This paper analyzes the school integration of Syrian refugee children in Turkey. Syrian refugees began fleeing to Turkey in 2011. Turkey currently hosts 3.7 million Syrian refugees, of which 47 percent are under 18 and a third are school aged (aged 5–17).
Initially refugees were hosted in camps, where refugee children attended schools following the Syrian curriculum and taught in Arabic. Camp schools were later turned into Temporary Education Centers (TECs) and established in out-of-camp areas as refugees gradually moved out of camps. From 2014-2015, Syrian refugee children were able to enroll in public schools, and in 2019-2020 TECs were closed, and students transitioned to public schools.
The analysis draws on data from the Turkey Demographic Health Survey (TDHS), which, for the first time in 2018, included a separate module (the TDHS Syrian Migrant Sample, TDHS-S) representative of the Syrian population in Turkey. The TDHS includes detailed data on the demographic and socioeconomic characteristics of children and their households. Data reveal:
- A substantial native–refugee gap in school enrollment. The average school enrollment rate is 90 percent for native children (ages 7–17) but only 64 percent for Syrian refugee children. While 1 percent of native children have never been to school, 11 percent of refugee children have never been to school. Syrian refugee children have lower enrollment rates, regardless of age and gender, however the gap widens after age 12.
- A substantial portion of refugee children are working. While only 5 percent of native children (ages 12–17) are in paid employment, 20 percent of Syrian refugee children are working. Almost a third (32 percent) of Syrian boys are in paid employment.
- Syrian children are more likely to be married. Five percent of refugee children (ages 12-17) have been married compared to less than one percent of native children.
- A larger proportion of Syrian children, particularly girls, are in NEET (neither in employment nor in education and training).
- Refugee children are 5 percentage points (pp) less likely to progress to the next grade level. Of this difference, 4 pp comes from a higher likelihood of grade repetition, and the remainder is due to a higher likelihood of school dropout. Refugee boys are about 4 pp less likely to progress than refugee girls.
- Most refugee children come from poor households. Over 80 percent of refugee children are in the bottom quintile for household wealth compared to 12 percent of native children. Syrian children are also more likely to live in crowded households, have younger and less educated household heads, and to have deceased parents compared to native children.
Main results from empirical analysis:
- There are significant differences in school enrollment between native and refugee children. In the baseline specification, Syrian boys are 29 pp less likely to be in school than native boys, while Syrian girls are 22 pp less likely to be in school than native girls.
- A substantial part of the difference in school enrolment of Syrian and native children (about half for boys and two-thirds for girls) is explained by differences in socioeconomic characteristics between Syrian and Turkish households. After accounting for socioeconomic characteristics, the baseline native–refugee gap reduces to 14 pp for boys and 8 pp for girls. Restricting the sample to children who arrived in Turkey at or before age 8 and accounting for socioeconomic differences, the native–refugee gap disappears for both boys and girls.
- Refugee children are less likely to be enrolled in school during their first year of residence. The refugee–native gap in predicted enrollment is 44 pp for refugee boys and 37 pp for refugee girls during their first year in Turkey.
- Refugee boys are more likely to be in paid employment than native boys, but the gap reduces significantly after socioeconomic differences between natives and refugees are considered. In the baseline specification, refugee boys are 25 pp more likely to be in paid employment than native boys, while refugee girls are 5 pp more likely to be in paid employment than native girls. Accounting for socioeconomic characteristics of households, the gap between native and refugee boys reduces to 18 pp and vanishes for girls. Restricting the analysis to refugee children who arrived in Turkey at or before age 8, the refugee–native gap in paid employment among boys also disappears.
- There is a strong association between the timing of school dropouts and the timing of labor market entry for refugee boys, suggesting similar factors drive these decisions.
- The timing of marriage does not seem to correspond with the timing of school dropouts among refugee girls. When marriage rates rise notably after age 15, there are no notable changes in enrollment rates, suggesting that girls who marry are already out of school.
- The native–refugee gap in school enrollment stems from refugees’ higher propensity for never enrolling in school. Refugee boys and girls are 10 and 7 pp more likely to never enroll in school at the baseline, respectively; and these gaps reduce to 5 and 3 pp once socioeconomic characteristics are accounted for.
- Accounting for socioeconomic differences between refugees and natives, refugees are no more likely to drop out than natives once they are enrolled in school.
- Refugee boys are less likely to progress to the next grade compared to refugee boys, despite being a more select group due to their lower enrollment rates. Conditional on enrollment, refugee boys are 5 pp less likely to progress to the next grade than native boys, whereas the native-refugee gap among girls is much smaller and not statistically significant.
- Both refugee boys and girls lag natives in terms of grade for age (about 0.7 grades), even after accounting for socioeconomic differences. For both genders, the lower likelihood of grade progression conditional on enrollment results from a higher probability of grade repetition rather than dropout. In terms of grade for age, refugees are about 0.7 grades behind both boys and girls, even after accounting for socioeconomic differences.
The results highlight the importance of age at arrival for refugee children’s school integration. Arrival just a few years after school start age is enough to cause deficits in educational attainment compared to native children. Differences in the socioeconomic characteristics between Turkish and Syrian households explain a sizable proportion of the differences in children’s school outcomes. Household wealth is a crucial factor because poverty pushes children into employment. Evidence for this is only apparent for boys, as employment data is limited to paid work, however the opportunity cost of refugee girls’ school enrollment is presumably also high due to their contribution to household chores and employment as unpaid family workers.