Notes: Each column represents a separate regression, estimated using SDI data, with flexible controls for lagged test scores and district-by-year and school fixed effects. Coefficients correspond to the effect of being ranked in the associated decile of within-district performance in the postreform period, compared to the middle six deciles. The SDI data were collected in 2014 and 2016 (the postreform period), but some variables were collected using the previous year (2013 or 2015) as the reference period. For each column, only two years of data are available: Columns 1, 4, and 5 use outcomes for the years of 2014 and 2016, and Columns 2 and 3 use outcomes for the years 2013 and 2015. The dependent variables in Columns 2 and 3 are inverse hyperbolic sine transformations (an approximation for the natural logarithm) and calculated at a per-student level, using enrollment data from 2014. Data from Column 3 are reported in Tanzanian shillings. The mean values reported in the penultimate row is of the untransformed outcome. We adjust for outliers in the following way: (i) we adjust downwards the per-student capitation grant to the maximum that a school can receive, 10,000 Tanzanian shillings; (ii) we set as missing one school that reported receiving 600 textbooks per student. Since the specifications include school fixed effects, schools with only one observation are dropped. Standard errors are clustered at the district level.