Merit-based Recruitment and Government Performance: Evidence from Random Assignment of Administrators

Abstract

How to recruit managers is a central question for public sector organizations. Exploiting a unique random assignment of administrators to counties in imperial China, we study how merit-based recruitment of managers affects government performance in multiple tasks. Combining personnel records and administrative archival data, we find that merit-recruited administrators through civil service exams (as opposed to other non-merit-based recruitment) performed better in public goods provision and tax collection, and were more likely to be recognized for their achievements by the local public. Evidence suggests that these administrators were associated with better management of subordinates. Examining their dual role as judges, we use criminal case data to show that cases judged by merit-recruited administrators had shorter durations, indicating enhanced judicial efficiency. Finally, merit-recruited administrators were less likely to come from elite families and have connections to incumbent top officials. Our results highlight the importance of competitive and open recruitment as a cornerstone of effective public organizations.