Title: Comparing standardized tests and teachers-based-assessments; Evidence from the A-levels fiasco in the UK.

Abstract: The distributional consequence of integrating teachers' assessments with standardized exam scores for college admissions remains challenging to assess. I focus on a unique event in the UK, when the pandemic completely replaced standardized exams with teacher-based assessments of their students, to investigate the changes in the pattern of grade assignment across students and schools. I find the method change led to substantial inflation; students' expected grade increased uniformly across groups with similar performances in a past standardized exam. Non-facilitating subjects experienced higher grade inflation than facilitating subjects. Institutional differences across school types and share of well-performing students also account for differences in the extent of inflation. My findings indicate an inclusion of teachers' predictions to college admissions reward students unevenly.