn what is currently a potentially controversial statistical exercise into a credible national education-data reform. The children of Nigeria deserve nothing less.