Is Private Education Worth it? Evidence from the Free Primary Education Reform in Kenya

Abstract: In 2003, Kenya introduced a nationwide Free Primary Education (FPE) reform, abolishing school fees in all public primary schools. As a result, enrolment rates in public primary schools rose by 15 percent, putting significant pressure on the educational system. Overcrowding and lack of school material caused many parents to turn to private school alternatives. Using a nationally representative cross- sectional household survey, I exploit intra-household variation in terms of school enrolment to measure private versus public school effectiveness in terms of math, English and Swahili test scores. My findings suggest that, on average, private school students score 18, 23 and 21 percentage points of a standard deviation higher than public school students, respectively. As my results are likely to be exposed to a selection bias stemming from high-achieving students being sorted into fee-charging private schools, I include household fixed effects and an extensive set of household-related controls. Moreover, I prove that when attempting to control for such sorting effects, half of the effect disappears. I thus argue that the remaining effect can be interpreted as evidence of private schools being more effective than public schools.