Keywords: Out-of-School Children; educational policy; mixed-methods research; Guinea-Bissau
July 2020
Rui Da Silva, Researcher at the Centre of African Studies of the University of Porto. PhD in Education
Jeffrey Marshall, International Development Consultant and Educationalist, PhD in Education
Mathilde Nicolai, Education Adviser at Cambridge Education. Master’s in sociology and economics
In Guinea-Bissau, most children still do not attend school and there are many reasons for this. In order to understand the context in question, the present study proposes a mapping of the situation through the use of mixed methods.
Context
The Global Out-of-School Children Initiative, a partnership between UNICEF and the UNESCO Institute for Statistics (UIS), supports countries in their analysis and monitoring of out-of-school children
Guinea-Bissau is a good example of a poor country where policy support to address the OOSC challenge is urgently needed. Significant progress has been made in recent decades towards meeting universal primary enrolment and completion. But the out-of-school children problem persists, along with low completion rates and early drop-out. The country faces sizable resource constraints, with a GDP per capita level (current US$777 in 2018) that is among the lowest in the world (World Bank, 2018).
Guinea-Bissau has also experienced an extended period of political unrest, military coups and institutional instability. The recent more stable political context has created space for addressing entrenched problems in social sectors, and the Ministry of Education has identified out-of-school children as a policy priority in the most recent sector plan (2017-2025).
The education sector in Guinea-Bissau faces a number of serious challenges. Less than two percent of GDP is allocated to education spending, and the education budget is almost entirely absorbed by salaries (Merchant et al. 2018). The remainder of the budget has low levels of actual spending execution, and the sector is heavily reliant on donor support (Ministère de l’Education Nationale. 2015b).
Methodology
This study comes from a larger report on out-of-school children that was commissioned by UNICEF and the Ministry of Education (author citation, 2018). It provides an analysis of the underlying causes based on quantitative data and existing research on school attendance. The data for the quantitative work comes mainly from the Multiple Indicator Cluster Survey (MICS), for years 2000, 2006, 2010 and 2014, the EMIS (2012-13 and 2014-15 data sets) and poverty assessment survey[1] data (2002 and 2012). The descriptive analysis is augmented with multivariate statistical analysis.
The study’s main contribution to OOSC research in Guinea-Bissau is an OOSC-focused qualitative survey with over 150 individual and focus group interviews with children, parents, community and religious leaders, and school staff [2], the latter incorporating the principles of the djumbai approach (Klute, Embaló, and Embaló 2006). The mixed-methods analysis allows for a more complete triangulation across data sources, with a focus on barriers and key bottlenecks that help explain observed patterns in OOSC.
Overview of OOSC in Guinea-Bissau
Despite the resource constraints the education system has been expanding.
Figure 1. School Attendance Profiles for Primary Aged Children, 2000-2014 MICS
Data source: MICS 2000, 2006, 2010 and 2014
Total primary enrolment (EB1 and EB2) has more than doubled between 2000 and 2016, and there has been a dramatic reduction in educational exclusion (Figure 1). Dropout rates have also declined over these age ranges, and in 2014 only about seven percent of children had left school by the age of 14. Gender ratios have steadily improved.
However, the data highlights serious systemic issues related to school attendance. The first is late entry into primary school, (23 percent of children aged 6-11 have not entered school (at any level). Overage enrolment is a key feature of the system: 30 percent of children in primary grades 1-3 (EB1) are 12 years or older, 15 percent of children in primary grades 4-6 (EB2) are 18 years or older, and roughly half of the student population in lower secondary is 18 years or older.
The total OOSC has been cut in half since 2000, despite considerable population growth in this period.
The multivariate analysis highlights covariates of ever attended school, current attendance and grade attainment using the 2014 MICS data. The analysis reveals that Socio-Economic Status and family background (parental education, family size, child relationship to head of household: child, grandchild, adopted, etc.) are consistently significant predictors of attendance and attainment. For example, young people from the wealthiest households (SES quintile 5) have accumulated more than four total years of educational attainment than their quintile 1 counterparts. Rural children complete about three years less education than urban children, even when controlling for differences in SES and other factors.
Table 1 summarizes the covariates of ever attended school, current attendance and grade attainment using the 2014 MICS data.
Table 1. Summary of Covariates of Current School Attendance, 2014 MICS
| Ever Attended School=1 | Current Attendance=1 | Grade Attainment | ||||
|---|---|---|---|---|---|---|
| (age 6-17) | (age 6-22) | (age 15-22) | (age 6-22) | |||
| Independent Variable | Female (1) | Male (2) | National | Female | Male | National (6) |
| -3 | -4 | -5 | ||||
| Female | ---- | ---- | -0.042** | ---- | ---- | -1.19** |
| (-5.49) | (-7.28) | |||||
| Age | 0.024** | 0.029** | 0.009** | -0.007 | -0.056** | 0.61** |
| (10.86) | (15.42) | (5.75) | (-1.28) | (-7.11) | (16.86) | |
| Relation to household head (reference= child): | ||||||
| Grandchild | -0.016 | -0.001 | 0.009 | 0.038 | 0.013 | -0.05 |
| (-0.79) | (-0.06) | (0.61) | (1.02) | (0.28) | (-0.17) | |
| Cousin | -0.033+ | -0.021 | -0.018 | -0.029 | 0.052+ | -0.61** |
| (-1.89) | (-1.21) | (-1.54) | (-1.25) | (1.75) | (-2.62) | |
| Adopted | 0.032 | -0.021 | 0.023 | 0.029 | 0.163* | 0.30 |
| (1.04) | (-0.66) | (0.74) | (0.49) | (1.93) | (0.47) | |
| Other without relation | -0.078 | -0.89 | -0.217** | -0.343** | -0.122 | -4.23** |
| (1.08) | (-1.12) | (-4.34) | (-3.41) | (-1.25) | (-3.85) | |
| Other relation | -0.095** | -0.081** | -0.138** | -0.030 | 0.005 | -3.09** |
| (-4.63) | (-2.77) | (-9.03) | (-1.22) | (0.14) | (-10.37) | |
| Household head= | 0.032* | -0.011 | -0.004 | -0.021 | -0.021 | 0.22 |
| Female | (1.97) | (-0.67) | (-0.31) | (-1.08) | (-0.68) | (0.90) |
| Household head religion (reference= Catholic): | ||||||
| Evangelical | -0.030 | -0.001 | 0.008 | -0.006 | 0.013 | -0.21 |
| (-0.87) | (-0.06) | (0.34) | (-0.17) | (0.28) | (-0.39) | |
| Muslim | -0.130** | -0.101** | -0.107** | -0.071** | -0.020 | -3.18** |
| (-5.06) | (-3.30) | (-5.89) | (-2.52) | (-0.46) | (-7.63) | |
| Anemista | -0.065* | -0.052* | -0.051** | -0.033 | -0.003 | -1.47** |
| (-2.42) | (-2.08) | (-2.72) | (-1.13) | (-0.07) | (-3.75) | |
| Mother education (reference=none): | ||||||
| Primary | 0.062** | 0.048** | 0.065** | 0.087+ | -0.116 | 1.27** |
| (3.87) | (3.73) | (4.32) | (1.86) | (-1.52) | (3.82) | |
| Secondary or more | 0.049+ | 0.053+ | 0.060* | 0.046 | 0.01 | 0.49 |
| (1.85) | (1.85) | (2.28) | (0.48) | (0.00) | (0.91) | |
| Mother not in home | -0.053** | -0.002 | -0.034* | 0.010 | -0.102* | -0.16 |
| (-2.74) | (-0.07) | (-2.14) | (0.35) | (-1.98) | (-0.48) | |
| Father education (reference= none): | ||||||
| Primary | 0.064** | 0.054** | 0.064** | 0.097** | 0.161** | 1.40** |
| (4.13) | (4.57) | (4.46) | (2.69) | (2.78) | (4.44) | |
| Secondary or more | 0.092** | 0.059** | 0.102** | 0.074 | 0.01 | 1.91** |
| (3.35) | (2.62) | (4.78) | (1.47) | (0.00) | (4.05) | |
| Father not in home | 0.021 | 0.039* | 0.028+ | 0.089** | 0.073 | 0.55 |
| (1.08) | (2.41) | (1.90) | (2.89) | (1.34) | (0.92) | |
| Family SES quintile (reference=Q1) | ||||||
| Quintile 2 | 0.050** | 0.024+ | 0.040** | 0.032 | -0.090* | 1.15** |
| (3.02) | (1.90) | (3.18) | (1.48) | (-2.53) | (4.17) | |
| Quintile 3 | 0.069** | 0.061** | 0.072** | 0.043+ | -0.002 | 2.01** |
| (3.44) | (4.07) | (4.69) | (1.83) | (-0.04) | (5.97) | |
| Quintile 4 | 0.118** | 0.090** | 0.100** | 0.073** | -0.033 | 2.90** |
| (5.37) | (3.92) | (5.51) | (2.66) | (-0.70) | (7.14) | |
| Quintile 5 (highest) | 0.124** | 0.104* | 0.141** | 0.115** | -0.024 | 4.36** |
| (3.93) | (2.46) | (5.63) | (3.13) | (-0.39) | (8.12) | |
| Number of household members: | ||||||
| 0-5 years old | -0.019** | -0.010+ | -0.030** | -0.018* | -0.017 | -0.68** |
| (-2.80) | (-1.66) | (-3.61) | (-2.32) | (-1.03) | (-6.38) | |
| 6-17 years old | 0.007 | 0.003 | 0.018** | 0.010 | 0.025 | 0.31* |
| (0.93) | (0.45) | (3.06) | (0.94) | (1.43) | (2.29) | |
| 18-59 years old | 0.011 | 0.018* | 0.010 | 0.008 | -0.002 | 0.35** |
| (1.35) | (2.24) | (1.58) | (0.88) | (-0.09) | (2.48) | |
| 60 or older | 0.005 | -0.002 | -0.005 | -0.009 | -0.020+ | 0.004 |
| (0.85) | (-0.50) | (-1.16) | (-1.12) | (-1.63) | (0.04) | |
| Language spoken in home (reference= Criolo): | ||||||
| Fula | -0.077** | -0.077** | -0.109** | -0.063** | -0.170** | -2.84** |
| (-2.77) | (-3.03) | (-6.00) | (-2.50) | (-3.74) | (-6.63) | |
| Balanta | -0.022 | -0.075** | -0.030+ | -0.014 | -0.086+ | -1.28** |
| (-0.91) | (2.79) | (-1.72) | (-0.52) | (-1.77) | (-3.17) | |
| Mandiga | -0.158** | -0.157** | -0.172** | -0.168** | -0.205** | -4.85** |
| (-6.06) | (-6.64) | (-8.07) | (-4.75) | (-3.47) | (-9.39) | |
| Manjaco | -0.005 | -0.01 | 0.021 | 0.027 | 0.050 | 0.40 |
| (-0.12) | (-0.02) | (0.64) | (0.58) | (0.81) | (0.57) | |
| Papel | -0.029 | -0.048 | -0.051+ | 0.002 | -0.132* | -1.45** |
| (-1.07) | (-1.45) | (-1.82) | (0.06) | (-1.94) | (-2.47) | |
| Family has bank account | -0.013 | 0.035 | 0.040* | 0.021 | 0.137** | 0.93** |
| (-0.48) | (1.08) | (2.20) | (0.83) | (3.05) | (2.45) | |
| Ratio of mosquito nets to HH members | 0.018** | 0.015* | 0.016** | 0.013+ | 0.013 | 0.41** |
| (2.69) | (2.46) | (3.20) | (1.67) | (1.10) | (4.00) | |
| Rural cluster | -0.087** | -0.069** | -0.110** | -0.108** | -0.180** | -2.79** |
| (-5.04) | (-3.79) | (-7.16) | (-4.81) | (-5.81) | (-8.23) | |
| FGM practiced in home | ---- | ---- | ---- | -0.055* | ---- | ---- |
| (-2.06) | ||||||
| Ever married | ---- | ---- | ---- | -0.236** | 0.011 | ---- |
| (-9.85) | (0.18) | |||||
| Ever had children | ---- | ---- | ---- | -0.188** | -0.138** | ---- |
| (-11.82) | (-3.53) | |||||
| Dependent variable mean | 0.82 | 0.83 | 0.69 | 0.66 | 0.53 | 2.94 |
| Region controls? | Yes | Yes | Yes | Yes | Yes | Yes |
| Sample Size (n) | 6.804 | 7.429 | 18.819 | 3.456 | 1.634 | 18.759 |
| Pseudo-R2 | 0.28 | 0.29 | 0.26 | 0.46 | 0.29 | 0.21 |
Data source: MICS (2014)
Notes: Current Attendance =1 if person reported attending school during 2013-14 school year. Additional predictors (not presented) include control for month of survey and region controls. T-statistics (parentheses) are corrected for clustering at sample cluster level. **Coefficient significant at p<=0.01 level; *Coefficient significant at p<=0.05 level; +Coefficient significant at p<=0.10 level
Muslim children are significantly less likely to be reported as being in school in MICS, and with less completed education. For language, the results show that children from Fula and Mandiga households are less likely to be in school, and complete 3-5 years less education than Criolo-speaking youth.
Gender gaps have been shrinking in Guinea-Bissau, but Muslim females are significantly less likely to be in school than non-Muslim females, whereas for males there is no significant difference by religion.
OOSC Barriers: Mixed methods evidence
There are multiple factors explaining school exclusion and drop-out. Here, we only focus on a reduced set of factors that stand out in the qualitative survey and policy discussions.
Low school quality and grade repetition
The interview data paint a picture of public schooling with deep-seated systemic problems that are recognized by respondents of all ages and background, including institutional weaknesses and government spending inadequacy.
The scope of grade repetition in Guinea-Bissau is unusual, even by regional standards where grade repetition is prominent (UNESCO 2013). Summaries from 2014-15 show repetition rates above 15 percent in the entire grade 1-grade 9 basic education cycle, and roughly 15 percent in grades 10 and 11 of upper secondary. Furthermore, these rates have increased substantially in the last 10 years. Qualitative data collection suggests that grade repetition decisions can be somewhat arbitrary and not necessarily linked to poor academic results.
Several barriers are underlying causes of late entry, but the first is school supply. About 40 percent of students in primary grades 1 and 2 are studying in incomplete schools that only offer grades 1-4[1]. A number of parents expressed a preference to wait to enrol their children until they are considered mature enough to face the challenges posed by the journey to a far-away school.
Economic barriers
One issue that stands out in Guinea-Bissau is school fees. Public education is officially free until grade six since 2011. However almost all parents complained about not being able to pay for the school fees during the fieldwork interviews. Some children have to drop out of school for a year or two, and re-enroll when their family can afford it, which explains part of the overage issue.
Child labour is another frequently mentioned influence on school attendance. MICS (2014) data show that children in the 5-11 age range report 8-10 hours per week of work. But the average working hours during the school year may tell only part of the story. Cashew harvest-related labour was the most frequently cited cause of dropout during the fieldwork interviews (across all regions and respondents).
When children go back to school after the harvest, they may fail end of year exams or are not accepted back at school.
Religion and exclusion
Frequent references are made to religion in discussions of school attendance and OOSC in Guinea-Bissau. The system does not take into account all schools: religious schools (Catholic schools and Islamic madrassas that use an officially-recognized curriculum) are counted as part of the formal education system, but Madrassas and Koranic centers are considered as part of the non-formal education offer. So, the central question is therefore the degree to which koranic centres and, to a lesser degree, madrassa schools act as complements or substitutes for non-religious public school options.
There is evidence of a small group of children who are effectively excluded from formal schooling by attending only a koranic center:
- Local research shows that Koranic centres tend to be free of charge for parents, but always include some sort of child labour (in the fields, begging, etc.).
- Despite these concerns, the majority of community leaders, parents, and religious leaders in Muslim communities were positive about religious education and their objective is to have children attend madrassa/koranic centres and public schooling in the same day when possible.
Gender
Female school attendance rates have improved in the last 15 years, but the country is still a long way from reaching gender parity in basic education. Attendance begins to significantly diverge around the age of 14 or 15, and the gender gap becomes very large by age 18.
The most frequently cited cause of girls dropping out of school was due to marriage and/or pregnancy, especially in the Muslim regions and in rural areas in general. Interestingly, participants used the terms “casamento forçado” (forced marriage) and “casamento precoce” (early marriage). For girls who want to combine schooling with raising a family there are structural barriers and stigmatization to overcome: most schools refuse to allow pregnant teenage girls to continue their schooling. MICS data analysis shows that both females and males are less likely to be in school when they report having children. But married females are 24 percent less likely to be in school, whereas for married males there is no significant impact.
If the issue of early marriage and pregnancy is relevant in Guinea-Bissau, it is important to put it in proper context in terms of how it impacts out-of-school children. Less than 10 percent of 15-17-year-old girls report being married or having children in the 2014 MICS, although marriage and pregnancy rates increase substantially beginning at age 18. However, in Guinea Bissau, the massive age-grade distortion means that at this age, girls are still enrolled in school, which is why pregnancy and marriage are often mentioned as reasons for early drop-out.
Policy Recommendations
For tackling out-of-school children directly the first steps include developing a national strategy and policy for inclusive education with attendant mapping of uncovered populations. This should cover:
- Actions directly targeting girls’ enrolment and education
- Incorporating Koranic centres into mainstream (formal) schooling.
Beyond reaching the remaining groups of out-of-school children the most pressing challenges facing the education system are improving efficiency and grade attainment, through tackling the following entry points:
- Preschool expansion is already contemplated in the current Education Sector Plan and should be scaled-up.
- Improving school supply to reduce commuting time and provide possibilities for school continuity.
- Address late entry head-on with community awareness campaigns, with support from the communities and religious leaders.
- Rethink the current grade repetition policy with a goal towards reducing the percentage of repeaters in the system.
- Address teachers poor working conditions to avoid strikes (teacher motivation framework, mobile payment, etc.)
- Encourage clarity on school fees, making sure that no fees are collected in public schools when they shouldn’t.
- Encourage communities’ involvement in the education sector, through autogestão schools – public schools that have more autonomy to raise fees and pay teachers, which allow school to be improved through local initiative.
- Support older students who are most vulnerable to dropping out, through conditional cash transfer, accelerated learning programs
- Adjust the school calendar so that long holidays are scheduled during the cashew harvest season.
References
Klute, Georg, Birgit Embaló, and Augusto Idrissa Embaló. 2006. “Local Strategies of Conflict Resolution in Guinea-Bissau : A Project Proposal in Legal Anthropology.” Recht in Afrika (2):253-272.
Merchant, C. Melissa, Angela Demas, Emily Elaine Gardner, and Myra Murad Khan. 2018. Guinea-Bissau School Autonomy and Accountability: SABER Country Report 2017. Systems Approach for Better Education Results. Washington: World Bank.
Ministère de l’Education Nationale. 2015b. Rapport d’état du système éducatif. Pour la reconstruction de l’école Bissau-Guinéenne sur de nouvelles bases. Bissau: Ministère de l’Education Nationale.
UNESCO. 2013. Guiné-Bissau – Relatório da Situação do Sistema Educativo. Margens de manobra para o desenvolvimento do sistema educativo numa perspetiva de universalização do Ensino Básico e de redução da pobreza. Dacar: UNESCO.
UIS. 2019. Fact Sheet no. 56: September 2019 – UIS/2019/ED/FS/56. Montreal: UIS.
UNICEF. 2015. Global Out-of-School Children Initiative Operational Manual. New York: UNICEF.
[1] EMIS data from 2014-15
[1] Inquêrito Ligeiro de Avaliação da Pobreza
[2] Conducted in 6 provinces: SAB, Cacheu, Oio, Tombali, Bafata and Gabu
