Are Refugee Children Learning? Early Grade Literacy in a Refugee Camp in Kenya

Benjamin Piper, Sarah Dryden-Peterson, Vidur Chopra, Celia Reddick, and Arbogast Oyanga

Journal on Education in Emergencies, Volume 5, Issue 2 (2020), Pages 71-107


This study assesses literacy outcomes for refugee children in lower primary schools in Kakuma refugee camp in Turkana County, Kenya. The authors also examine how the literacy outcomes for refugee children in Kakuma compare with those of Kenyan nationals outside the camps, how literacy outcomes vary by refugees’ country of origin, and what policy-relevant factors are associated with literacy outcomes.

The analysis is based on data collected in March 2018 from the Early Grade Reading Assessment (EGRA) for children in grades 1-3 in all 21 schools in Kakuma refugee camp that have lower primary classes, and from two schools in the adjacent, more recently established Kalobeyei settlement. Simple quasi-random sampling was used at the school level; the sampling strategy targeted 10 students (5 male and 5 female) in grades 1, 2, and 3 in each school, for a total of 30 students per school. All the schools follow the national curriculum of Kenya, have both Kenyan and refugee teachers, and teach using English and Kiswahili, Kenya’s official languages.

Main findings:

  • Early literacy outcomes for refugees in Kakuma were exceedingly low. Grade 2 students in Kakuma scored below students in the national Kenya baseline on all fluency measures in both English and Kiswahili. Only 8.6 percent of grade 3 students in Kakuma met the Ministry of Education’s grade 2 benchmark for reading fluency in English and Kiswahili, and average oral reading fluency rates were some of the lowest seen in any study in lower- and middle-income countries. In comprehension skills, which are critical predictors of later academic success, grade 3 students in Kakuma scored substantially below grade 2 students in the national Kenyan baseline (4.7 percent correct compared to 22.0 percent correct).
  • Oral reading fluency in English varied across cohorts of students from different origin countries. Somali students tended to score higher and students from South Sudan tended to score lower. Other research suggests that refugee students’ country of origin may influence learning outcomes, possibly due to family literacy rates, ease of connection to schools and school culture, previous educational experiences in the country of origin, and length of stay in Kenya. For example, Somali refugees have been in Kenya longer, and students are more likely to have siblings or parents who were educated in English in Kenya.
  • Students’ assessed reading fluency was higher if the language of assessment was the primary language used for instruction at their school. According to Kenyan policy, students should be exposed to English as a primary language of instruction, however teachers of refugees sometimes do not have the language skills to instruct in English or make the decision to use other languages so their students can understand the lessons.
  • Expectations of returning to their home country were associated with oral reading fluency. One might expect that students who predicted that they would continue their education in Kenya would invest in English, however students who expected to be in Kakuma three years later had lower reading fluency in English. A possible explanation for this counterintuitive finding, is that restrictions on refugees’ rights to move freely and work in Kenya could limit the perceived usefulness of a Kenyan education. The results also showed that children who planned to be back in their country of origin ten years in the future had lower English fluency.

The authors identify several implications of relevance to policy makers and researchers:

  • The lower learning outcomes in Kakuma compared to elsewhere in Kenya underscore the necessity for further research on the different learning needs of refugees, which may not be met entirely by the national education system and may require adjustments to national education systems to meet differentiated learning needs.
  • A better understanding of students’ educational histories, their parents’ educational histories, and their exposure to the languages of instruction could inform policy responses and instructional practices for students from different countries of origin.
  • Teachers’ in Kakuma could be trained in ‘translanguaging’ practices that enable them to capitalize on languages shared with students, such as Arabic or other home languages, while at the same time exposing students to English and Kiswahili, which, as languages of instruction and assessment, are critical to their educational success.
  • It is also important to further explore how refugee students and their families perceive the value of their education in exile, given their expectations for their futures, how this perception overlaps with the languages used in their country of origin, and how these perceptions and plans affect their investments in schooling and learning.