When Private Beats Public: A Flexible Value-Added Model with Tanzanian School Switchers

Despite the increasing popularity of private secondary education in sub-Saharan Africa, little is known about the return. In this paper, I estimate a private school learning premium in Tanzania, using administrative exam records for 635,000 students. I compare secondary school students with their primary school classmates who achieved the same primary school exam scores and control for peer effects and unobserved ability. On average, private schools improve exam scores by 0.54 of a standard deviation in 2 years. A regression discontinuity design suggests that the effect is causal, and subject-specific estimates are all positive but higher for mathematics relative to Kiswahili and English.