Review of Research on Basic Education Provision in Nigeria

Description

1.1 Methodology
A vast amount of education sector research has been undertaken in Nigeria, both by
local and state ministries as well as by International Development Partners and Non-
Governmental Organization‘s (NGO). The circulation and availability of this
information is unfortunately quite limited, many are available only in paper format,
and dissemination of results is impeded by organizational, printing, financial, and
bureaucratic issues (Theobald & Sanni, 2007). Many relevant studies were identified
but could not be accessed as they were published in Nigerian journals1 which are not
widely available outside of Nigeria. For these sources no access could be found
online and the references do not appear on standard academic search engines such as
ERIC, Eldis, or Google Scholar. Ango et al. (2003) conclude from their review of 60
reports that the literature is ̳dominated by speculative studies‘ with empirical research
largely coming out of higher education institutions and typically undertaken by
individuals rather than collaboratively (p. 7). The authors also question the quality of
the research reviewed and call for future ̳rigorous, vigorous, collaborative and
national action-research oriented studies‘ (p 8).
This review included analysis of the Nigeria Demographic Health Survey (NDHS)
datasets for 1990, 1999 and 2003. From the data it was possible to evaluate access
trends and the progress made towards expanding access for all children. Finally, the
review concludes with identification of knowledge gaps for which new research is
required. The identified research areas in some cases takes into account some
evidence from the review of the grey literature on basic education in Nigeria by
Bakari (2009).
2 Educational Access
This section re-examines educational access data in Nigeria. According to Theobald et
al., (2007) latest data suggests that there are about 3.6m primary-age children in
Nigeria who are not enrolled in school (of which 2.6m are girls). This rises to 7.85m
at junior secondary level, equally divided between girls and boys. Indeed, Nigeria is
estimated to have the largest number of un-enrolled children in the world including
children who have never enrolled as well as a high number of drop outs (EFA GMR,
2007). Most analysts point out that the challenges around educational access and
meeting MDGs and EFA goals in Nigeria are vast.
The section starts by re-analyzing educational access data based on household surveys
collected in three time periods to assess changes in access to basic education in
Nigeria. Examining access trends allows judgments to be made about whether
progress is being made, and what remains as challenge if Nigeria is to meet its EFA
goals. Finally the section reviews mainly empirical literature on educational access to
identify gaps that might be addressed in future research.
1 Journal of the Curriculum Organisation of Nigeria, The Nigerian Teacher Today, The Nigerian
Journal of Curriculum Studies, Nigerian Language Studies, Journal of Science Teachers Association of
Nigeria, etc.

7
This section focuses on the following questions:
 What is the extent of variations in access to primary years schooling for children in
different states, within states, and across income groups?
 What is the evidence about the barriers to educational access at primary and junior
secondary level?
 What evidence exists about policies that impact on pupils‘ participation in
education?
 What is the evidence about access for marginalised and disadvantaged children?
2.1 Analysing access trends
Data for this part of the analysis was based on the 1990, 1999 and 2003 Nigeria
Demographic Health Surveys (NDHS). All Nigeria DHS surveys are national
samples, covering all parts of the country. The 1990 NDHS was sampled on the basis
of the 1973 census enumeration areas. Within each state, enumeration areas were
stratified into three sectors (urban, semi-urban and rural) and were selected with equal
probability. A sample of about 10,000 households in 299 enumerated areas was
identified with a twofold oversampling of the urban stratum. Thus, the 1990 NDHS is
a weighted sample, nationally representative of the population and also representative
of 4 regions (Northeast, Northwest, Southeast and Southwest).
The 1999 and 2003 NDHS sampling frames used the enumerated areas for which the
country was delineated for the 1991 census, hence the geographical areas changed
from the 1990 NDHS. The sample was stratified into urban and rural areas and a two-
stage sampling procedure was used to obtain eligible women. Urban areas were
oversampled in order to obtain reasonable urban estimates. The surveys produced
nationally representative statistics, and also regional (Central, Northeast, Northwest,
Southeast and Southwest) as well as urban and rural.
2.1.1 Sample Selection
The analysis attempted to focus exclusively on 5 states (i.e. Kaduna, Kano, Kwara,
Jigawa and Lagos), for aggregate information on children‘ schooling. It is important
to highlight that the administrative division of the country changed between 1990 and
1999, so in the analysis it is sometimes difficult to make comparisons in trends
between the states of Kano, Jigawa and Kwara. This is because in 1990 the state of
Kwara comprised part of the state of Niger while in 1999 it comprised part of the state
of Kogi. Similarly, the state of Jigawa did not exist as an administrative locality in
the political division of Nigeria in 1990. Instead, it was part of the state of Kano.
With this limitation, trends in schooling can only be compared for two of these states
(Lagos and Kaduna) from 1990 to 2003 and for all five states between 1999 and 2003.
For each of the NDHS‘s analysis focused on children aged 6 to 14 years, and those
who were of the school age for primary and secondary schooling in Nigeria. Table 1
shows the number of children aged 6 to 14 in each of the five states in 1990, 1999 and
2003. Differences in total number of children selected reflect differences in sample
sizes for each of the surveys. For instance, sample size for the 1990 was over 47,000
individuals; while for 1999 and 2003 was 37,624 and 35,173 individuals, respectively.

Table 1: Number of children aged 6 to 14 in each of the five states
Male Female Total Male Female Total Male Female Total
Kaduna 271 265 536 201 211 412 180 218 398
Kano 634 591 1225 313 309 622 235 213 448
Kwara 160 156 316 84 78 162 81 72 153
Lagos 792 795 1587 177 148 325 132 131 263
Jigawa n.a. n.a. n.a. 110 101 211 80 95 175
1990 1999 2003
Source: Nigeria DHS 1990, 1999 and 2003. Note: Administrative divisions, i.e. states, are not
comparable between 1990 and the rest of the years for Kwara and Kano.
Another limitation in the analysis is the small number of observations in some of
these states. This arises because the DHS sample was designed to be representative at
national and regional levels, but not at administrative or state level.2 Because of this,
and our interest in investigating different indicators of schooling such as net
enrolment rates, age-grade dispersions, among others, most of the analysis was
undertaken at a regional level.
As pointed out earlier, because regions used in the DHS contain different states in
1990, 1999 and 2003, state level variables were used to aggregate data and obtain
homogenous regions which give comparable statistics over time. The 1990 NDHS
contained 4 regions (Northeast, Northwest, Southeast and Southwest). The analysis
used states in the 1999 and 2003 that were part of the regions in 1990 to recreate, as
much as possible, the 1990 regional classification.3 States in 1999 and 2003 included
in the 4 regions from 1990 were: North East: Taraba, Adamawa, Gombe, Borno,
Bauchi, Yobe, Kano, Jigawa, Plateau and Nasarawa. North West: Kebbi, Kaduna,
Katsina, Zamfara, Sokoto, FCT, Niger, Kwara and Kogi. South East: Ebonyi,
Anambra, Enugu, Abia, Imo, Cross River, Akwa Ibom, Rivers and Benue. South
West: Lagos, Oyo, Osun, Ogun, Ekiti, Ondo, Delta, Edo and Bayelsa.
Table 2 shows the number of children aged 6 to 14 in each of the 4 regions.
2 According to DHS, only aggregate information may be representative at the state level. For instance,
we may be able to estimate the proportion of children aged 6 to 14 not in school in these states, but not
this proportion for a particular age group.
3 All regions in 1999 and 2003 can be matched to the 4 main regions in 1990 except for the state of
Kogi, which did not exist in 1990. In 1991, parts of the states of Kwara in the northwest and of Benue
in the southeast were divided to form Kogi. For this report, all individuals in Kogi are included as part
of the northwest region.
Table 2: Number of children aged 6 to 14 by regions
Male Female Total Male Female Total Male Female Total
Southeast 1786 1800 3586 1278 1165 2446 976 957 1933
Southwest 1866 1826 3692 1109 944 2054 668 637 1305
Northwest 1026 1015 2041 1068 1023 2091 972 959 1931
Northeast 1447 1439 2886 1255 1209 2464 1342 1323 2665
1990 1999 2003
Source: Nigeria DHS 1990, 1999 and 2003. Note: Regions were homogenised according to the 4
regions in 1990.
2.2 Out of school children
To identify out of school children, data based on the questions ―has the child ever
been to school?‖ and ―is the child still in school?‖ was analysed. Table 3 shows the
proportion of out of school children in the states of Kaduna, Kano, Kwara, Lagos and
Jigawa. In general, there is a huge variation with respect to children not in school
between these five states, although there has been some progress towards reducing the
number of out of school children, especially between 1999 and 2003. The proportion
of children out of school in 1990 was 9% in Lagos and as high as 73% in Kano. The
difference in the proportion of children out of school between Lagos and Kwara and
the rest of the areas continues to be significant in 1999 and 2003. For instance, in
Lagos in 2003 as little as 2% of boys and 5% of girls are not in school, compared with
60% of boys and 89% of girls in Jigawa.
As noted there has been some progress in reducing the number of out of school
children in 5 states, although the proportions are still large (see table 3). Between
1999 and 2003 the proportion of children out of school declined across all five states,
with some States reaching significant levels. For instance, in Jigawa where nearly all
children were out of school in 1999, 25% are now enrolled in school (although a large
gender gap persists).
Table 3: Out of school children, aged 6 to 14, in five states by gender
Male Female Total Male Female Total Male Female Total
Kaduna 59.8% 61.3% 60.5% 57.0% 65.2% 61.2% 27.9% 36.2% 32.5%
Kano 66.7% 80.9% 73.5% 56.9% 66.8% 61.8% 39.4% 49.7% 44.1%
Kwara 29.1% 34.1% 31.4% 22.9% 24.6% 23.7% 6.2% 6.7% 6.4%
Lagos 7.8% 10.1% 9.0% 8.5% 10.2% 9.3% 2.2% 4.6% 3.4%
Jigawa n.a. n.a. n.a. 92.5% 96.5% 94.4% 60.2% 89.1% 75.8%
1990 1999 2003
Source: Nigeria DHS 1990, 1999 and 2003. Note: Administrative divisions, i.e. states, are not
comparable between 1990 and the rest of the years for Kwara and Kano.
Table 4 shows the proportion of primary age scold children out of school across all
states, grouped in 4 regions. A sustained decline in the proportion of children out of
school is notable in all regions excepting the southwest region (inclusive of Lagos
state). Again, regional differences are large, with the southern regions demonstrating a
much smaller proportion of children out of school (as little as 9%) as compared to
northern regions where nearly half of children are out of school. These trends suggest
that although the out-of-school population in Nigeria is still high, there appears to be

some progress in reducing this population. Also, it is more helpful to think of the
problem of out of school children in terms of geographical location and the context-
specific factors which contribute to the variations in the population of out-of-school
children.
Table 4: Out of school children, aged 6 to 14, by regions and by gender
Male Female Total Male Female Total Male Female Total
Southeast 26.5% 32.8% 29.6% 15.7% 19.8% 17.7% 9.5% 10.0% 9.7%
Southwest 12.8% 13.6% 13.2% 21.3% 19.8% 20.6% 8.3% 9.1% 8.7%
Northwest 71.2% 81.4% 76.1% 50.0% 60.2% 54.9% 37.8% 53.1% 45.3%
Northeast 63.6% 74.1% 68.8% 61.5% 65.4% 63.4% 43.5% 54.4% 48.8%
1990 1999 2003
Source: Nigeria DHS 1990, 1999 and 2003. Note: Regions were homogenised according to the 4
regions in 1990.
2.3 In school: overage children
Overage enrolment has consequences on timely progression and completion of basic
education. It also has pedagogical implications as teachers trained to a mono-grade
curriculum are ill prepared to teach classes which are essentially multi-grade in
composition. This point is important because documented poor learner achievement
could partly be explained by teachers‘ inability to provide suitable instruction for a
wide age range at different levels of cognitive development.
For the analysis, ̳overage‘ was defined as children who were 3 or more years older
than the age appropriate for their enrolled school grade. For example, in Nigeria, the
recommended age for primary grade 1 is 6 years; children aged 9 and above by this
definition would be considered overage for their grade.
For the five states, large differences exist in the proportion of overage children (see
table 5). In 1990, Lagos had the lowest proportion of overage children, with 4.3%; the
largest proportions were found in Kano and Kaduna, with 15% and 17% overage
children respectively. In these states it may be that children were either starting school
late or repeating more frequently.
Between 1999 and 2003, there is a drastic increase in the proportion of overage
children. For instance, in 1999 in Lagos State, only 0.7 of all girls attending school
were overage. By 2003, this proportion increased to 2%. Similarly for boys in Lagos
State, the proportion of overage students increased from 2% to 7% between1999 and
2003. This is an increase of more than triple the number of children overage.4
4 Caution should be taken with the proportion of children overage for Jigawa. As shown in table 3,
almost all children in this state were out of school in 1999. It could be that the very small proportion of
girls was in an age-appropriate grade. By 2003, there are more children attending school and we
estimate a large proportion of children overage.

Table 5: School attendance, overage children, aged 6 to 14, in five states by gender
Male Female Total Male Female Total Male Female Total
Kaduna 17.3% 16.2% 16.8% 16.9% 13.0% 15.2% 24.2% 25.5% 24.9%
Kano 14.8% 14.5% 14.7% 11.7% 7.9% 10.1% 24.3% 24.5% 24.4%
Kwara 7.7% 10.0% 8.8% 3.1% 0.0% 1.6% 8.0% 4.3% 6.3%
Lagos 4.6% 4.0% 4.3% 1.9% 0.7% 1.4% 7.3% 2.0% 4.7%
Jigawa n.a. n.a. n.a. 12.6% 0.0% 8.6% 33.4% 19.0% 29.9%
1990 1999 2003
Source: Nigeria DHS 1990, 1999 and 2003. Note: Administrative divisions, i.e. states, are not
comparable between 1990 and the rest of the years for Kwara and Kano.
Regional variation in the proportion of children overage is significant, with the
southeast having more than three times the proportion of children overage in 1990
than the southwest. Between 1999 and 2003 there is an increase in the proportion of
children overage nationwide, across all regions, for both boys and girls. In the
northern regions of the country, the proportion of children overage more than double
between 1999 and 2003; from 11% to 22% in the northeast and from 9% to 25% in
the northwest. This is a significant trend which suggests the existence of factors
operating in these states which either increases the chances of children enrolling late
or repeating grades.
Table 6: School attendance, overage children, aged 6 to 14, by regions and gender
Male Female Total Male Female Total Male Female Total
Southeast 24.0% 26.7% 33.9% 14.8% 14.0% 14.4% 17.6% 18.8% 18.2%
Southwest 10.1% 9.2% 10.7% 7.2% 4.4% 5.9% 12.6% 15.0% 13.8%
Northwest 17.6% 11.5% 17.9% 10.5% 7.7% 9.3% 22.9% 27.3% 24.8%
Northeast 17.3% 17.5% 21.0% 12.2% 10.2% 11.3% 21.5% 22.9% 22.1%
1990 1999 2003
Source: Nigeria DHS 1990, 1999 and 2003. Note: Regions were homogenised according to the 4
regions in 1990.
2.4 School Attendance by Wealth Groups
Given the small number of observations that we have for the five states, the following
analysis is only done at the regional level. For each region, we investigated the
proportion of children attending school according to household wealth. Household
wealth was obtained following the methodology developed by Filmer and Pritchett
(1997). DHS data contain detailed information about the characteristics of the
household dwelling and ownership of various assets. Filmer and Pritchett (1997)
suggest using information from more than twenty of these assets variables and
principal component analysis to obtain a total score which represents the wealth index
for the DHS sample. From this index, it was possible to generate wealth quintiles,
where households are grouped from the lowest to the highest wealth quintiles.5
Figure 1 to figure 3 details the proportion of children aged 4 to 14 attending school in
1990, 1999 and 2003 by welfare quintile. For all years and in all four regions there is
a positive gradient with a higher proportion of children living in wealthier families
5 Wealth indices were already created for Nigeria 1990 and 2003 DHS. For Nigeria 1999 the wealth
index from raw information on assets and principal component analysis was obtained. We remain
cautious about comparability of wealth over time, which is an issue that this report has not investigated
in detail.

attending school. However, there is considerable variation with respect to the
steepness of this gradient between regions. For instance, in 1990, the steepest gradient
was for the northwest region and the mildest gradient was for the southwest region;
meaning variations in proportion of children attending school by wealth status was
higher in the northwest than in the southwest.
It is worth pointing out that across all regions and all wealth quintiles there has been
an increase in the proportion of children attending school. However, the gradient of
school participation from the richest to the poorest does not show large variations
across all regions over time. Since one cannot be sure about the reliability of the
wealth index to make comparisons over time, no further inferences about possible
changes in relative inequalities in access to education are made. Nevertheless, one can
say that southern states demonstrate greater equity as compared to northern states
when it comes to differences in attendance by welfare quintile; i.e. disparities in
attendance rates between the poorest and richest are less marked in the south as it is in
the north.
Figure 1: School attendance, children aged 4 to 14, by wealth quintiles and region, in
1990
0 .2 .4 .6 .8 10 .2 .4 .6 .8 1
Lowest 2 3 4 Highest Lowest 2 3 4 Highest
Lowest 2 3 4 Highest Lowest 2 3 4 Highest
southeast southwest
northwest northeast
Graphs by region