Traditionally, pay analysis in the public sector has been based on cross section data, such as average or median wages. This study differs in that micro longitudinal data are used to explain and compare pay determination in the French and Italian civil services. Based on examples, the report presents different readings obtained from aggregate data and individual longitudinal data. It focuses on the weight of various components of human capital -- gender, age, year of birth, seniority, educational level -- on trends in civil servants' pay throughout working life. This type of analysis sheds new light in the area of human resource management in the civil service and, in particular, may be seen as a key tool for analysing gender pay discrimination.