Ahead at the Start: Selection and the Private–Public Achievement Gap in French Primary Schools

Whether private schools raise achievement or reflect selection remains an open question. Using a national panel of 15,188 French children followed from first grade to the end of primary school, with standardized mathematics and French assessments at both points, we estimate the effect of attending a publicly funded private school throughout primary school. Private-school students outperform their public-school peers at both school entry and the end of primary school. We compare three complementary identification strategies: a value-added specification controlling for baseline achievement and family background, an instrumental-variables strategy exploiting local private-school supply, and within-student comparisons of sector switchers. Accounting for selection eliminates the French advantage and reduces the mathematics advantage by about three quarters, leaving an estimated effect of 0.080 standard deviations. The value-added and instrumental-variable estimates point to the same conclusion, with the former estimated much more precisely. Sector switchers reveal a distinct selection pattern: more advantaged families move initially lower-achieving children into the private sector, and these children finish primary school ahead of comparable public-school students. The only consistent exception arises in large urban markets, where mathematics effects remain positive across specifications, although these are also the settings in which both admissions discretion and potential threats to the instrumental-variable strategy are likely to be greatest. Overall, most of the observed private-school advantage reflects selection rather than school effectiveness, although effects may differ across local education markets.