After one year, outsourcing the management of ninety-three randomly-selected government primary schools in Liberia to eight private operators led to modest learning gains. In this paper, the authors revisit the program two years later. After the first year, treatment effects on learning gains plateaued (e.g., the intention-to-treat effect on English was .18σ after one year, and .16σ after three years, equivalent to 4 words per minute additional reading fluency for the cohort that started in first grade). Looking beyond learning gains, the program reduced corporal punishment (by 4.6 percentage points from a base of 51%), but increased dropout (by 3.3 percentage points from a base of 15%) and failed to reduce sexual abuse. Behind these average effects, the identity of the contractor mattered. Despite facing similar contracts and settings, some providers produced uniformly positive results, while others present stark trade-offs between learning gains, access to education, child safety, and financial sustainability.