The learning divide
Understanding the state of learning in forced displacement contexts - Mexico and Mauritania
Apr 4, 2025
Learners around the globe are in learning poverty
In 2022, the World Bank estimated that around 70% of children aged 10 in low-and middle-income countries could not read and understand a simple text, an increase from 57% in 2019 (World Bank, 2022), highlighting that the world was experiencing a learning crisis. These increases occurred partly as a result of school closures that took place during the COVID-19 pandemic, but the projected rise in learning poverty was different by region, with Latin America and the Caribbean seeing the most notable increase.
In forced displacement contexts, we do not know if children are learning
We do not know what proportion of forcibly displaced learners have literacy and numeracy skills. The importance of reporting on this is highlighted in SDG4, indicator 4.1.1 which calls for countries to report on the proportion of children and young people achieving at least a minimum proficiency level in (i) reading and (ii) mathematics, by sex. We know even less about whether these learners have social and emotional skills, demonstrated not only to be necessary for the acquisition of literacy and numeracy but also essential for success in later life.
A refugee child in the Mbera camp in Mauritania takes a learning assessment with a trained facilitator. ©RET International
The objective of the Learning Measurement project is to fill this data gap to effectively inform education program design and policy to address learning for refugee children. We want to know: To what extent do forcibly displaced children have literacy and numeracy skills? To what extent do they have social and emotional skills? To find out, we conducted two learning assessments in the contexts of Mauritania and Mexico.
The findings in both countries tell us a complex story about the state of learning in these very different contexts.
Mauritania
In July 2024, UNHCR and RET International conducted an educational assessment in Mauritania. The Holistic Assessment of Learning and Development Outcomes (HALDO), developed by Save the Children, is an interactive assessment tool that measures learning progress among populations impacted by crises.
people live in Mbera camp. More than half of these refugees are children (2024)
primary and secondary school aged refugee children as a percentage of the total Mbera camp population (2024)
refugee primary gross enrollment rates (2023)
refugee secondary gross enrollment rates (2023)
The graphic at right helps to visualize the results of the Mauritania assessments. Each bubble represents 1 per cent of the students. There are 100 bubbles for both Grade 4 and Grade 6.
The assessment results for refugee children in both 4th and 6th grades, can be separated into five categories:
Level 1 Level 2 Level 3 Level 4 Level 5
By clicking the tabs at the top of the graphic, you can switch between the results of the literary assessments and numeracy assessments.
Students graded as Level 4 or Level 5, are considered to have achieved minimum fluency.
Of these, a smaller percentage achieved Level 5. In the literacy assessment, these means they can read fluently and correctly answer 4 or 5 reading comprehension questions. For the numeracy assessment, level 5 implies that the students could perform two-digit addition and subtraction. It is important to note that level 5 implies grade level fluency.
Only 9% of Grade 4 students achieved level 5 in both reading and math. A larger percentage achieved level 5 in Grade 6 in both reading (29%) and math (30%). This shows a noticeable improvement as the students get older.
On the other hand, the assessments show that the majority of children are not achieving basic fluency.
These levels represent a range of learning. Level 3 literacy means students can identify infrequently used letters, but struggle to read the first five words in the given text. For numeracy, level 3 indicates that students can recognize infrequent numbers, but cannot do basic math operations.
Level 2 in literacy means students can identify common letters, but not uncommon ones. In numeracy, it represents students who can identify simple numbers but not harder numbers.
Level 1 is the most concerning. In literacy it implies that students cannot name five commonly used letters. This applies to 14% of grade 4 students and only 5% of grade 6. In math, level 1 means that students cannot identify five basic numbers nor answer any word problems. This category includes a staggering 37% of grade 4 students, and 28% of grade 6.
It should be noted, a large portion of students in both grades are over age, and not at the appropriate age for their grade level. A positive finding was that Grade 6 students consistently outperformed their Grade 4 peers, indicating that some learning is taking place in the transition from Grade 4 to 6. Lastly, none of variables on socio-economic background of the students, such as possessions in the home, are predictors of assessment scores, except having books at home.
A facilitator conducts activities with children in Mbera camp. ©RET International
The HALDO instrument includes elements of Socioemotional learning (SEL), which “refers to a child’s growing capacity for independence and confidence in a range of routine activities.” The SEL section begins with five basic questions about the child’s self and environment, including full name, age, etc. All students scored above 90 percent. The high scores on this content are not surprising for students in this age range.
Questions 6-11 of the SEL assessment become more subjective and ask students to consider their future aspirations. The sequence begins with Q6 where the students are asked to name one thing that they “hope/wish will happen in your life in the future”. The progressively lower scores in questions 7-11 suggest that about half of the overall sample have difficulty in contemplating the future, what may happen to them and what kinds of factors may impact their lives.
Mexico
In Mexico, another learning assessment was given to children in Tijuana, Tapachula and Monterrey. The assessment relied on a sample of 443 children, aged 4 to 18. 75% of the sample are out-of-school children or in non-formal education setting. Data collection includes children from 15 different countries of origin, although a majority of them came from the Northern Triangle of Central America and Haiti.
asylum seekers (2023)
of asylum seekers are children (2023)
refugee primary gross enrollment rate (2024)
refugee secondary gross enrollment rate (2024)
HALDO on the Move was used for the assessment in Mexico. The tool is an adaptation of the original HALDO. It is designed to assess children from 5-18 years old and tailored to assess the learning needs of children on the move, particularly those observed along the Central American migratory route as well as Mexico. For this assessment in Mexico, children aged 4 were also included.
Like the results from Mauritania, children in all age groups perform better with easier questions but struggle as questions become more demanding.
A trained facilitator conducts a learning assessment with a child in Tijuana, Mexico. ©Save the Children
Age is a key predictor of children’s performance for all domains, while gender is not a key predictor. Children from Haiti perform worse in various domains which may be due to the language barrier. The tool administration was done in Spanish while their native language is Creole and French.
Across all domains, there is significant differences in performance between 4-8 age group and the older age groups. However, there is little difference between 9-13 and 14-18 age groups. This shows learning development is stagnant between early and late adolescents.
Tests were given to students in three age groups. The graphic at right will show the results in both literacy and numeracy. You can switch by clicking the button at the top of the visual.
As with Mauritania those who scored the lowest are considered as Level 1. For reading this means the students could not identify letters or combine syllables. Alarmingly, 68% of children aged 4-8 fall into this category. For math, it means that students cannot grasp the concept of numbers.
Level 2 in reading means the student can combine vowels and consonants but cannot read whole words. In math it means an ability to count objects or identify simple numbers.
Level 3 reading marks an ability to read words but not sentences. Level 3 math indicates students who can identify larger numbers, but cannot do simple operations (one-digit addition or subtraction).
Students achieving level 4 in reading can read, understanding simple sentences, but cannot demonstrate reading comprehension. In math, level 4 signals students who can do simple operations, but not harder ones.
Those at level 5 are the highest level. Only 11% of 4-8 year readers reached level 5 which means they can read and comprehend written texts. The percentages of those who reach this level for 9-13- and 14–18-year-old age groups are much higher at 61% and 60%, respectively, although this is not advanced considering the age and expected grades.
In numeracy, only 13% of those aged 4-8 can reach the highest level which means they can do harder operations (i.e. two-digit addition, subtraction or one-digit multiplication). Notably, the percentages for those 9-13 and 14-18 are substantially higher, at 65% and 73%.
SEL performance also shows similar trend as literacy and numeracy. Children struggle to answer the questions as they become more demanding.
34% of age 4-8 can reach two highest levels of the HALDO SEL question, which mean they can demonstrate empathy and rationalize others’ actions or through indirect information. These percentages are 51% and 42% for 9-13 and 14-18 age groups, respectively.
Key messages
- Levels of learning are low and students lack the basic competencies they should have, but some learning progression seems to be occurring.
- It is important to generate data to inform effective program design and policy to support improving quality of education for refugee children.
- Increased investment and prioritization of resources for learning is crucial to help forcibly displaced children obtain the skills they need. Including components of socio-emotional learning in the curriculum and/or advocate for the importance of SEL to the national authorities is essential to ensure holistic learning.
For more information on this study, check out our blogs of the assessment in Mauritania and Mexico, and our technical reports here.