What Determines Entry Into Top Civil Service?

Abstract:     
This paper examines the quality of the evaluation of 40,000 candidates to top civil service positions in the Spanish Administration. The evidence casts doubts on the design of the evaluation system. First, randomness abounds: candidates' success rates vary significantly depending on factors unrelated to their quality such as the day of the exam, the hour of the exam, a potential deferral and, specially, the order of evaluation. Second, the evidence is consistent with the existence of endogamy, sexism and localism in the evaluation. Candidates who are relatives of current civil servants tend to be more successful in non-anonymous tests, but not in anonymous tests. Moreover, success in civil service exams depends on the gender composition and geographical origin of the evaluation committee. We propose to reform entry exams in three ways: (1) anonymity and double external examination should be used in all written exams; (2) a preliminary, qualifying multiple choice test should be introduced; and (3), a supervision committee should be created to improve accountability in hiring.

Note: Downloadable document is in Spanish.

Number of Pages in PDF File: 60

Keywords: Evaluation Bias, Discrimination, Civil Service

JEL Classification: J45, J70, H83

working papers series