This paper establishes a new fact about educational production: ordinal academic rank during primary school has long-run impacts that are independent from underlying ability. Using data on the universe of English school students, we exploit naturally occurring differences in achievement...